The Russian Problem and the Onset of the Pacific War, March−December 1941
THE PREVIOUS TWO CHAPTERS have shown just how important the rise of Russian power was to Japanese leaders of all stripes from 1895 to 1940. Chapter 4 in particular discussed Tokyo’s preparations after 1935 for an all-out war with Russia, and how those plans got diverted by an undesired war with China. This chapter will focus on the nine months leading up to the attack on the United States on December 7, 1941. The puzzle is a simple one: Why did Japanese officials, despite their obsession with reducing Russian strength and recognition of the unlikelihood of winning any war with the United States, decide to take on the US giant rather than find a negotiated peace that would have avoided a costly and risky war? This riddle becomes even more profound once we realize just how earnestly both sides sought a peaceful way out. On three separate occasions—in early June, early September, and late November 1941—Tokyo and Washington almost came to an agreement that would have maintained the peace. Yet something kept getting in the way. What was this something, and why did it prove so intractable?1
For pretty well every international relations theorist who has studied this question, the primary obstacle to peace was Japan’s unwillingness to leave China and what this said about Tokyo’s larger obsession with the building of a regional coprosperity sphere. This view nicely aligns with the commercial liberal argument, insofar as domestic and psychological pathologies within the Japanese state can be said to have been unleashed during a period of reduced bilateral commerce, especially after 1939 and the imposition of increasingly severe US trade restrictions.2
As this chapter demonstrates, this contention is almost certainly wrong. For one thing, right from the start of talks in March–April 1941, the US government was aware that Japan did want out of China and that it desperately sought Washington’s help in finding a face-saving deal with Chiang. In return for US help, Roosevelt and Hull wanted one thing above all: a Japanese promise not to start a war somewhere else in the Pacific region, either against the United States and Britain in the south or the Soviet Union in the north. The primary objective of US policy was to defeat Hitler in Europe, and Roosevelt and Hull did not want Japan opening a second front in the Pacific should the United States enter the war against Germany. Thus Washington kept pushing Tokyo to agree to ignore its alliance obligations with Hitler, even if the United States was seen as the initiator of war against Germany. More broadly, and especially after Germany attacked the Soviet Union in June 1941, Roosevelt and Hull wanted Tokyo to commit to not attacking the north in support of Hitler’s eastern campaign. After June 22, it was this latter requirement that became the key sticking point in all the negotiations. Tokyo wanted to maintain the option of going north if the Soviets were about to be defeated by Germany. The Japanese leadership therefore could not agree to Washington’s final condition for peace, even if this meant the continuation of harsh economic sanctions against Japan.
As this chapter reveals, the traditional argument suffers from a profound irony. The United States had to worry that Japan would indeed get out of China, perhaps with US help, and that hundreds of thousands of battle-hardened Japanese troops would then be redeployed to the Siberian front for an attack on Russia. If this happened and Stalin was forced to fight a two-front war, the Soviet Union might lose its war with Germany. This would leave Hitler in a commanding position in Eurasia—the worst of all outcomes. Soviet survival, in short, was critical to Roosevelt’s whole war plan, and the Japanese could not be allowed to interfere with it. Keeping Japan bogged down in China, as the US leadership well understood, directly served US interests. Roosevelt and Hull were willing to help Japan exit China only on one condition: that Tokyo make a credible commitment to not going north or south. In the end, it was Japan’s unwillingness to commit to not attacking Russia—rather than any reluctance to leave China per se—that got in the way of a deal to avert war.
From a theoretical standpoint, this chapter supports the assertion from chapter 1 that third-party concerns can cause less dependent states to restrict trade flows with more dependent states, driving the latter into hard-line behavior and war. In this case, the highly dependent Japanese would have preferred to avoid a war with the United States. But this was only because Tokyo’s main priority was to launch a decisive preventive war against Russia. From 1938 to early 1941, Roosevelt used trade sanctions on military equipment, iron ore, and aviation fuel to restrict the growth of Japanese military power without directly provoking Tokyo into a war. He was well aware that going directly after Japan’s oil supply might push Japanese leaders into the ultimate act, and he wanted to be able to focus on the European war without the bother of a Pacific conflict. The sanctions gave Washington leverage: if Tokyo would commit to not redeploying forces in China for war against either Russia or the United States, Roosevelt and Hull in return would mediate with Chiang to help Japan get out of China, and would relax the economic restrictions. After Hitler’s attack on the Soviet Union in June, however, the priority was to interfere with Japan’s buildup against Russia and draw Japanese military resources south. This would allow Stalin to transfer the bulk of his crack Siberian forces to the German-Soviet front. The complete oil embargo of July 1941 against Japan was thus part and parcel of a larger strategy to help Russia avoid a two-front war, even as it gave Washington additional leverage in the negotiations. As I show, the traditional view—that Roosevelt was unaware of the plan to impose a total oil embargo and, due to bureaucratic momentum, could not reverse it once he found out—is a misinterpretation of the evidence. Roosevelt was intimately involved in the planning and fully informed of what was in fact an ingenious strategy to enhance Washington’s flexibility in its subsequent dealings with Tokyo.
Ultimately, then, it was Washington’s unwillingness to relax devastating economic sanctions that drove Tokyo to initiate a war south against the United States instead of north, as the Japanese preferred. Highly pessimistic trade expectations combined with extreme dependence on raw materials pushed Japan into war on December 7, 1941. Yet the US reluctance to make a deal reflects Roosevelt and Hull’s keen sense that such a deal would send Japan north against the only US ally that stood a chance of reversing Hitler’s bid for Eurasian hegemony. The United States would have preferred a deal committing Tokyo to peace both north and south. But if it could not get that—and Japanese leaders continually refused to promise not to go north—then a policy of oil sanctions drawing Japan south was preferred to a deal that left the United States out of a Pacific conflict, but only at the cost of having Japan attack United States’ most important ally and key to the eventual defeat of Nazi Germany. In short, Roosevelt and Hull were forced to choose the lesser of two evils to prevent Japan from going north. And because they did, the Japanese, in the face of such pessimistic trade expectations, had to attack south to avoid economic collapse.
In what follows, we will see that neither the liberal nor realist perspective on trade and conflict is able to explain the failure of diplomacy along with the outbreak of war. The liberal focus on the pathologies unleashed as trade falls can explain neither Tokyo’s effort to get out of China through US-mediated diplomacy nor its effort to avoid war with Washington in 1941. It also cannot explain the almost exclusively security-driven discussions within the Japanese state for the year before the war. Economic realism captures the strong feelings of vulnerability caused by raw material dependence that were shared by essentially all Japanese political and military elites. But it cannot capture the ups and downs of US-Japanese diplomacy in 1941, nor why Tokyo would try so hard to avert a war. As will become apparent, the fluctuations in relations over 1941 directly reflected Japanese perceptions of Washington’s willingness to relax its economic sanctions. It was only after the clear failure of negotiations in November 1941 that Japanese trade expectations became so pessimistic that war seemed the only logical way out.
The chapter proceeds as follows. I begin with a look at US-Japanese talks in spring 1941. For those who see Japan’s “unwillingness” to exit China as the underlying cause of the Pacific War, there is a surprise. Not only did Tokyo and Washington come close to an agreement that would have kept the peace but the United States also was prepared to sacrifice Chinese interests in the process. To get Tokyo to commit to not assisting Germany should the United States enter the European war, Washington was willing to help it negotiate a deal with the KMT that would have allowed Japan to hold on to Manchuria and maintain Japanese troops in northern Chinese provinces. In the second section, I explore why the US-Japanese agreement fell apart in late June 1941. After Germany invaded Russia on June 22, the United States feared that Tokyo would attack Siberia, knocking Russia out of the war. No agreement that might aid this attack could be allowed.
In the third section, I examine Roosevelt and Hull’s willingness to restart talks with Tokyo in late August and their subsequent suspension of the talks in early September. This odd behavior is best explained by the successes of Germany’s renewed offensive against Stalin coupled with the worry that Japan would use a peace with the United States and restoration of trade to facilitate its long-desired attack on Russia. The last two sections discuss the last three months of peace and aborted final round of negotiations in late November 1941. I show that US officials were highly aware that Japan’s desperate economic situation, caused by the American oil embargo, was driving Tokyo into war. The key puzzle here is why Roosevelt and Hull would pursue a modus vivendi in late November, and then decide on November 26 to, in Hull’s words, kick the whole thing over. The answer is not, as traditionalists would have it, that the United States received word that China objected to any deal compromising its fight with Japan. The answer is much simpler and more consistent with everything that had transpired over the previous six months. New information had arrived late on November 25 showing that the Soviets were on the ropes and Moscow might fall. Roosevelt and Hull were compelled to end negotiations in order to force Japan south, allowing the Soviets to avoid a two-front war. Only this action would give the Allies a fighting chance to save the world from Nazism.
THE FIRST ROUND OF NEGOTIATIONS, MARCH−JUNE 1941
The Japanese nation entered 1941 with a number of unsettled issues, including the nature and timing of any further advance south. The army was pushing for additional moves against the French and Dutch colonies, despite recognizing that a conflict with United States would hurt efforts to end the China Incident and develop an offensive capability against Russia. The navy and Planning Board were still hoping for a peaceful penetration of Southeast Asia. The Planning Board’s pessimistic assessments through summer and fall 1940 had shown that Japan’s already-severe economic difficulties would get immeasurably worse should Washington block all oil sales, as it might well do if it saw its policy failing to deter Tokyo (Barnhart 1987, 174). Japan, however, now found itself in a corner from which there seemed little way out. The China quagmire and Russian war preparations combined with increased trade restrictions had left the country without sufficient resources to either end the war with Chiang on reasonable terms or match the Soviets’ massive industrial and military buildup. And yet to coerce the French and Dutch, and perhaps even the British, to join Japan’s economic sphere would only cause Washington to increase sanctions or counterattack, leading to a war of attrition that Japan was unlikely to win.3 Meanwhile, after imposing an embargo on scrap iron and aviation fuel in July 1940, the Roosevelt administration kept adding new materials to the list of restricted goods every month. By early 1941, only oil was still off the list. And this was only because Roosevelt rightly feared that oil sanctions would provoke an attack on the Dutch East Indies, given that Japan still needed the United States for 80 percent of its general oil needs and 90 percent of its gasoline (Heinrichs 1988, 7, 10–11).
By early 1941, there seemed to be only one way out of the predicament, and only one way to preserve access to raw materials and still protect Japan’s security in the core region of Northeast Asia. And that was to sign a deal with Washington that would achieve two simultaneous objectives: the ending of the China Incident through US pressure on Chiang, and the reestablishment of a US-Japanese economic relationship that would end US sanctions. This solution had obvious appeal, since by re-creating the pre-1937 status quo, it would allow Japan to refocus attention on the 1936 total-war plans for the north. In late 1940 and early 1941, the outlines for a deal began to gain acceptance among key ministers and officials. The question was how to implement it in a way that did not reveal Japan’s strong need to get out of southern China. If Washington felt it had a powerful bargaining position, Tokyo would have trouble securing a critical requirement for the peace, one that Konoe had stated in his 1938 New Order speech: the continued presence of Japanese troops in northern China, especially in the Inner Mongolian provinces of Suiyuan and Chahar. Troops in these areas were essential to the protection of Manchuria against Soviet forces in Mongolia, as the 1939 battle of Nomonhan made even clearer than before.4
Fortunately, Roosevelt and Hull were willing to talk. Since 1939 the two men had one obsession, and that was the destruction of Hitler’s Germany. In 1940–41, Roosevelt was desperately trying to keep Britain in the war through economic and military aid. He was also more than willing to place US ships in harm’s way in the Atlantic, hoping that a clash with Germany would bring the United States into the European war. The last thing in the world he needed was a war in the Far East. Thus when talk of a peace deal with Japan started to emerge in early 1941, Roosevelt and Hull were indeed interested. And as we will see, their bargaining behavior in the first half of 1941 was purely geopolitical in nature. They were more than willing to overlook moral issues such as the stationing of Japanese troops in northern China as long as Japan met one key demand: the United States be allowed unimpeded entry into the European war.
The Japanese government used two routes to send peace feelers. One was through Foreign Minister Matsuoka Yosuke, who sent signals in March and April 1941 that he was interested in having Washington mediate a peace between Japan and China (Tsunoda 1994, 13–14). The second, more indirect path was through a Japanese banker, Ikawa Tadao, who had ties to both Prime Minister Konoe and the White House. The Japanese army leadership took a keen interest in the venture. It authorized a staff officer, Colonel Iwakuro Hideo, to join Ikawa in Washington to facilitate negotiations and ensure army desires were well represented. Both of the two routes would be overseen by the new ambassador to Washington, Nomura Kichisaburo. The choice of Nomura was an important one, since it signaled Tokyo’s willingness to make a new start. Nomura was well known for his anti-Axis leanings and had shown a strong interest in better US-Japanese relations during his time as foreign minister in 1939. Moreover, as a former admiral, Nomura would give the navy a de facto representative at the table to complement Iwakuro’s presence. When Konoe through Nomura and Iwakuro offered proposals, the United States could assume that both civilian and military officials were on board back in Tokyo.
The second diplomatic path proved more effective than the first. On April 9, 1941, Ikawa presented a “Draft Understanding” to Hull detailing Japan’s demands but noting its potential concessions. Hull and Roosevelt had used a March meeting with Nomura to signal their two primary concerns: Japan’s continued adherence to the Tripartite Pact and its lack of economic openness regarding China (FRUSJ, 2:396–98). So when Hull saw the initial draft on April 9, there was much in it that was pleasing, at least as a first step. Japan pledged that the Tripartite Pact was defensive only; it guaranteed that Japanese activities in the “Southwestern Pacific” would be “carried on by peaceful means, without resorting to arms”; and it noted the desire to sign a new treaty of commerce. Japan in return wanted Roosevelt to use his office to convince Chiang to “negotiate peace with Japan.” This peace, the draft stated, would include the withdrawal of Japanese troops from Chinese territory and resumption of the open door. But it would also include “recognition of Manchukuo,” presumably by both Chiang and the United States (ibid., 2:398–402).
When Hull and Nomura met on April 16 to open what would become two months of intense negotiations, Hull told Nomura that the draft contained “numerous proposals with which my Government could readily agree.” Hull laid out four broad principles to guide the talks: respect for territorial integrity; noninterference in internal affairs; support for the principle of equality, including equality of commercial opportunity; and nondisturbance of the status quo except by peaceful means. The Japanese ambassador noted that point four might interfere with Japan’s position in Manchuria. Hull quickly made a surprising concession, stating that “this status quo point would not … affect ‘Manchukuo,’ but was intended to [apply] to the future [after] adoption of a general settlement” (ibid., 2:406–9). This was a highly significant signal. It showed that the United States was not going to let the legal or moral status of Manchuria—an area almost one-fifth the size of China—get in the way of a peace deal. In all subsequent drafts of the agreement over the next two months, the independent status of “Manchukuo” was affirmed by both sides, taking this issue off the table. Already at this early stage in the talks we see that Roosevelt and Hull were ready to operate as clearheaded geopoliticians rather than starry-eyed idealists. But this should come as no surprise: Manchuria had never impeded US-Japanese negotiations since Teddy Roosevelt had helped Japan secure its foothold there some four decades before.
As the talks proceeded through May and into early June, the US side kept reiterating three objectives. The first was advanced the hardest: Japan must distance itself from Germany. It must make a statement indicating that if the United States got into a war with Germany, even as the result of an offensive move by Washington, Japan would not invoke the defensive clause at the heart of the Tripartite Pact and respond with an attack on US possessions. The second objective was Japan’s agreement to refrain from any further aggressive acts in the region after a US-mediated China-Japan peace was arranged. The third aim was the assurance by Japan that it would reinstate the open door in China.5
Oddly missing from the list of demands, at least from the perspective of the traditional argument for why negotiations ended in late November 1941, is any US insistence that Japan put an end to the Sino-Japanese War and exit from China. This contention sees China as the key sticking point that led to the breakdown of talks and start of war: the United States wanted Japan to leave China before oil and other raw material exports would be reinstated, and Japan kept refusing. Yet in the April to mid-June negotiations, at a time when Washington should have been pushing its maximalist position, Roosevelt and Hull made no such demand. This was for one simple reason—a point that, once seen, completely alters one’s view of the causes of the Pacific War. The reason was this: getting Japan out of China was Tokyo’s objective, not Washington’s, and Roosevelt and Hull were using this as leverage to achieve their own ends.
Indeed, from a larger geopolitical standpoint, the United States had a distinct interest in keeping Japan bogged down in China. This would keep Japan from causing havoc elsewhere, either north against the Soviet Union or south against Britain and the United States. As Hull’s senior adviser Hornbeck told him on April 7, one must remember that the Sino-Japanese conflict “had become part of a world conflict.” Hence,
so long as and while Japan remains a member of the Tripartite Alliance, it would not be in the interest of the United States or in the interest of Great Britain that the Japanese-Chinese hostilities be brought to an end by any process which leaves Japan’s military machine undefeated [and] intact.… The world situation being what it is … Japan’s present involvement in China is to the advantage of the United States and Great Britain. (quoted in Tsunoda 1994, 51–52)
It was Tokyo that had approached Washington for help in ending the Sino-Japanese War. Through the April–June negotiations we therefore see Japan pleading for Hull’s guarantee that he and Roosevelt will actively push Chiang to make peace. The only two things that Tokyo sought, in fact, were this guarantee along with reinstated US trade. The US negotiators, understanding this, held back their assurances on these two items as bait to lure Japan into an agreement on the three US demands. This back-and-forth dynamic permeated almost every discussion held over these crucial three months (see FRUSJ, 2:passim). As we will see, US leaders were keenly aware through all three sets of talks before Pearl Harbor that ending the Sino-Japanese War would aid Japan at the expense of US interests.
The larger point being made is straightforward, but crucial to any rethinking of the Pacific War. Roosevelt and Hull did not care about China per se. What they cared about were Foreign Minister Matsuoka’s repeated statements that Japan would not relinquish its commitments to the Tripartite Pact. As the price for ending the Sino-Japanese War and allowing Japan to deploy its divisions elsewhere, Roosevelt and Hull wanted assurances that Japan would not use its renewed strength for attacks elsewhere in the Pacific, exploit China’s postwar economy for its own purposes, and above all, interfere in US effort to destroy Nazism. During the first three months of conversations, the US side expressed no outrage at the immorality of Japan’s war, and never demanded that Japan leave China on moral or ideological grounds. Such ideological language would be employed only after June 22 (and especially November 26), when the talks fell apart. Through the whole first round of negotiations, the United States operated in the realm of pure realpolitik. It wanted a deal binding Japan to good behavior after a Sino-Japanese peace, and made it clear that it would offer US help as an honest broker only if Japan made a credible promise to this effect.
There is not the space to discuss the details of this first round of talks, but a brief review of the efforts in May and June to revise the draft understanding will capture the essence. On May 12, Nomura presented a revised draft stating that Japan was still committed to its Tripartite Pact obligations. When Hull and Nomura met two days later, the main issue was the question of the United States’ “self-defense” against Hitler. On May 16, they met again when the US side presented its suggestions for a revision of the draft understanding. In reviewing the thinking behind these suggestions, Hull once again reinforced that Washington and Tokyo had to work out what the United States would be allowed to do against Nazism for its own security. Hitler’s plan of “unlimited conquest” required states to take “appropriate measures of self-protection and self-defense.” This meant that the United States might indeed have to become involved in the European war “even if its territory was not attacked.” Hull handed Nomura a long annex of the secretary’s recent public statements about the US need to fight Germany to defend the hemisphere.
Nomura got the message. He responded that while Japan desired only peaceful relations, it might still have to carry out its obligations under the Axis alliance (i.e., if the United States initiated war against Germany). Nomura then turned to Japan’s need for US mediation to end the China war. Hull’s draft suggestions had indicated that Washington might play this role should a broader deal be signed, so he was able to reassure Nomura on this point (ibid., 2:420–34).
When Hull met Nomura on May 20, there was an air of optimism. The ambassador assured him that there was strong support for the revised draft from the army, navy, foreign ministry, and Hirohito. Aware that Nomura was talking about the Japanese version of the draft, Hull asked Nomura about two details that concerned him: the idea that China and Japan after a peace would build a “joint defense against Communism,” and the statement that some Japanese troops would remain “in certain parts of Chinese territory” (ibid., 2:434–36).
As he went on, it became apparent that Hull was concerned not because he feared Japan might keep troops in northern China but rather because explicit statements to that effect might hurt the United States’ diplomatic position and global image. Hull stressed that he was not discussing the merit of the two points; he was simply wondering whether Chiang would view them as problematic, making it more difficult for Washington to broker a peace. It would be embarrassing for the administration, he noted, to have Chiang reject US suggestions because these two points were explicitly written into a US-Japanese agreement. Maxwell Hamilton, head of the Far East Division and Hull’s senior negotiating partner, subsequently added that the specific wording was important, since any agreement’s chances for success depended on US public opinion and the administration’s ability to counter critics.
In the subtle language of diplomacy, these comments represented a major concession. The United States had already acquiesced to Japan holding on to Manchuria. Hull and Hamilton were now indicating that the stationing of Japanese troops in northern China was not a deal breaker. If Tokyo wanted to bring it up with Chiang, fine. But it would be up to Japan and Chiang to decide the nature of China’s postwar reality. All the US side wanted was language in the US-Japanese agreement hiding the fact that Japanese troops might indeed remain in the north. Hull told Nomura that he wondered whether it would be possible “to cover these two points under some broader provision, such as a provision which would call for special measures of protection for Japanese nationals and property interests against lawlessness in areas where special measures for safeguarding the rights and interests of nationals of third powers were necessary” (ibid., 2:434–36). In short, Washington would not challenge Japan’s stationing of troops in the north. Knowing that Tokyo wanted to highlight its right to do so, however, Hull was suggesting vague language that would signal this right (under the cover of protecting citizens of third parties) without making it look like Washington had betrayed the Chinese.6
When they met the next day, Hull plunged back into the topic of Japanese troops in China. Once again his goal, far from trying to force Japan out of northern China (let alone Manchuria), was to accommodate Japanese wishes while protecting the administration’s image. Using sympathetic language designed to persuade, he suggested that Chiang’s forces might work with Japanese troops to protect areas where “special arrangements” to safeguard third-party nationals were needed. In any schedule for the troops’ departure, forces occupying the provinces of northern China could “come out last.” A commission also could be established to study the problem of keeping order in these areas. In making such suggestions in the most diplomatic way possible, Hull was again signaling that he did not want the issue of Japanese forces in northern China to derail US-Japan talks (ibid., 2:437–39).
Negotiations that day went well. Hull reiterated that the central purpose of the talks was to guarantee a regional peace. A “mutually acceptable tentative formula” was agreed on that found its way into the subsequent US draft—namely, the principle that “the controlling policy underlying this understanding is peace in the Pacific area.” Both sides pledged not just to maintain this peace but also that they had no “territorial designs in the area mentioned” (ibid., 438). This was a positive development. Back in Tokyo, though, the mercurial Japanese foreign minister threw a wrench into the works. On May 26–27, Hull received a flurry of reports from Grew and others that Matsuoka was again stating strongly that if Washington started a war with Germany, Japan would have to respond according to Article 3 of the Axis pact (see FRUS [1941], 4:224–38). When Hull met with Nomura on May 28, Matsuoka was the first topic of conversation. Given Matsuoka’s many statements on Japan’s obligations under the Tripartite Pact, Hull remarked, it was essential for Tokyo to clarify its real policy. After all, the United States might soon be drawn into the European war “through action in the line of self-defense.”
Nomura tried to reassure Hull that Matsuoka’s statements were designed for domestic consumption and the foreign minister sincerely desired peace with the United States. It would be difficult to eliminate draft language that talked about Japan’s adherence to the pact, Nomura observed, since the Japanese government would then be in a difficult position vis-à-vis Germany and pro-Axis forces inside Japan. But his government “would make its own independent decision and would not be dictated to by Germany” regarding Japan’s obligations under the alliance. Nomura was essentially asking Hull to go beyond the exact wording of the Axis document, and trust that Japan would look the other way should the United States and Germany fight (FRUSJ, 2:440–41).
At the meeting of May 28, Hull again broached the question of the withdrawal of Japanese troops from China, politely asking Nomura to “indicate what the Japanese Government had in mind.” Nomura noted that the general withdrawal might take as much as two years. He then introduced a new element that must have taken Hull by surprise. Nomura explained that the evacuation “would not include troops retained in China under the provision of cooperative defense against communistic activities. He contemplated an arrangement being negotiated with China similar to the Boxer Protocol under which Japanese troops would be stationed for an indefinite period in north China and Inner Mongolia.” For the first time, Nomura was stating explicitly that in any peace with China, Japan planned to keep forces in northern China for an undefined period. Had the nature of the Sino-Japanese peace been terribly important to Hull—as his outward behavior in late November 1941 might suggest that it was—the secretary should have ended all negotiations at that point. He did not. Instead, he simply commented that this might make a permanent friendship between China and Japan difficult. Continuing in the sympathetic mode used the week before, he even spoke of Washington’s own recent past interventions in Latin America (!) and how the US “experience” there had shown that a policy of employing troops did not pay. The subsequent discussion made it clear that Hull’s concerns revolved solely around the question of his role as broker should the Japanese seek to remain in northern China permanently (ibid., 2: 440–43).
That US leaders had no strong objections to Japanese forces staying in northern China and Manchuria—contrary to the traditional view that it was the fundamental divide that led to war in 1941—is confirmed by what followed over the next three weeks. On May 30, Hull sent Hamilton to meet with Iwakuro to clarify Japan’s thinking regarding the maintenance of troops in northern China. Iwakuro told Hamilton that the troops were absolutely critical to Japan’s effort to prevent the spread of Soviet Communism. Sketching a map of northern China, Manchuria, and Siberia, he indicated the specific Chinese provinces where Japan needed to keep forces to protect against the “the entry of communistic elements” from the Soviet puppet state of Mongolia. When asked how many troops would be stationed there, he replied one-fifth to one-tenth of the current number. When questioned about the length of their stay, Iwakuro responded that it would not be permanent but instead only long enough “[to protect] against communistic activities from external sources.”
Despite these revelations, Hamilton pushed on with his agenda. He reiterated that the question of Japanese troops mattered only insofar as Washington had been asked to mediate a China-Japan peace. Hamilton asked Iwakuro whether it might be possible to “work out a formula” whereby the stationing of Japanese troops in China after a peace would not be mentioned in the US-Japanese agreement. He even offered one possible strategy for Tokyo to consider. The document could state that in any China-Japan peace, the evacuation of Japanese troops in the north could be “left to the last.” Then Japan, “shortly before the due date of [the] completion of [the] evacuation,” could propose to China that it consult “in regard to measures called for by any situation that might exist, including any situation [involving] communistic activities.” The two sides bandied ideas back and forth on how to word such a delicate matter within the draft. With it getting late, Hamilton suggested that they leave the exact wording for now and work on it independently (ibid., 2:444–45).
The incredible nature of Hamilton’s suggestion may not be fully evident at first glance. Here was Hull’s second in command not only stating that Washington would accept language obscuring the issue as to when Japanese troops in northern China would leave, as significant as this concession was. He was actually giving the Japanese advice as to how they could use future concerns about “communistic activities” to skirt any commitments made to China regarding a fixed withdrawal date!
That Hamilton had Hull’s full confidence when he made these suggestions is shown by the US revised draft of the agreement presented on May 31. In section 3, on “action toward a peaceful settlement between China and Japan,” the US text for the first time formally accepted that Washington would approach Chiang to encourage him to make peace with Japan. In addition, the United States accepted most of the eight items from the Japanese draft of section 3. The two main changes made by Hull were important signals of the evolving US perspective. Item 6 on the withdrawal of Japanese forces from Chinese territory after a peace did add the phrase “as promptly as possible.” Yet by saying nothing about a specific time frame, the United States was giving Japan flexibility on this question. This allowance was linked to a crucial change on item 2. Here, the United States reworked the Japanese draft, which had simply called for “joint defense against communism,” to read:
2. (Cooperative defense against injurious communistic activities—including the stationing of Japanese troops in Chinese territory.) Subject to further discussion.
This was the first written confirmation that the US side would accept the possibility of Japanese troops staying in northern China under the guise of fighting Communism and protecting foreigners (including US citizens). In line with the Hamilton-Iwakuro talks, Hull was now agreeing to discuss the issue until a mutually agreeable wording or “formula” was found (see ibid., 2:446–51; cf. ibid., 2:422–23).7
Hull had a price for his concessions, however. The May 31 draft retained the annex that he had put forward on May 16 and knew was strongly opposed by the Japanese. This was the annex that provided long excerpts from Hull’s speeches on the need for US involvement in Europe for its own “self-preservation.” The secretary again wanted it to be clear that any US entry into the war against Hitler should not be considered an offensive action, regardless of how Washington got involved. In section 2 on the “attitude of both governments toward the European war,” he thus excised a paragraph from the Japanese draft stating that Japan’s obligations under the Tripartite Pact remained in place. In a concession to Japanese sentiments, Hull kept Tokyo’s line indicating that the purpose of the Tripartite Pact was defensive and designed to prevent the participation in the European war of nations not presently involved in it. But he immediately added, “Obviously, the provisions of the Pact do not apply to involvement through acts of self-defense.” For the United States, the main objective was to keep the peace in the Pacific to allow Washington to focus on defeating Nazism. If China had be convinced to accept some Japanese troops on its soil to achieve this end, so be it (ibid., 2:446–51).
The negotiations in the first half of June went well. Japanese diplomats were able to assure Hull that Japan, notwithstanding recent statements by Matsuoka, did indeed accept the United States’ right to self-defense, even if this meant a US entry into the European war (see ibid., 2:454–70). Japan’s willingness to compromise showed in its revised draft agreement, dated June 8, but presented to the United States on June 9. The basic thrust of the US draft of May 31 was kept, with revisions confined largely to adjustments of language. US changes to the critical section 2 on attitudes toward the European war were allowed to stand, including the line that Washington’s perspective on the war would be determined by considerations of “self-defense” (FRUS [1941], 4:256–59).8
The Japanese compromises seemed to do the trick. For the next two weeks, the talks went smoothly, revolving around the specific wording of what was now taken as the common baseline document: the US draft of May 31, as adjusted by Japan on June 8. There was still some discussion regarding the all-important question of Japan’s commitment to the Tripartite Pact, but it was limited and moderate in tone. The focus was on more secondary issues, in particular the question of US access to China’s economy after a Sino-Japanese peace. In their revised draft of June 15, the Japanese negotiators finally conceded that commerce in the Pacific region would conform to the principle of nondiscrimination—something Hull had been pushing for since March (FRUSJ, 2:472–76). Subsequent talks centered largely on the specifics of nondiscrimination in China, with the Japanese accepting that US trade could flow freely, but arguing that Japanese mining companies be given some preference due to the significance of raw materials to Japan’s economy (ibid., 2:476–83). That the conversation could concentrate on such details reveals just how far the negotiations had come on the big issues that had dominated the early talks. The two countries seemed headed for a peace deal, and both negotiating teams knew it.9
THE BREAKDOWN OF NEGOTIATIONS AND THE SUMMER OF DISCONTENT
Everything fell apart with Germany’s surprise invasion of the Soviet Union in late June. Both Washington and Tokyo had been aware for months of Germany’s massive buildup on the Soviet border. Until the invasion began, however, US opinion was divided as to whether Hitler was preparing for war or only trying to coerce Stalin into additional concessions. Germany’s war with Britain, after all, was still ongoing. But once the German attack on Russia began in earnest on the morning of June 22, US officials agreed that this single act changed everything. If Germany were to beat the Soviet Union and push it back behind the Urals, as many observers anticipated, then Germany would be the first empire to truly dominate the Eurasian landmass.10 It would then be extremely difficult to dislodge, even if Britain stayed in the war. From this position, Hitler could then mobilize for the destruction of his final main adversary, the United States (Tooze 2006).
Since the early spring, when reports of a possible German attack began to gain credence, Roosevelt and Stalin had begun to work on repairing US-Soviet relations (see Heinrichs 1988, 54–56). Yet the German invasion in June suddenly made the relationship between the two future superpowers the most important “axis” in the world. For the next six months, the allied effort to prevent a Soviet collapse would be Roosevelt and Hull’s number one preoccupation, with everything else seen through its lens. This would include the US relationship with Japan.
Roosevelt and his closest colleagues knew they had to act quickly if Russia were to survive. On July 24, the president publicly announced that Washington would do everything possible to provide aid to the Soviet Union. Two days later Sumner Welles, the acting secretary of state, told the Soviet ambassador that “any request” that Moscow might make for material assistance “would be given immediate attention.” In the three days after the Soviet attack, Roosevelt unfroze $39 million in Soviet assets to facilitate purchases of US goods. By June 30, the Soviets had requested $1.8 billion in military equipment and industrial hardware (FRUS [1941], 1:769–70; Herring 1973, 9–10). The US problem in meeting these requests was a simple one, though: not only was the $7 billion allotted to the first lend-lease appropriation for Britain almost exhausted, but isolationists and anti-Communists in Congress opposed giving any material assistance to Soviet Russia. Roosevelt therefore worked behind the scenes to purchase Russian gold, giving Moscow the cash needed to buy US goods until he had secure a political deal on lend lease. He also agreed to send his closest adviser, Harry Hopkins, to Moscow by the end of July to assure Stalin that much more would be on the way.11
Yet Roosevelt also still had to deal with the possibility of war in the Pacific. This section will examine Japanese decision making through late spring and summer 1941 as the German-Soviet struggle began to dominate all events. Along the way, I will provide a window into the internal debates within the Japanese polity—debates that set Foreign Minister Matsuoka against the majority of the Japanese cabinet and military. This section will also show how Roosevelt and his officials sought to deal with growing information from various sources suggesting that Japan was gearing up for a war against the Soviet Union.
It must be reiterated that the vast majority of Japanese ministers and military officials had been keen to reach an agreement with Washington during the spring negotiations. Leaving southern China and reestablishing US trade were both essential for any future military move north. The main stumbling block, as the United States had gathered through decoding Japanese diplomatic transcripts (via the “Magic” program), was in fact Foreign Minister Matsuoka. Matsuoka was already well known as a strong supporter of the northern option, and had been instrumental in securing the Tripartite Pact back in September 1940. During thirteen intense Liaison Conferences between mid-April and mid-June 1941, called to coordinate civilian and military decision making, Matsuoka displayed a stubborn unwillingness to make any deal with Washington if it meant relaxing Japan’s military obligations under the treaty. As the internal debates unfolded, it became clear that he did not oppose a US-Japanese agreement per se—after all, some form of agreement was essential to his cherished goal of attacking the Soviet Union. He nevertheless believed that Japan could not fight Russia without Germany taking on the Soviets, and that Tokyo needed to maintain a good relationship with Hitler. Herein lay the tension. As Navy Minister Oikawa Koshiro pointedly told Matsuoka on April 22, if Japan made German acquiescence a precondition to any US-Japanese deal, Japan could not achieve the key purpose of such a deal: the ending of the Sino-Japanese War (JDW, 19–24).
In late May, word arrived that the Dutch East Indies would not likely extend a six-month trade agreement covering tin, rubber, and oil exports (ibid., 36–43). By the May 29 and June 11 meetings, when it seemed the Dutch would probably start cutting Japan off, Matsuoka advocated a strong stand. Army Chief of Staff Sugiyama countered that in the view of both the army and navy, confronting the Dutch East Indies would probably draw in Britain and the United States. As such, Japan should end the Dutch East Indies talks and accept whatever raw materials it could get (ibid., 43–51).
What these internal debates up to early June reveal is that the army and navy, far from seeking war in the south to satisfy organizational needs or out of some crazed paranoia, were responding rationally to developments as they arose. They were seeking to restrain their emotional foreign minister from taking steps that would automatically throw Japan into a war with the United States. For Japanese military and civilian officials, the overarching goal remained the securing of a deal with Washington to help end the China war and reinstate the raw material imports needed for the northern buildup. For all concerned, including Matsuoka, war with the United States was still an undesired last resort.
On June 25, three days after Germany’s invasion, the first Liaison Conference on the implications of the German-Russian war was held. Another split opened up between Matsuoka and the military, this time over whether to strike the Soviets right away or wait. Matsuoka argued that with Soviet Siberian troops being sent west, Japan should strike north immediately. Army and navy leaders both opposed this, asserting that any decision depended on conditions in China, the north, and the south, and if Japan attacked the Soviet Union prematurely, “the United States will enter the war” (ibid., 56–60). In five follow-up meetings over the next six days, it was agreed that Japan did not have the capability to go both north and south simultaneously, and so must prepare for both options and see how the situation develops. A firm decision to attack Russia, as Prime Minister Konoe summarized on June 30, could be made only after consideration of the “the world situation,” meaning, most important, the state of talks with Washington (ibid., 60–75). Since these talks might still bear fruit, making a move north that much easier, Japanese leaders recognized the value of waiting.
An Imperial Conference was held on July 2 to secure Emperor Hirohito’s approval of a document titled “Outline of National Politics in View of the Changing Situation”—in view, that is, of the German attack on Russia. Reflecting the consensus of the extended cabinet, the document noted that Japan would not immediately enter the German-Soviet war but instead continue its buildup against Russia while pursuing negotiations with the United States. Then, “if the German-Soviet war should develop to the advantage of our Empire,” Japan would use armed force to “settle the Northern Question.” As this buildup was going on, Japan would push into southern Indochina, to both secure rubber and tin for the northern operation as well as prepare for a strike that might be needed against the Dutch East Indies. In carrying out this southern advance, the document underscored, Japan would not be deterred by the possibility of war with Britain and the United States, even if, as Navy Chief of Staff Nagano Osami told the gathering, this was not the desired outcome.
It was up to Privy Council President Hara Yoshimichi to speak on Hirohito’s behalf (by Japanese tradition, the emperor, being “above” politics, did not speak at such conferences). From Hara’s words, it was apparent that Hirohito was concerned about a war with the United States and wanted to make sure the focus was kept on the Russian problem. The German war with the Soviets had given Japan the chance of a lifetime: “Since the Soviet Union is promoting Communism all over the world, we will have to attack her sooner or later. Since we are now engaged in the China Incident, I feel that we cannot attack the Soviet Union as easily as we would wish. Nevertheless, I believe that we should attack the Soviet Union when it seems opportune to do so” (ibid., 77–90). Such a stance was perfectly consistent with Hirohito’s anti-Soviet views since the mid-1920s. And as Bix (2000) has shown, he took his role as the supreme commander of Japan’s military seriously. So when Hara spoke again near the end of the meeting and invoked “the Supreme Command,” it was evident to all that Hirohito knew exactly the direction he wanted to see Japan take:
I think that the Government and the Supreme Command are in agreement on this point: that is, we will try our best to avoid a clash with Great Britain and the United States. I believe that Japan should avoid taking belligerent action against the United States, at least on this occasion. Also, I would ask the Government and the Supreme Command to attack the Soviet Union as soon as possible. The Soviet Union must be destroyed, so I hope that you will make preparations to hasten the commencement of hostilities. I cannot help but hope that this policy will be put in to effect as soon as it is decided.
War Minister Tojo Hideki immediately jumped in to say that he shared Hara’s (i.e., the emperor’s) opinion, although the continuing war with China was still a constraint. Army Chief of Staff Sugiyama then spoke of the need, given Soviet current superiority, to build up the army in Manchuria. Yet with Soviet forces being shipped westward and Germany doing so well against the Red Army, he argued, Japan could be ready for war against Russia in as little as two months. Until then, Japan should continue to negotiate with the United States and Britain. The navy signaled its consent (as Sugiyama remarked) by remaining silent. When the conference ended, the note taker recorded that “the Emperor seemed to be extremely satisfied” (JDW, 77–90). This was not surprising. Hirohito now had both military and civilian officials on board for his primary objective, the single overarching objective of Japanese leaders for four decades: the destruction of Russian power in the Far East. Within days, the army moved to mobilize 850,000 men and an unprecedented 800,000 tons of materiel to Manchuria by the end of July (Barnhart 1987, 213).12
Back in Washington, Roosevelt was made aware through Grew’s reports and the Magic transcripts that the Japanese government was intensely debating whether to go north or south. On July 1, Roosevelt wrote Petroleum Coordinator Harold Ickes to say that the question of ending oil exports was a delicate one, given information he had received on Japan’s political situation. Ickes had been arguing that all oil exports to Japan should be ended given severe US gas shortages. Roosevelt, however, had known for over a year that an oil cutoff would likely lead Japan to strike south, and he needed peace in Asia to focus on the war in Europe. He thus told Ickes,
I think it will interest you to know that [the Japanese] are having a real drag-down and knock-out fight amongst themselves and have been for the past week—trying to decide which way they are going to jump—attack Russia, attack the South Seas (thus throwing in their lot definitely with Germany) or whether they will sit on the fence and be more friendly with us. No one knows what the decision will be but, as you know, it is terribly important for the control of the Atlantic for us to help to keep peace in the Pacific. I simply have not got enough Navy to go round—and every little episode in the Pacific means fewer ships in the Atlantic. (FDRPL, 1173–74)
As will become increasingly clear, by “peace in the Pacific” Roosevelt meant peace in the whole Pacific region, north and south. Already by late June, Roosevelt and Hull had shifted gears in order to deter a Japanese attack against Russia as well as the south. On June 22, the day of the German attack, Hull met with Nomura and came right to the point. He wanted to know whether “Matsuoka’s intention was … to have [Washington] subject China to pressure to come to an agreement with Japan,” so that Japan was then “free to take action with reference to the European war.” This was a thinly veiled reference to a possible Japanese attack northward to support Germany’s invasion of Russia, justified by the “defensive” clauses of the Tripartite Pact. Nomura cagily replied that Japan did not want war with the United States, but could not “bind itself in advance with regard to some future eventuality” (FRUSJ, 2:492–94).
That Roosevelt and Hull along with their subordinates were worried about a Japanese move north is demonstrated by their behavior after June 22. On June 23, Hamilton wrote a memorandum for Hull noting that some in Tokyo might argue that Germany’s attack gave Japan the freedom to move south, but others will urge “that this is Japan’s opportunity to remove the Russian menace to Japan and that Japan should attack Russia.” Japanese thinking would follow the second line, not the first, he maintained. Thus while the main purpose of the proposed US-Japan agreement was to “preserve the peace in the entire Pacific area,” Washington might not be able to deter Japan from going north. The US problem was simple but profound. Japanese leaders understood that going south against British and Dutch holdings would lead to war with the United States, but they didn’t believe the United States would respond to an attack on Russia. Tokyo would therefore be eager for a peace deal with Washington. Yet because such a deal would increase Japan’s ability to go north, Hamilton concluded, the United States had less desire for such an agreement (FRUS [1941], 4:276–77).
A lengthy paper written for Hamilton and Hull that day by subordinate Max Schmidt supported Hamilton’s contentions. Titled “The Effects on Japan of the Present War between Germany and Russia,” it claimed that Japanese leaders “cannot be expected to forget what they have for many years considered as the Russian dagger pointed at the heart of Japan.” While Japan might still decide to go south, any further military adventures were “far more likely … to be in the direction of action against Russia” (ibid., 4:983–84). Two days later, another subordinate produced a report reinforcing these concerns. German successes on the battlefield would be “an impellant to Japan to move against Siberia.” This posed a huge problem, since it would leave the Soviets fighting a two-front war that would shorten the period they could resist the Nazis. To help defeat Hitler, Washington should try to “immobilize Japan as regards an attack upon Siberia and as regards an attack against Singapore or the Dutch East Indies.” To achieve this, the report offered a fateful recommendation: “[increase] restrictions upon the export of petroleum products to Japan,” while freezing Japanese assets (ibid., 4:278–80).
On July 2, Willys Peck, from Hamilton’s Division of Far Eastern Affairs, provided a more Machiavellian set of ideas. Peck also believed that given Hitler’s recent successes, it was “on the whole more probable that Japan will decide to invade Siberia than continue her southern expansion.” Under these circumstances, “one fact stands out as indisputable,” namely, that it was
more than ever urgent that China’s resistance to Japan shall be intensified [and] to the end that more and more of Japan’s armed striking force shall be immobilized and dissipated in the “China Incident.” … The advantage of this situation to the United States seems clear. By encouraging China to ever greater efforts against Japan, by dragging Japan into deeper and deeper involvement in the China hostilities, the United States can work powerfully toward the achievement of some of her principal objectives, among them the maintenance of the status quo in the Far East, the preservation of our rubber and tin supplies, [the] aiding of Russia against Germany, and the aiding of Britain.
Peck’s recommended policy of sending China more lend-lease aid was straightforward realpolitik: in helping the Chinese achieve their ends, he contended, “we shall be taking the most effective course open to us of achieving our own ends” (ibid., 4:288–89). Peck’s arguments aligned with ideas that Hornbeck had been championing since March. So as Hull’s senior adviser on political relations, Hornbeck felt free to pen a note next to Peck’s line about Japan being more likely to go north than south. He wrote: “Contingent on German success or lack of success in Russia” (ibid., 4:288n62). Hornbeck was making sure Hull and Roosevelt recognized that Japan’s desire to attack Russia would rise should Germany appear to be winning, and US policy would have to adjust accordingly to deter a move north. As we will see, Hull and Roosevelt well understood this concern.
Given all these dire analyses, it is not surprising that US diplomacy took a decidedly hard-line turn after June 22. We have already seen that Hull that day signaled his worry to Nomura about Japan using a peace with China to take action elsewhere. On June 27, Welles instructed Grew in Tokyo to tell Japanese leaders that it was the fixed policy of the US government “to aid Great Britain and other nations” in their fight against Hitler. With Britain and Russia the only two powers left against Germany, this was an obvious reference to Roosevelt’s new commitment (announced three days earlier) to do everything possible to help the Soviet Union, including supplying aid through Vladivostok. Washington was now aligned with Moscow, and Japan must know this fact (FRUS [1941], 4:987).
Roosevelt and his associates attempted to deter Japan from going north in other important ways. Probably the most explicit was Welles’s follow-up telegram to Grew on July 4. Grew was instructed to give Konoe a message sent at “the specific request of the President.” It read:
The Government of the United States is receiving reports … that the Government of Japan had decided to embark upon hostilities against the Soviet Union. As the Government of Japan is aware, the Government of the United States had earnestly desired to see peace maintained and preserved in the Pacific area.… It goes without saying that embarkation by Japan [upon such] a course of military aggression and conquest would render illusory the hope which this Government has cherished and which it understood the Government of Japan shared [namely,] that [the] peace of the Pacific might not be further upset.
The message concluded with the hope that the reports were not based on fact, and that Washington would deeply appreciate Konoe’s assurance to that effect (ibid., 4:994–95).
This message to Konoe came as close as Roosevelt dared, given isolationist and anti-Communist sentiment in Congress, to an explicit deterrent commitment to Russia. Not surprisingly, it arrived only two days after the Imperial Conference that had decided to prepare for a northern advance within two months, should conditions permit. Roosevelt and Hull had learned of the conference through Grew and Magic, and the US ambassador in Paris affirmed this gathering’s focus on the northern option. The ambassador had information indicating that a decision was reached in Tokyo “in favor of a move against Russia.” The “so-called northern party seems to have ‘won out’ … and has been strongly supported by German influence in Tokyo.” This dispatch came just after a telegram from the ambassador in China, suggesting that the Chinese had received reliable information that “an attack by Japan on Siberia is certain” (MB, 2:56–57, appendix; FRUS [1941], 4:994–96).13
When Grew got back to Hull and Welles on July 6, it was clear that Konoe was playing for time. Konoe had avoided seeing Grew, so Grew sent a hard copy of Roosevelt’s message through an intermediary. Konoe’s two-sentence note back thanked Roosevelt and stated that Matsuoka would provide a response. Grew opined that the president’s message had probably reopened internal debates as to the advisability of the course marked out in the Imperial Conference—one that “envisages an attack on [the] Soviet Union” (FRUS [1941], 4:997–98). In short, from Grew’s perspective, US diplomacy was having its intended effect: to get Tokyo to rethink any decision to move north.
Despite Grew’s cautious optimism, the fact that Konoe was unwilling to say that Japan would not go north was undoubtedly troubling. Moreover, his passing of the buck to Matsuoka—the individual known to be the most in favor of the northern option—was not a good sign. When Matsuoka handed Grew a formal reply on July 8, these concerns were only increased. The statement read that Japan had always had a sincere desire for peace and had “not so far considered the possibility of joining the hostilities against the Soviet Union.” This hedging of bets was repeated in Matsuoka’s July 2 statement to the Soviet ambassador (which Grew attached as an annex), affirming that Tokyo did not “at present” feel compelled to modify its policy toward Russia (FRUSJ, 2:504).14
It was apparent that more had to be done to protect the United States’ most valuable ally from a two-front war. On July 14, the administration chose to take the highly unorthodox step of demanding that Japan change the makeup of its government. That day, again incapacitated by illness, Hull had Hamilton deliver a blunt message to Nomura asking whether Tokyo was willing to assert control over “elements” in the government that “supported a policy inconsistent with the policy of peace” (FRUSJ, 2:505–6). Given Hull’s known concerns about Japan’s foreign minister, the message was clear: Matsuoka had to go if Japan expected any sort of deal with Washington.
The normally mild-mannered Nomura, plainly taken aback, told Hamilton the next day that the United States had no business raising questions about particular individuals in the Japanese government. Hamilton disclaimed any idea that Washington was trying to interfere in Japan’s internal affairs. The goal was only the achievement of peace, not just between Japan and the United States, but “among all countries in the Pacific area” as well. Nomura’s reply was not reassuring. For the first time since the negotiations had begun in April, he referred to Japan’s encirclement by hostile powers and that Japan might be forced to “take appropriate precautionary measures” such as the United States had done recently in occupying Iceland (ibid., 2:506–9).
Back in Tokyo, however, Hull’s remonstrance was having its desired effect. Army, navy, and civilian leaders had been concerned for some time about Matsuoka’s undiplomatic stance toward Washington and tenacious adherence to the Tripartite Pact. Given this consensus, Konoe decided to use a procedural trick to both satisfy both US demands and get rid of a difficult colleague. On July 16, he secured the resignation of the cabinet. Two days later, the emperor authorized Konoe to form his third government, and Konoe asked every cabinet member to rejoin him—except Matsuoka.15 With Matsuoka out and Toyoda Teijiro in as foreign minister, Konoe could proceed with his larger plan: to make a peace with Washington that would give Japan the time and resources to prepare for war in the north.
For the United States, Matsuoka’s dismissal was a good sign. But Konoe had still not committed to the status quo, either for Russia or Southeast Asia. Something else was needed, especially as it became evident through US intelligence that Japan was gearing up to occupy the southern half of Indochina. This something was the final weapon in the economic arsenal: tight restrictions on all oil sales to Japan. The way the story is typically told is that the licensing restrictions announced on July 25 were only to punish Japan for its move against southern Indochina. Roosevelt did not intend for them to turn into a complete cutoff of all oil sales, but he became a victim of bureaucratic overzealousness. In this way, Roosevelt inadvertently pushed Japan into all-out war by depriving its economy of its most vital resource.16 This story would fit nicely with the trade expectations argument, since it suggests that Japan might not have attacked Pearl Harbor had Washington not ended oil exports.
These elements of the traditional account are nevertheless only partial truths at best, and as such, they obscure the larger dimensions of Roosevelt’s global strategy. There have always been two big problems with this account. First, it assumes that Roosevelt, who had known for a year that sanctions on general oil products (as opposed to specialty items like aviation fuel) would almost certainly drive Japan to attack the Dutch East Indies, could have allowed lower-level bureaucrats to formulate the details of oil policy without his direct oversight. Second, it asserts that Roosevelt returned from his Atlantic conference with Winston Churchill, realized that the bureaucrats had turned the sanctions into a complete embargo, and then did not have the courage to correct this mistake immediately, despite knowing of its implications. In essence, we are asked to believe two things at odds with everything we know about this consummate politician: he was of both weak intellect (he didn’t understand the licensing arrangement) and weak character (he couldn’t reverse the complete embargo for fear of seeming soft on Japan or upsetting his underlings).
There is clear evidence, however, that Roosevelt did in fact know and understand the exact details of the licensing arrangement before he left to see Churchill, and approved them. On July 18 Roosevelt told Ickes, who was still keen on an oil embargo, that he agreed with Hull that such a move would precipitate a Japanese attack on the Dutch East Indies. At a cabinet meeting that day, the president ruled out a complete embargo, but agreed that the United States should freeze Japanese assets and limit US sales of oil to Japan to “normal” amounts—that is, pre-1939 levels (Barnhart 1987, 225–28).17 Over the next six days, a program was developed by the State Department with the following characteristics. Japanese firms would first have to ask for licenses for specific amounts from the department’s Division of Controls. To then receive the oil, the firms would need a second license from the Foreign Funds Control Committee (chaired by Assistant Secretary of State Dean Acheson) in order to release the funds to buy it.
This complicated series of bureaucratic steps had one great appeal to Roosevelt and his secretaries when the cabinet met on July 24 to discuss the matter. Whereas a full embargo, announced as a “regulation,” would merely stop all oil purchases until it was rescinded, the licensing system would give Roosevelt and Hull maximum flexibility to increase or decrease the level of economic pressure on Japan, depending on Japan’s behavior and US objectives (ibid., 230). The system would also achieve the key tactical objective recommended by the Division of Far Eastern Affairs back in late June: to keep Japan uncertain about US oil policy (FRUS [1941], 4:278–80). With reports in late July that Russia was starting to slow the German juggernaut and would probably survive until winter, such a policy would reinforce any Japanese reluctance to go north while still giving Tokyo the hope that oil could be reinstated if its behavior improved (Heinrichs 1988, chap. 5).
Roosevelt and his cabinet approved the licensing recommendation at their July 24 meeting. The next day the White House released a statement. By its very wording, it is obvious that this statement was designed to create confusion and hence uncertainty as to what the United States was really up to. It mentioned the new licensing system, but said little about how this system would actually work, noting only that the measure “brings all … trade transactions in which Japanese interests are involved under the control of the [US] Government” (FRUSJ, 2:266–67). The statement thus signaled that the US state would now dictate every aspect of Japanese-US trade, down to the smallest detail. But it left the Japanese with little clue as to how to get the bureaucracy to actually approve of oil sales.
The British and Dutch were as puzzled as the Japanese as to what the United States was really announcing on July 25. Still, to ensure maximum economic pressure, they immediately renounced their trade treaties with Japan (see Heinrichs 1988, 135–36). Back on July 8, President Roosevelt and British ambassador Edward Halifax had talked about the importance of using oil sanctions to keep the Japanese guessing (see ibid., 134, 141). Roosevelt quickly saw that the new strategy was working. On July 26 he sent a letter to Churchill through his personal assistant Hopkins, who was on his way to Moscow to discuss US aid for Russia. Roosevelt instructed Hopkins to tell Churchill that “our concurrent action in regard to Japan is, I think, bearing fruit.” Revealing how much Roosevelt already knew about Japanese decision making (via Magic) a mere day after the freeze announcement, the president went on to say that “I hear their Government [is] much upset and no conclusive future policy has been determined on” (CR, 1:225). The coordinated sanctions policy was already pushing the Japanese to rethink the wisdom of further expansion, either north or south.
Even if it is clear that Roosevelt fully understood the strategic logic behind the new policy, one might still argue that he thought the sanctions were only partial when in fact hard-line bureaucrats were acting to turn them into a complete embargo on all oil exports. Yet this part of the traditional story is also a distortion. Roosevelt’s time frame was this: he planned to meet Churchill at Argentia from August 6 to 9, leaving Washington just before August 6. On July 30, George Luthringer of the Office of the Adviser on International Economic Affairs submitted a long memorandum on the execution of the licensing system to Acheson and his boss, Hull. In the first main section, it noted that for the immediate future—specified in brackets as “approximately two weeks”—there should be absolutely no indication of what Washington’s general policy was. During these two weeks, a complete two-way embargo would be imposed: no license applications on exports or imports would be granted. But in the “intermediate period” that followed, Japan would then face a trial period in which Washington could adopt “a cautious attitude … with respect to trade with Japan.” Some export and import trade might be reinstated, including a limited amount of petroleum exports to Japan, but this would depend on “general political developments in the Far East” (FRUS [1941], 4:844–46).
This document was commissioned by Acheson to reflect the conclusions of an Interdepartmental Policy Committee meeting held the previous night (ibid., 4:844n13). It shows that the intention right from the beginning was to completely shut down all US-Japanese trade—not just exports, but imports as well. This would send the clearest possible signal that Washington could hurt Japan if it wanted to. Yet by avoiding any sign that this was part of a new general policy, Tokyo would retain enough hope that it could get the trade reinstated if its behavior improved. Once again, we see a deliberate policy at the highest levels to create uncertainty and anxiety in the Japanese government, and nonetheless retain the flexibility needed to negotiate a peace.
But was Roosevelt aware of this Machiavellian strategizing at the highest levels before he left for Argentia? He certainly was. On July 31, with Hull again out sick, Welles sent the president a long memo on the execution of the sanctions policy, providing in essence a summary of Luthringer’s report, yet also incorporating other details discussed internally over the previous week. Welles told Roosevelt that some Japanese applications for export licenses had been submitted, but “[for] the time being, [Acheson’s] Foreign Funds Control Committee is holding these applications without action.” It was desirable, Welles continued, that this committee along with export-control authorities “be given instructions as to the policy you desire it to follow.” He described a list of embargoed goods that the Foreign Funds Control Committee would continue to deny. With regard to petroleum products, though, depending on developments, Welles noted that it might be worth reinstating trade in gasoline and lubricating oils at “normal” 1935–36 levels—that is, prior to the increase in Japan’s oil purchases caused by the China Incident. Just to make sure Roosevelt got the point, Welles ended by saying that the committee would “continue to hold without action” all oil-export applications and subsequently grant licenses “only in accordance with the policy to be initiated by export control”—that is, after instructions from the president (ibid., 4:846–48). Roosevelt approved of Welles’s recommendations by noting: “SW [Sumner Welles] OK. FDR” (ibid., 846n16).
Thus we see that Roosevelt left for his conference with Churchill perfectly aware that for two weeks, all sales of oil products to Japan would be ended. And he knew that it would be up to him to decide what policy would replace this total embargo. The strategy was indeed a brilliant one. The fact that Japan would get no oil for two weeks would shake Japanese leaders out of their complacency. Yet by retaining the flexibility to reinstate oil sales after two weeks simply by allowing license applications to go through, Roosevelt could reward Japan for any improved behavior without much damage to his reputation as a reasonable negotiator. The fact, too, that Luthringer’s memo explicitly stated two weeks as the time frame for the immediate period is instructive: two weeks from the imposition of sanctions (July 25–26) was when Roosevelt was slated to return from his Argentia meeting. Roosevelt would have time to discuss the issues with Churchill, observe what Japan was doing on the diplomatic or military fronts, and then make a final decision as to whether to persist with the complete embargo. The real story, then, is the exact opposite of the one traditionally told: far from Roosevelt leaving for the conference believing sanctions were only partial, he knew from the beginning they were total, and it was up to him to decide whether such an interim policy should be continued.18
The Japanese government, believing that the oil embargo was only related to the July move into southern Indochina, sought to appear accommodating. In a note delivered to Hull on August 6, Nomura proposed both that Japan would go no further than Indochina and all troops there would be withdrawn after the peace with China. The note expressed hope that in return, Washington would reinstate normal trade relations and even help Japan secure raw materials from Southeast Asia (FRUSJ, 2:549–50). Hull, however, came to the meeting ready to signal the United States’ tough new posture. Instead of reading the note, Hull stuffed it unceremoniously into his pocket. Two days later he handed Nomura a reply stating that Tokyo’s proposal was “lacking in responsiveness” to Roosevelt’s earlier suggestions (ibid., 2:546–53). The president received Hull’s reports of the exchange while he was at the Atlantic Conference. On August 11, the last full day of meetings, he decided to give Churchill a copy of Japan’s August 6 proposal. The two agreed that Japan would have to promise not to station additional troops in Southeast Asia and withdraw the ones now in Indochina before Washington would accept Tokyo’s terms. If the Japanese government refused, Roosevelt was to inform it that “various steps” would have to be taken, even if they resulted in war (FRUS [1941], 1:356–60).
The two then considered whether the warning should include a statement covering any aggressive steps by Japan against Russia. Welles reiterated that the main US goal was to keep the peace in the entire Pacific region, regardless of whether Japan intended to go north or south. Hence, the statement Roosevelt should make to Nomura should be “based on the question of broad policy rather than premised solely upon Japanese moves in the Southwestern Pacific region.” This of course had been Roosevelt, Hull, and Welles’s thinking all along: to deter not just a Japanese move south but also north. Roosevelt agreed with Welles’s suggestions, reasoning that this strategy would help prevent further Japanese aggression for at least thirty days (ibid., 1:356–60). The “thirty days” remark shows just how pessimistic Roosevelt had become regarding Japan’s short-term intentions: while he might not be able to deter a Japanese attack on Russia, he might still be able to buy a month. With the Soviets still on the ropes in the west, a month might be just long enough to preserve the Soviet Union through the winter. The global strategy to defeat Hitler thus demanded one more effort at negotiations with Japan.
THE ABORTED EFFORT AT A SECOND ROUND OF NEGOTIATIONS, AUGUST−SEPTEMBER 1941
That Roosevelt still wanted peace in the Pacific region is shown by his behavior over the next three weeks. True to his word, on August 17, soon after returning to Washington, he called Nomura to the White House. Having signaled Konoe’s interest in a summit meeting over a week before, Nomura now stressed Konoe’s strong desire for such a meeting at some midway location, perhaps Honolulu. Roosevelt replied that before discussions could continue, it was necessary to clarify the US position. He read and then handed the ambassador a long statement reiterating Washington’s new deterrence posture. Japan had continued to deploy its forces around the Far East and had occupied southern Indochina, the document noted. The president wanted Tokyo to know that if Japan took any further steps as part of a policy of “military domination … of neighboring countries,” Washington would be compelled to take “any and all steps” to safeguard US security in the region (FRUSJ, 2:556). As he and Churchill had agreed, Roosevelt was now explicitly extending the US deterrent to the whole region, including the Soviet Union (one of the “neighboring countries”).
Roosevelt then turned to the question of Japan’s desire to restart negotiations. He read and handed to Nomura a separate statement that began by bluntly declaring that any resumption of talks would require an agreement covering “the entire Pacific situation.” Five of the next six sentences repeated this refrain—any deal had to involve all the countries in the Pacific region. If, after an agreement, any of these countries were menaced, Washington would immediately extend assistance to them. Following this stick, Roosevelt dangled the carrot: through a peace agreement, Japan could secure everything it wanted, including the “satisfaction of its economic needs.” Tokyo, however, must furnish a clearer statement of its present attitude and plans. Nomura replied that he would communicate Roosevelt’s views to his superiors, adding that they were “very desirous” to preserve the peace between the two countries (ibid., 2:557–59).
Roosevelt was still determined to leave no stone unturned to secure peace for the entire Pacific region. While we might dismiss Roosevelt’s diplomacy at this stage as designed only to buy time, it was that and much more. A peace deal with Japan protecting both the north and south would achieve his primary objective: the freedom to pursue a war with Germany without the diversion of a second front, and with Russia and Britain free to focus solely on the European theater. As such, on August 18, Roosevelt met again with Nomura in a secret meeting (with only Hull present) to confirm his interest in negotiations. He told Nomura that it might be difficult to meet in Honolulu, given the overseas flight, but that perhaps the summit could be in San Francisco, Seattle, or even Juneau, Alaska, if that worked for Konoe (MB, 3:26–29, appendix).19
Back in Tokyo, things were coming to a head. The US oil embargo had had a dramatic cooling effect on those wanting an immediate invasion of Russia. As Barnhart shows, even those most inclined to a northern move agreed that a southern advance was now necessary. War in the north was always dependent both on Germany’s success in the west and adequate raw materials. Two reports in the first week of August confirmed that a northern attack before the snows began in October was unwise now. The first report by the General Staff Intelligence Division predicted that Russia, displaying more resilience at this point, would now not surrender in 1941, and perhaps not even in 1942. The second report by the Army Ministry’s Equipment Bureau stated that given the Western oil embargo, without the oil fields of Southeast Asia, Japan could not fight the Soviets with any degree of confidence (Barnhart 1987, 239). Roosevelt’s nicely timed oil licensing system had had one of its desired effects: it had ended all talk about an autumn attack on Russia. On August 9, all operations planning for the attack was temporarily suspended and instead redirected toward southern preparations. While the army led this charge, the navy was in agreement. With no oil imports, both army and navy analyses estimated that Japan had only two years of stocks remaining, even if it undertook no major military operations (ibid., 240–41).
Declining expectations of future trade had shifted the Japanese focus from north to south. The main issue during the August Liaison Conferences was now not strategy but rather the inevitability of war with the United States. Konoe had secured army and navy support on August 7 for the idea of a direct summit with Roosevelt. Everything depended on the terms, though: the army signaled that it wanted peace with the United States, but only if Roosevelt agreed to allow raw materials to flow into Japan unimpeded. At the August 26 conference, participants concurred that Japan should respond to Roosevelt’s latest démarche in a reasonable way to help persuade Roosevelt to proceed with the summit (JDW, 85–112).
On August 28, Nomura met Roosevelt and handed him two documents. The first expressed the desire for “the peace of the Pacific” and thus “urgent necessity” of a summit meeting. The second signaled some of the concessions Japan might be willing to make. Japan would be willing to withdraw its forces from Indochina “as soon as the China Incident is settled or a just peace is established in East Asia.” Moreover, Japan would take no military action against Russia “as long as the Soviet Union remains faithful to the Soviet-Japanese neutrality treaty and does not menace Japan or Manchukuo.… In a word, the Japanese Government has no intention of using, without provocation, military force against any neighboring nation” (FRUSJ, 2:572–75).
These were significant changes in Japanese policy. By including the line regarding a just peace in East Asia (i.e., a US-Japanese agreement), Tokyo was now indicating that it would not make the ending of the Sino-Japanese War a precondition to a pullout of Indochina. More important, for the first time the Japanese government was willing to promise that it would not attack the Soviet Union unless provoked. Japan, in essence, was recommitting itself to its April 1941 nonaggression pact with Moscow, notwithstanding the pressure being placed on it by Berlin to attack Russia’s eastern front. Hull had reiterated the significance Washington placed on this assurance just five days earlier, at a meeting on August 23, when he told Nomura point-blank that should Japan “project herself into the Russian-German situation,” there would be no chance at a US-Japanese agreement (ibid., 2:567).20 Clearly Tokyo had got the message.
Given the war that broke out in December, one might conclude that Japan had no intention of honoring these concessions, and Roosevelt and Hull knew this. This view is too simple on two counts. For one, the Japanese intention was indeed to honor the Japanese-Soviet pact for at least six months, given that the earliest feasible time for a move north was now seen as February 1942. A peace deal that reinstated trade would help Japan prepare for this northern war.
Roosevelt and Hull also plainly saw the Japanese statements on August 28 as crucial, at least in terms of buying time for the United States’ own buildup and perhaps as a way to keep the Pacific peace. That night, at Hull’s request, Nomura came to Hull’s apartment to talk about the specifics of both the reinstated “informal” negotiations and a subsequent summit meeting. What is most significant is not the details discussed but instead that Hull proceeded as if it were already a given that the negotiations and summit meeting would go ahead. The meeting’s tone was similar to the May–June meetings—so much so that when Nomura reintroduced the notion that Washington act as intermediary to end the China war, Hull again accepted that Washington would play this role, provided Chiang was willing. The US position was remarkably reasonable. Hull spoke of his willingness to “work together” to make the most of China’s potential as a trading nation, and the conversation itself revolved around the agreed assumption that Japan would pull its troops out of Indochina after the China Incident was resolved—as opposed to such a withdrawal being a precondition for US help in facilitating a Sino-Japanese peace, as Roosevelt had emphasized earlier in the month (ibid., 2:576–79).
Despite this promising restart of the aborted talks from June, six days later everything fell apart. On September 3, Roosevelt called Nomura into the White House and presented two statements to him. Before any summit could go forward, he stressed, the Japanese government had to show that it stood “earnestly for all of the [four] principles” that Hull had used as guidelines for talks back in April. Yet even if the Japanese were somehow able to prove their adherence to such vague principles—including respect for territorial integrity, noninterference, and commercial equality—Roosevelt added another condition. The US government would have to first “discuss the matter fully” with the British, Chinese, and Dutch to secure their acceptance of any deal (ibid., 2:588–92).
These two new conditions were deal killers, as Roosevelt certainly understood. The first one was bad enough. Whereas discussions over the previous two weeks had pragmatically revolved around the June draft agreement, Roosevelt was now forcing Japan to meet the unstated standards of four obscure principles. In diplomacy, this is an old technique designed to signal lack of interest in the horse-trading needed for a deal. The second condition, on the other hand, was completely new. Roosevelt had never before indicated that he required the approval of even the British before proceeding, let alone secondary allies such as the Chinese and Dutch.
Back in Tokyo, Foreign Minister Toyoda was clearly worried after receiving Nomura’s meeting report (MB, 3:64–68, appendix). Later that same day (September 4, Tokyo time; September 3 in Washington), Toyoda called in Grew and explained that Konoe wanted to “leave no stone unturned” to secure a summit meeting. Toyoda gave Grew a memorandum summarizing seven concessions, drawing Grew’s attention to parts (c), (d), and (e). Here, the three main sticking points over the last six months—Japan’s adherence to the Tripartite Pact, the stationing of troops in China, and the issue of the open door—were eliminated in one stroke. Should the United States get involved in the European war, Japan’s interpretation of the Tripartite Pact would be “independently decided.” (Toyoda emphasized several times how different this was from Matsuoka’s position.) Japan would also remove its armed forces from China as soon as possible after a Sino-Japanese peace, while restrictions on US firms in China would end. These concessions had been cabled to Nomura, Toyoda mentioned, but he underscored the importance of Grew transmitting them immediately to Washington (FRUSJ, 2:593–94, 608).
The next day (September 4, Washington time) Nomura rushed a new version of the draft agreement to Hull, incorporating the concessions. The section on the Tripartite Pact, borrowing from an undiscussed draft Japan gave Hull the day before Germany’s attack on Russia, stated that the pact was defensive only, designed to help prevent an “unprovoked extension” of the European war. The word unprovoked was critical, since it marked Japan’s willingness to ignore any obligations to Germany should the United States enter the war. (Japan could claim that US entry into the European conflict had been provoked by German expansionism.) In section 5, Japan accepted the principle of nondiscrimination in trade. In the section on China, Japan agreed to withdraw from Chinese territory “as promptly as possible” after a Sino-Japanese peace, and within a maximum of two years. Clause 2 on the stationing of Japanese troops in northern China, as the document noted, had been “entirely dropped” to accord with US wishes (ibid., 2:597–600, 608, 488).
These concessions were sweeping, to say the least. A comparison with the US draft from early June shows Japan now complying with Washington’s desired wording for almost every section of the document. In fact, the new document went further than the US draft on the troops issue. In June, the United States had agreed to discuss the stationing of troops in northern China to fight Communism and had simply asked for troops in southern China to leave “as promptly as possible.” By dropping the stationing question entirely and specifying a two-year maximum for withdrawal, Japan was communicating its willingness to withdraw completely from all of China (excluding Manchuria) after two years (ibid., 2:597–600; cf. ibid., 2:446–49).
Notwithstanding these concessions, Hull, in his meeting with Nomura, put up roadblocks every step of the way. He restated the need for Allied approval before proceeding. To Nomura’s point that Washington now should not fear Japan’s adherence to the Tripartite Pact—the primary US issue for the whole April–June period—Hull said that this would create difficulties with the Senate if there were no “explicit assurances” on these items (ibid., 2:595–96). These were clearly bogus reasons for not seizing the most important set of Japanese concessions to date. Indeed, Hull didn’t even try to exploit Tokyo’s obvious desperation to push his demands further, as most good negotiators would do. He simply refused to continue bargaining.
What was going on here? Why would Roosevelt and Hull become intransigent just when the deal they had been seeking for half a year was within their grasp? This is a key question, given that Roosevelt and Hull’s actions effectively shut down serious negotiations until mid-November. The oil embargo that could have been reversed by an agreement was continued, which of course only reaffirmed Japan’s already-pessimistic expectations for future trade. Roosevelt and Hull’s decision to abandon the summit and negotiations made war essentially inevitable in the absence of new negotiations. So why did they do it?
There are few explanations in the literature on this conundrum, largely because scholars tend to presume that neither side by August–September was really that interested in a deal.21 The above discussion shows beyond doubt that a lack of interest was not the problem. In August, Roosevelt had still wanted a deal so he could concentrate on Hitler and buy time for his naval buildup. The Japanese military and civilians wanted to avoid war with the United States so they could use raw material imports from the south and the United States to facilitate a war against Russia.
Because there are no documents revealing Roosevelt and Hull’s thinking during the first days of September, we must rely on circumstantial evidence and logic.22 One plausible explanation is that a number of factors interacted to convince them to shift to an intransigent stance. Three factors seem most salient. First, the existence of secret talks regarding a possible summit had leaked out in both the Japanese and US press around this time. For hard-liners in both Japan and the United States, such talks were anathema. This may have led Roosevelt and Hull to believe that continuing the talks were too politically dangerous, or that Konoe would no longer have enough support at home to ratify any agreement.
It is also possible that Roosevelt and Hull, given new evidence that Japan was preparing for a war in the south sooner rather than later, suddenly decided that negotiations were no longer useful and that Japan would only use such talks to buy time. Grew was sending reports from Tokyo in late August noting that Japan seemed to be mobilizing for war (see FRUS [1941], 4:408–19). More significant, the Japanese civilians and military held an important Liaison Conference on September 3 that affirmed a consensus that Japan, pending Hirohito’s approval at an Imperial Conference, had to attack south should diplomatic efforts fail (JDW, 129–33). Since this conference was held on September 2, Washington time, there was a day for Magic and other intelligence tools to learn of this meeting along with what had transpired.
Finally, in late August the Germans renewed their offensive against the Soviets in an all-out effort to destroy Stalin’s army before the onset of winter. The German army had been forced to halt in early August and for three weeks had remained in place on a line running from the Baltic through the middle of Ukraine. But on August 22, Germany began a full-scale drive toward Leningrad, the Caucasus, and ultimately Moscow. Reports began to arrive in Washington in late August and early September that the Russians were not doing well and were retreating (Heinrichs 1988, 146–79; Clark 1965, 109–49). With Roosevelt and Hull fully aware since June that Japan’s decision to go north depended on German success on the western front, this news could not have been pleasing. As Roosevelt had remarked to Ambassador Halifax in mid-August, the Japanese “were more influenced by the sway of battle in Russia than by regard for the United States.”23 The thinking behind Roosevelt and Hull’s sudden ending of negotiations thus probably mirrored their thinking in late June: with Japan now more likely to attack Russia, any deal allowing a diversion of troops north and renewal of raw material flows had to be scotched.
All three of these factors may have contributed to the about-face on September 3. There are good reasons, however, to believe that the first two were not nearly as salient as the third. The idea that Roosevelt and Hull would stop negotiations because the existence of these talks had been leaked seems suspect. The concessions offered by Tokyo on September 4, if codified in a final agreement, would have been a huge public relations coup for Roosevelt. It would have demonstrated that coercive diplomacy, including oil sanctions and a military buildup, could pay big rewards, not only in keeping the peace, but in renewing trade with China, too. As for the argument that Washington ended negotiations because of new evidence on Japan’s preparations for war, this also seems suspect. It is unlikely that it was Grew’s reports regarding increased Japanese mobilization that caused the shift in US behavior. After all, this news was consistent with what the United States had known for two months: that Japan, after the Imperial Conference on July 2, was mobilizing for total war either north or south, depending on subsequent conditions. A somewhat more plausible notion is that Roosevelt and Hull received information about the September 3 conference, and believing war essentially inevitable, decided to end negotiations.
Discussions at this conference did in fact reveal a growing desperation in Japanese thinking. Navy Chief of Staff Nagano began the conference by noting that “the Empire is losing materials: that is, we are getting weaker … [and] the enemy is getting stronger.” With the passage of time, “we won’t be able to survive.” Japan had to try diplomacy, he continued, but if this should fail, a decision for war must be made quickly. He was not optimistic about the chances of victory. Japan might do well in the opening months, yet even with an initial decisive battle, “it would be a long war,” and one that Japan, with its small resource base, was unlikely to win. Sugiyama then spoke of the need to achieve Japan’s diplomatic objectives by October 10 at the latest. His reasoning had nothing to do with weather or logistics, and everything to do with the army’s true objective: destroying Russia. Japan could not carry out large-scale operations in the north until February. “In order to be able to act in the North,” he explained, “we have to carry out operations quickly in the South. Even if we start right away, operations will take until next spring. Insofar as we are delayed we will not be able to act in the North. Therefore, it is necessary to move as quickly as possible [in the South].” Agreement was reached that day that Japan had to give diplomacy a chance. If there were no results by October 10, though, the leadership would then “immediately decide to commence hostilities.” The ministers also agreed to ask the emperor for an Imperial Conference three days later to approve this plan of action (JDW, 129–33).
Had the details of this important conference been picked up by Magic, it might have contributed to Roosevelt and Hull’s decision to abruptly end negotiations. But even if they had gotten wind of this conference, there is still a puzzle.24 What information from it would have driven them to stop the negotiations? It is highly improbable that the news of the decision to begin a move south after October 10 would have done it. After all, this decision explicitly depended on Konoe’s diplomacy failing. So if Roosevelt and Hull had wanted to prevent a move south, driven by resource scarcity, they should have increased their efforts to sign a deal giving Japan some oil and raw materials. Only this would have improved Japanese trade expectations and averted a tragic plunge into Southeast Asia. At the very least, it would have bought time for the US buildup—something, as we will see, that the US military desperately wanted.
If Roosevelt and Hull had received details on the conference, the thing that would have been most worrisome was Sugiyama’s comment that the army was still actively anticipating action against Russia as early as February. Japan, like the United States, had undoubtedly learned of the renewal of the German offensive on August 22 and its initial successes. If Roosevelt and Hull had learned of Sugiyama’s remark, they could only assume one thing: any US-Japanese agreement helping Japan to end the China war and restore trade would mean that hundreds of thousands of troops as well as millions of tons of resources would be sent north for the offensive against Russia.25 As such, a deal with Japan would have been the last thing Roosevelt and Hull wanted.
Yet even without information about the September 3 conference, news from the Russian front would have been upsetting enough to cause Roosevelt and Hull to suspend negotiations. Knowing that the decision to go north depended on Germany’s advance toward Moscow, and that Japan had gone from two hundred thousand to around six hundred thousand troops in Manchuria as a result of the July buildup, Roosevelt and Hull had good reason not to help Japan reinforce its northern army any further (see PHA, 13–14:1346–47). This does not mean, of course, that the pair was deliberately trying to draw Japan into an attack on Southeast Asia. Indeed, over the next three months, they both continued to be preoccupied by the need to get into the European war and avoid war in the Pacific. A two-front war had always been seen as highly problematic, and the imminence of it in November would bring them back to the idea of a temporary modus vivendi with Tokyo to buy at least six more months of peace. In early September, however, Roosevelt and Hull could not afford to facilitate a Japanese attack on the one state, Russia, that was critical to winning the war against Hitler.
To see this, we need to look briefly at the larger geopolitical context during late summer 1941. Other than provoking a war with Germany in the Atlantic, sending as much military and economic aid to Russia as possible remained Roosevelt’s primary geopolitical objective. For an hour on August 1, he angrily lectured cabinet members, particularly Secretary of War Henry Stimson, for their foot-dragging on this issue. With information confirming his hopes that Russia could hold Germany back until winter, he told the cabinet that whatever the United States could supply to Russia must be there by October 1, “and the only answer I want to hear is that it is under way” (quoted in Herring 1973, 13–14). He was even willing to sacrifice the United States’ own military buildup to achieve this end; he ordered Stimson to send the promised aircraft immediately, even though there were no planes to replace them (Henrichs 1988, 139–40).
Through August and into September, Roosevelt continued to berate his subordinates about the slow pace of Russian aid. Huge supply problems were arising as the US economy suddenly prepared for total war. The president was trying to give the Russians everything they asked for. But with shortages pervasive and the anti-Soviet legislators in Congress trying to bar the extension of lend lease to Russia, Roosevelt was in a bind (Herring 1973, 13–17). Hopkins’s trip to see Stalin in late July–early August had been a success: the two states were now in a de facto alliance, with Hopkins reassuring Stalin that aid was on its way. Yet there was little Roosevelt could really do for Russia in material terms in the short run. From the larger strategic perspective, however, there was certainly something that the president could do, and that was to keep Japan from opening up a second front in Siberia.
The need to end negotiations with Japan would have been reinforced by the wording of the above-mentioned statement of concessions presented to Grew late on September 3 (Washington time). In listing seven things Japan would agree to do, Toyoda had drawn Grew’s attention to items (c), (d), and (e)—the parts discussing the Axis pact, troops in China, and open door, respectively. Items (b) and (f), in contrast, while they seemed to be important concessions, contained a glaring omission. In item (b), Japan promised that it would not resort without justifiable reason to military actions against any regions “lying south of Japan.” In item (f), Japan promised that its activities “in the Southwestern Pacific Area” would be carried on only by peaceful means and according to the principle of nondiscrimination (FRUSJ, 2:608).26 For anyone who knew anything about Japan’s recent internal discussions, such language would have raised immediate red flags. Here was Japan promising to keep the peace in the south, leave China, and observe economic nondiscrimination, but making no explicit assurances regarding its all-important neighbor to the north, Russia. The very fact that Toyoda would not have tried to dissimulate by presenting his concessions as covering the entire Pacific area suggests that he either hoped any US-Japanese agreement would last past early spring 1942 (the projected date for the northern invasion) or was constrained by army leaders from making a larger concession. Regardless, Roosevelt and Hull would have easily read between the lines, thereby recognizing that the Japanese government’s desire for peace in the south was masking its larger plans for war in the north.27
Logic and circumstantial evidence thus point to one primary explanation for why Roosevelt and Hull abruptly ended negotiations with Japan on September 3–4, despite having restarted them in earnest only a week earlier. It seems that the only thing that could have propelled them to shift policy—especially given the known risks of ending Japanese hopes for renewed trade—was the new victories of the German army on the Russian front and their anticipated impact on Japanese decision making. The two other factors—the leaking of the existence of secret talks and evidence about Japan’s buildup for war—were too weak on their own. At best, they served only to reinforce US fears that Japan was preparing to go north and would need US raw materials to do so.
THE FINAL MONTHS OF PEACE, SEPTEMBER−NOVEMBER 1941
Almost everything that occurred over the next two and a half months, right up until mid-November, was driven by the US decision in early September to end negotiations. Japanese leaders knew that war was inevitable if talks were not restarted. Japan’s whole economic structure, plus the maintenance of its fleet and modern army, depended on oil and raw materials. Without access to these resources, Japan would decline so precipitously that it would become vulnerable to whatever power decided to pick it off in the future. The September–October Liaison and Imperial Conferences made these facts perfectly clear to any official who was not already keenly aware of them.
At the critical Imperial Conference of September 6, held two days after Washington’s change of heart, Hirohito was asked to approve a document stating that Japan would take “all possible diplomatic measures” to get trade restored by October 10. If these measures failed, however, a decision to commence hostilities would be made immediately. And this was no longer just an attack on the Dutch East Indies, as the army had previously wanted. It would be war against “the United States, Britain, and the Netherlands.”
Konoe began the conference by outlining the general problem. Given the coordinated policies of the US, British, and Dutch governments,
if we allow this situation to continue, it is inevitable that our empire will gradually lose the ability to maintain its national power.… [Japan] must try to prevent the disaster of war by resorting to all possible diplomatic measures. If the diplomatic measures should fail to bring about a favorable result within a certain period, I believe we cannot help but take the ultimate step in order to defend ourselves.
Navy Chief of Staff Nagano then told Hirohito that Japan could not delay a decision for war for long. Otherwise, the dwindling of oil supplies “[would] cause a gradual weakening of our national defense.” Meanwhile, the United States was quickly building up its military strength. By mid-1942, “America’s military preparedness will have made great progress, and [it] will be difficult to cope with her.” War would have to be waged now, should diplomacy fail.
Nagano spoke candidly of the low probability of victory. Even if Japan had initial successes, the United States would fight a prolonged war using “her impregnable position, her superior industrial power, and her abundant resources.” Japan did not have the means “to make [the Americans] give up their will to fight.” The only hope lay in Japan establishing a strong position early in the war and then waiting on “developments in the world situation.” These utterances are significant considering how often scholars claim that Japanese leaders entered the war believing it would be short, or that the United States did not have the will to fight more than a year or two.28 Nagano not only dispelled such illusions but also implied that only if the United States got bogged down in a war with Germany (the “world situation”) would Washington seek some sort of peace, and even then only after a long war in the Pacific. He ended by reiterating that Japan must spare no efforts to find a way to avoid war. But it also could not allow itself to decline, only to face an aggressive opponent later and with less power. His remarks nicely captured the tragic choice that would hang over Japanese ministers for the next three months: either allow decline and trust that adversaries would not destroy Japan, or attack now, hoping for a negotiated peace after a prolonged war.
Sugiyama spoke up to agree with Nagano about the problem of decline. Yet he went on to reassure the emperor, whose enthusiasm for a northern advance was known to all, that Japan was still continuing preparations for war against Russia. By spring 1942, a full-scale attack could be launched to take advantages of any “changes in the Northern situation.” His remarks reiterated what the group already took for granted: the nation’s core objective was the destruction of Soviet power. The northern operation would go ahead even if there was peace in the south. Conversely, if war in the south happened, southern raw material bases had be to secured quickly to facilitate the northern spring offensive.
Suzuki Teiichi, director of the Planning Board, informed Hirohito that the country’s only problem lay in its access to raw materials. Japan’s economy had developed through trade, he reminded the emperor, and depended on foreign sources for its vital goods. Due to Anglo-American economic sanctions, though, Japanese national power was declining day by day. It was thus “vitally important” for Japan’s survival that it “establish and stabilize a firm economic base.” There were no illusions about what Japan would have to face. If it went to war in the south, Suzuki cautioned, its short-term productive capacity would be halved. And even if the raw materials of Southeast Asia fell easily into Japanese hands, it would take two years to make full use of them.
Given this sobering assessment, it is unsurprising that Hara was not happy with the solutions offered by Konoe and his colleagues. Speaking for the emperor, he told the group that in a time of such poor US-Japanese relations, conventional diplomacy was no longer enough and every possible means should be taken to resolve the situation. The emperor had been led to believe that diplomacy and military preparations were to be undertaken simultaneously. Yet as he read the policy documents, it seemed that the emphasis was on war, not diplomacy. Navy Minister Oikawa quickly tried to reassure Hara that diplomacy had equal importance, as the prime minister’s determination to visit the United States showed. On this basis, Hara (i.e., the emperor) gave his consent to the policy proposal. Nevertheless, at the end of the meeting, Hirohito took the highly unusual step of voicing his concerns. He read a poem composed by his grandfather, Emperor Meiji: “All the seas in every quarter are as brothers to one another. Why, then, do the winds and waves of strife rage so turbulently throughout the world?” This statement was swiftly understood as an imperial censure of any tendency to favor war over diplomacy. Every possible effort for a peaceful solution would have to be made before war with the United States would be sanctioned (JDW, 133–51).29
The ministers’ subsequent actions showed that they got the message. But it was now apparent that if Washington refused to meet Japan halfway, war would have to begin sooner rather than later. The pressure to attack south if diplomacy failed was coming from two sources. One was the simple fact that the US navy would be much stronger in another year, leaving little hope of even a stalemated victory. The other was the army’s constant point that the military operation in the north had to start after February 1942, meaning that raw materials from either the United States (through diplomacy) or Southeast Asia (through war) had to be acquired before then. Many scholars have noted the strong preventive motivations of Japanese thinking by fall 1941.30 Yet they overlook the larger geopolitical picture: that from Tokyo’s perspective, it was still even more critical to stop the rising Russian colossus than to prevent the growth of the US giant. Destroying Russia by taking advantage of the German-Soviet war had been the emperor’s primary goal at the July Imperial Conference, and would remain the overarching Japanese aim through fall 1941, as the continued buildup in Manchuria confirms.
In the five Liaison Conferences in September, discussion focused on the state of diplomatic negotiations with Washington. Because Nomura’s dispatches had downplayed the roadblocks to a deal that Roosevelt and Hull were now putting up, Japanese ministers maintained an overly optimistic picture of the true state of relations.31 Only on October 2 did Nomura finally tell his foreign minister that the negotiations were deadlocked and the chances of a summit were remote. At the Liaison Conference two days later, both the army and navy chiefs of staff argued that Japan now had to move immediately to war. War Minister Tojo, who over the previous six months had shown himself to be a strong supporter of the emperor’s desires, was more cautious. Because this matter was “extremely critical,” he warned, Japan had to avoid any quick decision and give the matter more study (JDW, 179–81).
At the next conference (October 9), navy leaders were now reticent about war, too (see ibid., 181–84). By October 12, however, two days after the deadline for a deal and with nothing to show for all his hard work, Konoe knew he was in trouble. At a private meeting that day with just the foreign, war, and navy ministers (Sugiyama was excluded), Konoe tried to persuade Tojo that there was still hope for negotiations. Tojo was not convinced. He agreed with the army that war was the logical choice and Konoe should resign, to be replaced by a leader that could, if necessary, take Japan into war. Both men concurred on the choice of Prince Higashikuni Naruhiko as an individual who, as both a blood relative of the emperor and army officer, could achieve the unity needed for the crucial next steps. But Hirohito, not wanting a relative to be associated with taking Japan into war, surprised Tojo by asking him to become the new prime minister. Tojo was seen as the individual who could control the army and ensure that the emperor’s wishes were carried out.32
With reluctance, Tojo accepted his assignment, and appointed Togo Shigenori as foreign minister. By this point, resource shortages were already playing havoc with Japan’s military buildups both north and south. At the Liaison Conference on October 27, Finance Minister Kaya Okinori began the meeting with the key issue: “the acquisition of vital materials.” Suzuki explained that by the end of 1942, Japan would have exhausted all of its stockpiles. Restrictions on both civilian and government resource allocations had already been imposed. If further restrictions were required, “the productive power of the country will decline.” Tojo, speaking now as both prime minister and war minister, replied that the army was making its preparations “with emphasis on getting ready for the Soviet Union,” and would still be able to attack Russia in 1942 despite the shortages, as long it were given the same allocations as before (JDW, 190–93).
Follow-up conferences on October 28 and 30 focused on the possibility of using synthetic oil to overcome Japan’s supply problem, and whether southern operation could begin as late as March 1942 to give more time for diplomacy. Suzuki argued forcefully that Japan could not produce enough synthetic oil to overcome the oil shortage. The army, citing material shortages, rejected any delay past November. There were also discussions on whether Japan should simply accept the US demands. All participants except Foreign Minister Togo agreed that Japan would fall to the status of a vulnerable third-rate power if that came about. Thus three—and only three—options remained for future discussion: Japan should “avoid war and undergo great hardships” (i.e., accept the status quo); Japan must begin a war in the south immediately; and Japan should “decide on war but carry on war preparations and diplomacy side by side.” The participants agreed to meet on November 1 to make a final decision (JDW, 193–99).
What is perhaps most surprising about these late October meetings is that option number three was still on the table. The September 6 Imperial Conference had agreed that if there was no diplomatic solution by October 10, a decision for immediate war would be made. Yet here it was late October, with former army general Tojo now in the prime minister’s spot, and everyone but the army leaders was hesitating. When the group reassembled on November 1 for what would be a historic seventeen-hour debate on Japan’s future, the army leaders found themselves outvoted and had to accept another postponement of the final deadline for a war decision—this time to November 30. Proposal number three—to continue war preparations but keep striving for a diplomatic solution—had won out.
The discussion on November 1 of the first option nicely illustrates how actors in severe decline worry about the future power and intentions of rising states. If Japan put off war, Kaya asked, and “three years hence the American fleet comes to attack us, will the Navy have a chance of winning or won’t it?” Nagano replied that nobody could answer that question. Kaya then asked whether the United States would indeed attack Japan. Nagano answered that he wasn’t sure, but that the “chances are 50–50.” Kaya and Togo suggested that it was unlikely that the United States would attack. Nagano’s response was that “the future is uncertain; we can’t take anything for granted.” And in three years time, he noted, the United States would be that much stronger.
In the middle of the long conference, as it was becoming clear that option three would win out, Tojo (backed by Togo) turned to the army leaders and asked them to make an important promise. If Japan was going to make one final diplomatic effort, “you must give your word that if diplomacy is successful we will give up going to war.” In the subtle give-and-take tradition of Japanese decision making, this was a highly unusual and almost-brutal request. Tojo and Togo were essentially asking army leaders—in front of the whole group, no less—to put their honor on the line and refrain from pushing the nation into war should Washington prove accommodating.
Not surprisingly, the discussion suddenly took a nasty turn. Army Vice Chief of Staff Tsukada Osamu quickly responded that this was “impossible”; such a promise would throw the Supreme Command into confusion. Sugiyama jumped in to support Tsukada. Navy Minister Shimada Shigetaro then inquired if it might be acceptable to negotiate up until two days before a war started. Tsukada shut him down: “Please be quiet. What you’ve just said won’t do.” A heated debate followed. The group decided to take a twenty-minute recess, during which the army came to accept a compromise suggestion that negotiations continue until five days before war was to begin. By doing so, the army was essentially agreeing to Tojo’s demand that it accept a peaceful solution if one could be arranged before the deadline. The note taker summed up the mood of the room as this marathon seventeen-hour conference came to an end:
In general, the prospects if we go to war are not bright. We all wonder if there isn’t some way to proceed peacefully. There is no one who is willing to say: “Don’t worry, even if the war is prolonged, I will assume all responsibility.” On the other hand, it is not possible to maintain the status quo. Hence one unavoidably reaches the conclusion that we must go to war [should diplomacy fail]. (ibid., 200–207)
The significance of the above exchange between Tojo and his former military colleagues lies not in its unusual frankness but rather in the fact that it is the final nail in the coffin for all arguments suggesting either that the military hijacked the Japanese state and drove it into war, or that some cultural or ideological dementia had overtaken the Japanese leadership prior to the attack on Pearl Harbor.33 As in previous Liaison Conferences, the discussion was open and wide ranging, revolving around what would be best for the Japanese state instead of what was best for some organizational group or individual. And when army leaders proved intransigent, they were coerced into accepting that a diplomatic solution, even at this late date, was still the preferred option. A rational lesser-of-two-evils decision for war might still have to be made on December 1. But the prime minister and his foreign minister, acting in the name of the emperor and state, were still in control. The military obeyed them, and not vice versa.
On November 5, another Imperial Conference was held to secure Hirohito’s approval for the decision agreed to four days earlier. Diplomacy with Washington would revolve around two proposals, A and B. Proposal A was based around the US version of the draft agreement from early June. As it had on September 4, Japan would again promise to “act independently” on the Axis question and accept the open door. On the third big issue—that of China—Japan would again agree, as it had on September 4, to leave China as soon as possible (within a maximum of two years) after a Sino-Japanese peace agreement and withdraw from Indochina immediately after that agreement. Yet unlike the September 4 draft, which had removed Japan’s request to station troops in northern China, the Japanese now returned to the idea of leaving some troops there for a “necessary period,” presumably to fight Communism as per the June draft (JDW, 209–10; cf. FRUSJ, 2:600). The United States, one will recall, had made a concession on the question of troops in northern China when things were going well in early June. But in the face of current US intransigence, this was now likely to become a major stumbling block.34
The backup plan, proposal B, was designed as a temporary modus vivendi should proposal A prove too hard to negotiate given the limited time frame. It proposed that the two sides simply go back to the pre–July 1941 status quo: Japan would get out of southern Indochina, and the United States would restore most of its oil exports. It was this proposal that would tempt Roosevelt and Hull briefly to the bargaining table again in November.
When the November 5 Imperial Conference started, Tojo began by telling Hirohito of the November 1 decision to pursue both diplomacy and military preparations. Togo then noted that diplomacy was currently deadlocked, with Washington unwilling to make any concessions. These two presentations were followed by Suzuki’s long analysis of the resource problem. Expectations of future trade were pessimistic: “the probability that we will experience increased difficulties in obtaining materials is high.” In fact, Japan might be drawn into war, “even though we wish to avoid it,” to secure its supply. This was not just a question of keeping Japan’s military strong. Suzuki detailed a long list of materials critical to Japan’s domestic needs that were also unobtainable due to the Anglo-American embargo. Because the United States had free access to raw materials and Japan did not, he concluded, “differences in defensive power” would open up over time.
Finance Minister Kaya reinforced this gloomy picture, observing that without raw materials, the national economy would collapse no matter how perfectly the government managed it. Sugiyama followed with a detailed analysis of the military situation. With the US arms buildup, the power balance was becoming increasingly unfavorable. As for the north, Sugiyama underscored that the Russians had suffered massive losses on the German-Soviet front, and had sent forces westward from Siberia equal to thirteen infantry divisions, thirteen hundred tanks, and a similar number of airplanes. In the short term, then, Russia would not attack Japan’s army in Manchuria. Given that it was being reinforced by US aid, there was still a future danger. Thus Japan “must conclude its operations in the South as quickly as possible, and be prepared to cope with this situation”—namely, strike north.
It was at this point that Hara spoke up on Hirohito’s behalf. He commented that despite the September 6 consensus—that Japan would concentrate on negotiations with Washington—to his regret no agreement had been signed. Togo informed him that US concerns regarding the Axis pact had been resolved, since Japan had agreed not contest any US entry into the European war. The China question, however, had become the main stumbling block. The United States was now balking at Japanese requests to station some troops in northern China after a peace deal with Chiang.35
Hara turned to the specifics of proposals A and B. Japan had to settle the China Incident, he stressed. So it seemed odd to him that plan B had said nothing about China per se. Togo noted that given the short time remaining, a quick agreement was unlikely under plan A. Even the much simpler plan B was unlikely to succeed. Hara then turned to the military “to explain what will happen if negotiations break down.” The key issue was Japan’s ability to defeat the US navy. Nagano said that in the short term, Japan did have a measure of superiority. The US fleet was bigger overall, but with 40 percent of it in the Atlantic, Japan had a 7.5 to 6 ratio in its favor. Yet even if Japan won an initial decisive battle, “war will continue long after the Southern Operation.”
Hara then summarized the emperor’s perspective in a long statement. It was still essential to try to get an agreement, he emphasized. But with little prospect of a diplomatic breakthrough, if Japan missed the present opportunity to go to war, “we will have to submit to American dictation.” The country must therefore accept a decision for war should diplomacy fail. With his remarks indicating Hirohito’s approval of the plans, the meeting quickly came to a close. Before it did, Tojo summed up the situation. There was still some hope for success in the US-Japanese negotiations, he contended, since the United States needed time to complete its buildup. He somberly noted the uneasiness felt around the table regarding the prospect of a protracted war. On the other hand, how could Japan let the United States do as it pleases? “When I think about the strengthening of American defenses in the Southwest Pacific,” he declared, “the expansion of the American fleet, the unfinished China Incident, and so on, I see no end to difficulties.… [Still] I fear that we would become a third-class nation after two or three years if we just sat tight.” The participants filed out of the conference knowing that unless a diplomatic miracle occurred within three weeks, Japan would be embarking on its most dangerous venture since the Meiji Restoration (JDW, 208–39). Yet aside from perhaps Sugiyama, not one person in the meeting expressed any degree of optimism regarding Japan’s chances for even a stalemated victory. All knew the risk of defeat and national destruction was high. Nevertheless, in the face of a continued US embargo and Roosevelt’s unwillingness to accept the deal for China that had emerged in early June, the Japanese saw no other option but to go south. War was quickly becoming a true lesser-of-two-evils choice.
THE FINAL NEGOTIATIONS
From the above, we can see that the real puzzle surrounding the last three weeks of November was not why Japan was preparing for total war in the south but instead why Roosevelt and Hull were still unwilling to make a deal to avert the impending disaster. Ever since negotiations began in April, the pair had been seeking to avoid a Pacific war so they could concentrate on the war against Hitler. By early September, Japan had conceded to their primary initial concern: Tokyo would operate “independently” of the Axis should the United States enter the European war. With Roosevelt, Hull, and the military all wanting to avoid a draining war in the Pacific—especially one taking place before the US buildup was complete—one would think the rational strategy was at least to buy some more time through a compromise.
In fact, Roosevelt and Hull did decide, for good geopolitical reasons, to seek such a deal. The two key military advisers, Army Chief of Staff George Marshall and Chief of Naval Operations Harold Stark, had been arguing for weeks that Japan was being driven into war by the oil embargo. With US forces not yet ready for war, they wanted an agreement that would give the United States at least an extra six months. The president was sympathetic to this reasoning. In an hour-long meeting with Stimson (diary, in PHA, 10–12:5431) on November 6, Roosevelt told Stimson that he was “trying to think of something which would give us further time.” By mid-November, with Japan’s desperate situation plain for all to see, Roosevelt decided to proceed with a modus vivendi strategy. In a handwritten memorandum to Hull on or about November 17, he said that the United States should “resume economic relations—some oil and rice now—more later,” and it could again agree to play the role of intermediary to “introduce [the Japanese to the] Chinese to talk things over” (FRUS [1941], 4:626).
In return for US help, Roosevelt told Hull, Japan had to accept two things. First, consistent with long-standing US concerns, Japan must agree not to invoke the Tripartite Pact even if the United States got into the European war. Tokyo was unlikely to put up a fuss here, since it had already accepted this condition in early September. The president’s second condition, in contrast, was much harder to secure, and quite revealing of his deeper worries. Japan had to promise to “send no more troops to Indo-China or [the] Manchurian border or any place South—(Dutch, Brit., or Siam)” (FRUS [1941], 4:626; Heinrichs 1988, 208, 258n95). Once again, we see that Roosevelt was not just concerned about a southern advance but also a move north. He could not accept a deal that helped Japan exit China and kept the peace in the south if it only meant the redeployment of forces for an attack on Russia. This concern had been the root of the demand in previous negotiations that Japan promise to keep the peace in the entire Pacific region. Roosevelt now wanted something much more explicit. Japan not only had to agree not to attack Russia; it had to agree not to put additional forces in the north. His inclusion of Manchuria on the list of restricted countries could only have had one purpose: to get Japan to commit to peace with Russia while providing a simple mechanism—no further troop movements from either southern China or the home islands—to verify compliance.
By mid-November Roosevelt knew he was in a bind. Defeating Hitler was still the main priority. Through the fall, Roosevelt had continued to put US ships in harm’s way in the Atlantic in order to spark an incident that would spiral to war. Hitler refused to take the bait. Notwithstanding Roosevelt’s frustrations, the revisionist thesis that the president, seeing Berlin’s unwillingness to escalate, sought a backdoor to war with Germany by provoking a war with Japan is without foundation.36 Not only is there little direct documentation supporting it, but there is also one decisive fact against it: Roosevelt and Hull actively contemplated making a deal with Japan three times over six months—in early June, late August, and finally and most significant, late November. If one wants to provoke an adversary into attacking to get into a war elsewhere, one does not expend significant energy finding a way to satisfy its demands.37
Roosevelt’s goal was exactly the opposite of what the revisionist thesis supposes: he wanted to avoid all conflict in the Far East if Japan would accept his two (and now only two) conditions—to delink itself from Germany, and remain peaceful everywhere in the Pacific and, most important, in the north vis-à-vis Russia. It is self-evident why this scenario was his first preference. It allowed him to focus solely on the European war while ensuring that Russia could play its key role in defeating Nazi Germany.
In international politics, however, one cannot always achieve one’s first preference, and Roosevelt knew he might have to choose between two far less desirable options. The first was simply a war with Japan where the Japanese struck south to gain access to the raw materials being denied to them. The second was a compromise deal with Japan along the lines of his note to Hull, but one that Tokyo broke in order to attack Russia. Getting the Japanese to agree to not moving additional forces into Manchuria might help bind Tokyo to a regional peace. A severe form of the commitment problem nevertheless hung over any deal.38 In the end, Tokyo might decide to build up in Manchuria secretly or just attack Russia with what it had already deployed there.
Thus Roosevelt, in even considering a modus vivendi, was entering into a delicate and risky diplomatic game. If Japan agreed to his terms, he could accept the deal and hope that Japan lived up to its promises. Then if and when his Atlantic scheme or some other ploy proved successful and the United States found itself at war with Germany, his first preference would be realized. Yet the downside risk was clear. Renewing trade with Japan and helping it end the China Incident might lead to the worst of all outcomes: a Japanese war with Russia at a time when the Soviet Red Army was the only real barrier between Hitler and the German domination of Eurasia.
Roosevelt was perfectly aware of his military’s desire to buy time. On November 5, the war department sent him a memorandum noting that the US navy was presently “inferior to the Japanese Fleet” and incapable of offensive operations. With the United States and Britain still reinforcing the Philippines and Singapore, they would lack the capacity to deter Japan until early spring. The chiefs of the army and navy (Marshall and Stark) thus agreed that war between the United States and Japan “should be avoided while building up defensive forces in the Far East.” The “primary objective” of US policy, after all, remained the defeat of Germany, and any war with Japan “would only greatly weaken the combined [US-British] effort in the Atlantic against Germany” (PHA, 13–14:1061–62). The importance of buying time would be a constant refrain of Marshall and Stark through the remaining days of November (see Trachtenberg 2006, 120–21).
The president also understood the fragile nature of the Japanese-Russian peace that had held despite German advances. Roosevelt knew Japan was readying for war, but was waiting for a decisive German victory and possible Sino-Japanese peace before taking the plunge. From early July through November, Roosevelt received a series of top-secret estimates the Far Eastern situation directly from the war department.39 The July 17 report indicated that in spite of its Manchurian buildup, Japan might stay on the defensive, given that “the major part of the Japanese Army … [was] pinned down in China.” The August 16 report stated that Japan had increased its strength in Manchuria to six hundred thousand troops (from an original two hundred thousand in May), yet was unlikely to attack in the short term because Germany’s offensive had “gone awry.” The analysis of September 5 reiterated that Japan would refrain from war against Russia until peace with China had been secured. The September 23 evaluation concluded that Tokyo probably sought a summit to hide preparations for an all-out northern attack “timed with the expected Russian collapse in Europe” (PHA, 13–14:1342–57).
The October 2 report to Roosevelt was the most remarkable of the bunch. Should the Sino-Japanese War end, it asserted, Japan would free up twenty-one divisions and a thousand aircraft, most or all of which would be diverted to Manchuria. Tokyo would then initiate war if either Russia collapsed in the west or Japan achieved superiority due to transfers of Soviet troops westward. Given this, Washington should exploit the continued China quagmire to prevent a Japanese move north.
The initial feeling of revulsion over this apparent utilization of China as a cat’s paw in our plan of strategy will be alleviated by an examination of the situation of the anti-Axis powers in the light of cold reason. Our objective is the destruction of Nazism, and all-out aid to those powers actively engaged in resisting its aggressive drive for world domination. Russia is, as a matter of expedience, an ally in this cause. We must, among other things, do what we can with what we have at our disposal to aid Russia in her struggle with Germany. Any action on our part, therefore, which would liberate Japanese (pro-Axis) forces for action against Russia’s rear in Siberia would be foolhardy.
The document did note that Washington might use a summit to buy time for its own buildup. Most important, however, was the immediate objective of “weaken[ing] Hitler in every way possible.” In any negotiations, then, Washington had to ensure Tokyo showed “substantial evidence of sincerity, not to attack Russia in Siberia.” Only such a guarantee would free Russia, psychologically and militarily, for stronger opposition to Nazi Germany (ibid., 13–14:1357–59).
Given the above, it is impossible to accept the idea, critical to almost every historical argument, that Japan’s refusal to leave China in late November was the key reason for US unwillingness to renew the oil trade. US officials well understood that ending the Sino-Japanese War was Tokyo’s objective, not Washington’s, and that Washington had a strong interest in keeping Japan bogged down in China. For if Japan invaded Siberia, the war with Germany would then be as good as over, with Hitler the head of a new Eurasian superpower.40
But did Roosevelt himself really understand the significance of these startling analyses? He certainly did. On October 10, Roosevelt told British ambassador Halifax that he feared that Japan would attack Vladivostok (Heinrichs 1988, 191). Five days later, on October 15, he wrote to Churchill that the situation with Japan was “definitely worse and I think they are headed north” (CR, 1:250). The next day, October 16, he held a meeting with his top advisers at the White House. From diary notes made by Stimson (LC, October 16, 1941), we can see that Roosevelt was now actively considering the necessity of drawing Japan into war. The United States, the president argued, faced the delicate question of the “diplomatic fencing” that had to be carried out by Washington “so as to be sure that Japan was put into the wrong and made the first bad move.” Not coincidentally, Chief of Naval Operations Stark sent a telegram to his Pacific and Atlantic commanders that day stating that Japan was unlikely to make a deal with the United States, and as such, “hostilities between Japan and Russia are a strong possibility.” He went on to suggest that because the United States and Britain might be blamed for Japan’s economic situation, there was also a risk that Japan might attack them. The commanders were advised to engage in preparatory deployments, but to avoid maneuvers that “constitute provocative actions” (PHA, 13–14:1402). Such language indicates not only that the main US fear was of an attack on Russia, not Southeast Asia, but also that Japan had to be allowed to strike the first blow.
The information gathered by the war department after October 16 only reinforced the worry that Japan might attack Russia soon, despite the onset of winter in Manchuria. The department’s October 21 analysis to Roosevelt stated that the Soviets possessed air superiority, but this could be quickly altered “by [Japan’s] shifting from China … the bulk of [its] air forces.” On the ground, Japanese troops now numbered 684,000 to Russia’s 682,000. If more Siberian troops were moved westward and the Kwantung Army achieved a two-to-one superiority, it was “highly probable” that it would take the offensive. If this rose to a three-to-one ratio, the probability “will become a certainty” (ibid., 13–14, 1360–61).
Given this desperate situation, the report argued, Washington had to do everything possible to maintain Russia’s military equality. Two policies stood out. The first, a continuation of the cat’s-paw strategy, was to increase aid to China to allow it “to continue to pin to the ground … the bulk of the Japanese Army.” The second was to help Russia by bolstering aid to Soviet armies in both Europe and Siberia, meaning a funneling of more lend-lease aid through Iran and Vladivostok (ibid., 13–14:1360–61). A follow-up report to Roosevelt on November 1 noted that Japan now had thirty-three divisions in Manchuria—a large jump from the nineteen divisions of early September, and a quadrupling of the eight divisions there prior to Germany’s invasion of Russia. The Japanese might still increase southern troop levels, the report suggested, but this action might hurt the northern army and cause Japan to “‘miss the bus’ when [its] chance comes for [the] invasion of Siberia” (ibid., 13–14:1361–62).
By mid-November, therefore, Roosevelt found himself walking a geopolitical tightrope. He could only negotiate toward a modus vivendi that might give him his preferred outcome—total peace in the Pacific even if the United States entered the European war—at the risk of sending Japan north against Russia. Throughout the first half of November, encouraging reports were arriving indicating that Germany’s push toward Moscow had been halted (Heinrichs 1988, 201). Then on November 13, Grew forwarded a report strongly maintaining that Japanese leaders were reaching the end of their economic rope. Without the renewal of raw material imports, Japan’s economy “[could not] withstand the present strain very much longer,” meaning Tokyo “must accept the inevitable or fight” (PHA, 19–20:4051–57). It was likely this information combined with the sense that Russia might survive the winter that led Roosevelt to reverse policy and support a compromise deal on November 17. With Japan deterred a bit longer in the north, Roosevelt could buy time for his southern buildup.
This was now the third time since March that Roosevelt had restarted US-Japanese talks in hopes of reaching an agreement to keep the Pacific peace. Just like the other two instances, he would subsequently decide to end them abruptly, this time on November 26, much to the frustration and mystification of the Japanese diplomats (and US military). It is crucial to remember that there is no way to explain Roosevelt’s on-again, off-again interest in a negotiated peace by any version of the “backdoor-to-war” thesis. If he was manipulating Japan into war to get the United States into the European war, then he should have kept the pressure on throughout November (and through spring and summer, for that matter).
In fact, the only riddle left to explain is why Roosevelt would push for an immediate modus vivendi, reach a consensus on the need to secure one with his top advisers on Tuesday, November 25, and then reject the idea the very next morning. For anyone studying the Pacific War, this about-face on November 26 is the single most puzzling event of the whole March to December 1941 period. Roosevelt’s unwillingness to proceed with a modus vivendi seems even odder once we consider the exact sequence of events. The president knew from discussions with Nomura and his associates that Tokyo desired a modus vivendi should a broader agreement based on the May–June and August–September talks not pan out.41 This of course was plan B from the Imperial Conference of November 6: Japan would return to the pre–July 1941 status quo, pulling out of southern Indochina if Washington reinstated at least part of US oil exports. Thus when Roosevelt sat down at noon on Tuesday, November 25 for an hour-and-a-half meeting with Hull, Stimson, Stark, Marshall, and Navy Secretary Frank Knox, he knew almost exactly what the Japanese would both propose and accept. The draft agreement settled on by the end of the meeting emphasized Japan’s removal of forces out of southern Indochina. In return, Washington “[would] permit the export to Japan of petroleum … upon a monthly basis for civilian needs” (FRUS [1941], 4:662–64). Stark and Marshall, who had been pressing Roosevelt for a month for such a deal, were delighted. They left town Wednesday afternoon believing that the two sides would avoid war for at least a few more months, giving the military time to expand its presence in Hawaii, Guam, and the Philippines.42
When Roosevelt and Hull met early the next morning, Wednesday, November 26, they decided, in Hull’s words, to “kick the whole thing over.” Without consulting the other participants of the Tuesday meeting, they also decided to immediately call the Japanese diplomats into the White House to break the news to them. That afternoon, Roosevelt told Nomura and his associates that there would be no modus vivendi and in fact Tokyo must now agree to ten US demands before oil would be renewed (FRUS [1941], 4:645–46). Because Magic had made known the November 29–30 deadline for a Japanese decision to attack, Roosevelt and Hull had not expected to meet with Nomura until Friday or Saturday. So to ask him to come in Wednesday afternoon, not to secure a deal, but to lay down ten demands that they knew would be unacceptable, is certainly odd behavior.
That Roosevelt and Hull understood that the Japanese leadership could not possibly accept the ten demands is quite clear. Nomura and his colleagues on Wednesday afternoon, despite their shock, did not reject the ten demands but instead stated only that they would relay them to Tokyo. Yet the next day, the war department sent a message to the Pacific command telling it straightforwardly that diplomacy was now essentially over, war was possible at any moment, and were it to occur, “the United States desires that Japan commit the first overt act” (see PHA, 13–14:1328–29; Prange 1981, 402–3).43
So the puzzle remains: Why would Roosevelt and Hull suddenly end any chance of a diplomatic solution on November 26, when only the day before they had agreed to pursue one? Obviously something must have happened between 1:30 p.m. on Tuesday and the Roosevelt and Hull meeting on Wednesday morning to change their minds. To try to explain the riddle, historians have generally focused on the roles played by the Chinese and British leaders on the night of November 25. There is evidence that Chiang communicated to Stimson on the night of November 25 that he was upset about the possibility of a deal with Japan. In Chiang’s memorandum, written to Chinese envoy T. V. Soong and passed by Soong on to Stimson, Chiang wrote that the morale of the Chinese people would collapse if they believed that Washington was sacrificing their interests. If the Chinese army then fell apart, “[such] a loss would not be to China alone” (FRUS [1941], 4:660–61). Historians speculate that this message changed Roosevelt’s calculation: he realized that he could not maintain China as an ally if he made a deal with Japan.44
This explanation is reinforced by the British leadership’s reaction to a possible US-Japanese deal. Early on the morning of October 26, Churchill wrote to Roosevelt, noting that he had read Halifax’s dispatches regarding the proposed modus vivendi. There was only one point that disquiets us, Churchill told Roosevelt: “What about Chiang Kai-shek? Is he not having a very thin diet? Our anxiety is about China. If they collapse, our joint dangers would enormously increase” (FRUS [1941], 4:665). Combined with Chiang’s message, so the reasoning goes, Roosevelt and Hull had no choice but to end the effort to reach a compromise (Dallek 1979; Kimball 1997).
From a theoretical perspective, it is perfectly in line with the trade expectations argument to accept the traditional historians’ take on the events of November 25–26. After all, it suggests that an exogenous third-party issue—the survival of China as an ally—compelled Washington to maintain a hard-line policy. And it was this policy that confirmed Japan’s negative expectations for future trade, making war inevitable. Yet the “Chinese-connection” assertion must be called into question. It was another third-party issue—the now-familiar one of Russia’s survival in its war with Germany—that was almost certainly the key to Roosevelt and Hull’s about-face. In what follows, I lay out the evidence for this alternative perspective. But it is worth remembering that whether one accepts the Chinese-connection thesis or Russian-connection argument, both uphold the book’s contention that third-party constraints can drive two states into war over economic issues.
The problems with the Chinese-connection thesis are threefold. First, Roosevelt and Hull had known for months that Chiang was upset about the prospect of a US-Japanese agreement, believing it would reduce US aid. Chiang’s panicky telegram on the night of November 25 was not telling them anything new. Indeed, for the previous four days, ever since Hull briefed the Chinese and British ambassadors about a possible modus vivendi, the Chinese government had been registering its disapproval.45 In particular, on Monday, November 24, the day before Roosevelt’s inner circle met, Hull was told of Chiang’s “rather strong reaction” to the modus vivendi, believing that Washington was now inclined to appease Japan at China’s expense (FRUS [1941], 4:646–47). In light of known Chinese anxieties, why would Roosevelt and Hull proceed with the compromise idea on Tuesday and then kick it over the next morning?
The second problem is simply that Chiang’s concerns were not valid, and thus could have been easily placated by subsequent US behavior had Washington and Tokyo made a deal. The US drafts of the proposed agreement from November 21 to 24 made it clear that Japan would have to reduce force numbers in northern Indochina even as it pulled out of southern Indochina. The November 24 draft in fact specified that Japan could only keep a maximum of twenty-five thousand troops in northern Indochina after the agreement—a number deemed by the US military as too small for a successful attack on the Burma Road (the main conduit for supplies to Chungking) (ibid., 4:643). Moreover, the draft made absolutely no promise that US aid to China would not increase. In short, by stabilizing China’s southern frontier, keeping the Burma Road open, and allowing Washington to sustain its aid to Chiang, Washington would be actually helping China’s ability to fight Japan, not hurting it.
It is evident that Roosevelt understood these points and had little tolerance for Chinese concerns, as shown by his meeting with Soong on Wednesday, November 26. Early in the morning that day, just before throwing over the modus vivendi, Roosevelt had been on the phone with Hull when the latter mentioned Chiang’s anxious telegram from the night before. As bystander Henry Morgenthau (FDRL, November 26, 1941), Roosevelt’s treasury secretary, recorded in his diary, the president breezily replied, “Well, send [the Chinese envoys] in to see me today and I will quiet them down.” When Soong met with the president that afternoon, Roosevelt led him to believe that a deal with Japan was still being considered, even though he and Hull had already rejected it many hours before. Stringing Soong along, Roosevelt told him that “he thought Chiang Kai-shek was much too excited.” Any modus vivendi would help keep supplies flowing to China along the Burma Road, Roosevelt argued, and would buy three to six months to get ready in the Philippines (ibid., FDRL, November 27, 1941).46
One might believe that it was Churchill’s intervention that made the difference, since the United States needed Britain on its side regardless of the validity of Chinese concerns. But this notion also has little foundation. Halifax had made it clear to Hull before November 25 that the British were anxious to avoid a two-front war in the Pacific and that London strongly supported the modus vivendi. Churchill himself had been signaling for months that he did not want war in the Pacific. So the only “new” piece of information was the passage in Churchill’s letter of November 26 asking if Chiang was on too thin a diet. Historians regularly quote this passage to support the notion that Hull suddenly realized that neither the Chinese nor British would support the deal. Hull’s (1948, 1080–82) own memoirs argue along these lines.47
Once we examine the surrounding content of Churchill’s letter, however, it is apparent that this letter could not have been decisive. Prior to the section on Chiang’s thin diet, Churchill tells Roosevelt that he has read Halifax and his foreign secretary’s assessment of the proposed modus vivendi. He goes on to say, “Of course, it is for you to handle this business and we certainly do not want an additional war.” Only then does he introduce the phrase “there is only one point that disquiets us” as a precursor to discussing Chiang’s thin diet. The larger context thus makes it plain that Churchill is giving authority for any decision over to Roosevelt and his main priority is, like the president’s, to avoid a two-front war. Churchill’s language is muted and circumspect. By saying “there is only one point,” he is deliberately moderating his tone to indicate that he is only making a suggestion. And by highlighting Chiang’s thin diet, he is underscoring the importance of material aid for China rather than trying to reverse any US decision regarding a deal with Japan. Churchill had no reason to communicate opposition to a deal that he as well as everyone else knew was the only way to avoid a Japanese attack south at that point. So the idea that Hull later propagates—that he went to Roosevelt on the morning of November 26, and told him that given China’s opposition along with “the half-hearted support or actual opposition of the British, the Netherlands, and the Australian Governments,” the modus vivendi should not be presented—is pure fabrication (see Hull 1948, 1082). The British, Dutch, and Australians were already on board, and given their geographic vulnerability in Southeast Asia, understandably anxious to secure a peace.48
The third and final problem with the Chinese-connection argument is the most straightforward: it cannot explain why Roosevelt and Hull torpedoed possible negotiated deals on two previous occasions (in late June and early September). There is an evident pattern here—a serious US interest in an agreement and then the deliberate undermining of promising developments—that the argument about Chiang cannot cover. We have seen that the most likely reason for the two previous terminations of talks was Russia’s situation vis-à-vis Germany and whether Japan would exploit a Russian setback to attack Siberia. Could this again be the best explanation for the about-face on November 26?
Although we will probably never have complete smoking gun evidence for Roosevelt and Hull’s change of heart that Wednesday morning, the need to keep Russia alive in the global struggle against Nazism is the only explanation that fits all the available evidence, and certainly the only one that aligns with the evidence for the two previous shutdowns in talks. Roosevelt and his advisers, as we have seen, had been obsessed for months with the possibility that Japan would go north. Within this context, the information on Russia’s situation that arrived on the night of November 25 could only have been seen as deeply troubling. Back in early November, the German campaign that had begun in late August had finally been halted just a few hundred miles outside Moscow. For a few days Hitler had held his forces in place, and it looked as if he might pull them back to stabilize supply lines and wait for better weather. But instead of retreating, Hitler engaged in yet another perilous roll of the dice. On November 15, he began a new major offensive to take Moscow before the full depth of the Russian winter set in. For the next nine days, the front line ebbed and flowed in what had become the most crucial single battle of the war so far. On the night of November 25, however, new information came in indicating that the Germans had scored major victories that day. They had moved within thirty-one miles of the city, and were making a beeline for the Moscow-Volga Canal in the north and Oka River in the south. If either fell, the road to the Soviet capital would be open. Hitler was throwing everything he had into the struggle to take Moscow, knowing that if Stalin had to retreat behind the Urals, the war would be essentially won.49
As Morgenthau’s diary notes reveal, Roosevelt had become highly concerned with these developments by the early morning of Wednesday, November 26. That morning, Morgenthau went to the White House to discuss recent US efforts to increase the flow of lend-lease goods to the Soviet Union. Catching Roosevelt in the middle of his breakfast, Morgenthau handed him a memo. The news was discouraging. The number of ships that had delivered goods to Russia during November was less than a third of the planned amount. Yet to Morgenthau’s surprise, the president did not want to talk about the “Russian matter.” This, for Morgenthau, “makes me believe he knows the situation is bad.” (It was just after this that Roosevelt spoke on the phone with Hull, telling him he would “quiet down” the Chinese.)
At this point in his notes, Morgenthau reminds himself of the essence of the proposed modus vivendi (which at that point had not yet been kicked over). His spare summary is instructive in terms of what it mentions and what it leaves out. He writes that the United States is to unfreeze Japanese assets to allow Japan to buy as much oil as it needs for industrial uses and export as much as it wants. In return, the “Japanese are to agree not to attack Russia in Siberia.” Nothing more is mentioned. Morgenthau was well informed about the state of US-Japanese relations. Nevertheless, he makes no mention of a US demand to “get out of China”—surely a strange omission if Japan’s exit from China was the United States’ primary concern in these tense November days.50
Near the end of the meeting, Morgenthau’s diary relates, the phone rang again. This time it was Eleanor Roosevelt on the line. After some banter about the resignation of a lower-level official, Eleanor asked him about how things were going. Roosevelt replied, “Everything is terrible. The Russian situation is awful. Moscow is falling.” The British might also lose Libya, he continued. The situation, the president told his wife, “looks very bad.” Roosevelt seemed so worried by the course of events, Morgenthau observes, that he couldn’t finish his breakfast.51 The meeting wrapped up, and an hour or so later Hull arrived to see the president. It was in that one-on-one meeting that Roosevelt and Hull made their fateful decision to toss out the compromise and slap the ten demands on Japan that made war inevitable.
While perhaps not quite a smoking gun, Morgenthau’s diary notes are highly suggestive, to say the least. His notes are essentially the only contemporaneous documents available to shed light on the early morning of November 26, 1941.52 They show a president who had little concern for Chiang’s objections to a possible modus vivendi, but who was deeply anxious about the imminent fall of Moscow. Scholars clinging to the backdoor-to-war thesis might claim that the new information Roosevelt was receiving on Russia simply reinforced that he had to get the United States into the European war, and quickly. Notice, though, that even an immediate US entry into the war would have done nothing to help Russia in the short term, had Japan still gone north. The war for Eurasia would have been decided long before the first US soldier touched foreign soil.53 Germany would then have been able to redirect troops southward to finish the job of capturing the vast oil potential of the Middle East (as Roosevelt’s concern about the British in Libya suggests). If the continental balance of power was to be restored, Communist Russia had to be saved from a two-front war. There was no other option. Japan had to be drawn south to keep it from going north.
CONCLUSION
The analysis of this chapter provides strong support for the thesis that third-party factors can lead states into a downward trade-security spiral that forces them into severe economic restrictions and war. Neither the Americans nor the Japanese wanted war with each other when negotiations restarted in April 1941. The Americans wanted to fight Germany, not Japan, and the Japanese wanted to end the China conflict so that they could regroup for the long-desired war against the rising Soviet colossus. By early June, a deal had been hammered out that appeared to satisfy both sides. But after Nazi Germany struck Russia on June 22 and began its steady push toward Moscow, Tokyo would not commit to staying out of the German-Soviet war. Roosevelt and Hull kept hoping they could achieve their first preference—namely, peace in the Pacific as the United States entered the European struggle. By late November, however, it seemed clear that any deal with Japan to restore the oil flow would likely only send Japan north against Russia. Roosevelt and Hull thus reluctantly decided to keep the embargo on, knowing that pessimistic trade expectations would force Japan to make a plunge toward the oil-rich Dutch East Indies and into war with the United States.54 As undesired as such a conflict was, it would at least help preserve the Soviet Union as an ally, giving the United States a fighting chance to save the world from Nazi tyranny.
It is important to see that this explanation for the Pacific War not only aligns with the evidence but also covers the anomalies that exist for alternative explanations. Economic realism can explain Japan’s concern for its high dependence. It nonetheless cannot explain the Japanese desire to resolve their raw material supply problem through diplomacy, nor why Japan would switch from diplomacy to war by late November 1941. The constant of high dependency cannot explain a variable. For liberals, the Pacific War should be one of their best cases: as trade levels fell from 1939 to 1941, preexisting domestic pathologies should have been unleashed. Yet even the strongest argument for the role of domestic politics—Snyder’s (1991) notion that parochial groups, specifically the army and navy, hijacked the Japanese state and pushed it into war because of internalized imperial myths along with self-created domestic pressures—faces deep problems. For one thing, civilian leaders and Hirohito were very much involved in the decision-making process, with all top officials working hard to create a consensus for war. Moreover, many in the military, particularly in the navy, were reluctant to fight the United States, and as such, actively supported efforts to secure a deal that would restore Western oil exports. Finally, the documentary evidence shows overwhelmingly that the primary motive for war with the United States was the fear of long-term decline caused by the double whammy of economic sanctions and the US naval buildup. There is little proof that Japanese officials felt any pressure from the general population or lower-level military officers to start a war with the United States. And if they had felt such pressure and were just unwilling to express it, they would not have sought a negotiated peace across nine months of on-again, off-again talks.
Purely historical arguments fare no better. The backdoor-to-war thesis that sees Roosevelt manipulating Japan into war in order to get into the war in Europe falls apart in the face of one key fact: Roosevelt and Hull worked hard to get a peace deal for the Pacific and almost made such an agreement on three occasions. The argument of this chapter shows why they would pursue such a deal. It also demonstrates why they would have to repeatedly scuttle it, even at the risk of a US-Japanese war, when Tokyo refused to commit to not going north. The traditional assertion that Roosevelt and Hull kicked over the final negotiations because of Chiang also proves unsustainable. Chiang’s concerns were well known and largely immaterial, since Roosevelt and Hull were more than prepared to keep supplying China with economic and material aid in the event of a modus vivendi with Japan.
In the end, the war between Japan and the United States cannot be viewed as overdetermined by a mix of powerful forces. Rather, it was driven by one primary cause: Japan’s increasingly pessimistic trade expectations in a situation of high dependence. All other factors can be seen to have had little salience or acting merely as facilitating conditions. Needless to say, when we look at the war in its broader context, Japanese leaders—civilian and military as well as the emperor—are hardly blameless. Had they been willing to give up their four-decades-long drive to reduce Russian power in order to protect Japan’s economic and strategic position in Northeast Asia, a deal for peace in the Pacific could have been made. Yet in explaining the total war of 1941—and why it was launched against the United States and not against the Soviet Union—we must turn to Roosevelt and Hull’s decision to impose as well as maintain a complete embargo on oil and raw materials despite knowing that they were pushing Japan into war. The two understood full well that by manipulating Japanese trade expectations in either a positive or negative direction, they held the fate of the Pacific region in their hands. Still, the force of circumstances—the importance of keeping Russia in the fight against the greatest menace the world had ever seen—led them to make a lesser-of-two-evils choice for war in the south to avoid a war in the north.
In the final analysis, we can say that they made the right choice, both on moral and geopolitical grounds. The only irony is that by creating the conditions for Hitler’s eventual defeat, they also set down the conditions for a future bipolar struggle between Russia and the United States that would consume the next half century of world history. The country that Roosevelt helped save in 1941 would become the United States’ main rival and long-term threat a mere four years later. Great power politics, however, often requires that distasteful means be used to avoid having to face a disastrous outcome. And it is hard to argue that the destruction of Hitler’s Germany was not an end that justified any means.
_____________
1 This case represents, perhaps more than any other, a direct challenge to a fundamental premise of the bargaining model of war: that states fall into war when they lack information about the other’s power and resolve (Fearon 1995; see also chapter 1). As we will see, both US and Japanese leaders were fully aware of their adversary’s capabilities as well as willingness to fight, and yet war still occurred in spite of their best diplomatic efforts.
2 On the supposed irrationality of Japanese decision making, see especially Snyder 1991; Kupchan 1994; Taliaferro 2005.
3 On the navy’s pessimism regarding its chances of winning a war with the United States, see Tsunoda 1980, 258–59.
4 See Coox 1985; Montgomery 1987; Duus, Meyers, and Peattie 1989.
5 For the discussions between May 2 and June 17, see FRUSJ, 2:411–83.
6 With Japan showing no signs of readying for war on Russia, Hull could make this concession without worrying about the impact on the European war.
7 The United States also adjusted item 3 on economic cooperation to note the importance of commercial nondiscrimination, thereby reinforcing Washington’s interest in open-door access to China.
8 The line “Obviously, the provisions of the Pact do not apply to involvement through acts of self-defense” was taken out, but Iwakuro carefully explained to the Americans that day that this was merely to avoid adverse public reaction at home, and that the US right of self-defense was so universally recognized that this line seemed superfluous (FRUSJ, 2:468–70). Sensitive to Japanese concerns, the US side did not hold up negotiations over this line, suggesting instead that Japan state that the purpose of the Tripartite Pact was defensive and designed to prevent the “unprovoked extension” of the European war (ibid., 2:478, 471–83; FRUS [1941], 4:260–74). Japan formally accepted this language on June 21, the day before the German attack on Russia (FRUSJ, 2:488).
9 Indeed, the very language of diplomacy had changed. Both sides by mid-June would talk in terms of “the present understanding” and “now pending” agreement, with US concerns centered mainly on whether hard-liners such as Matsuoka might try to undermine the emerging deal. See Nomura’s diary, June 15 and 19, in PWP, 142–43.
10 On the expectations that Germany would likely win, see Sherwood 1950, 304–5.
11 Herring 1973, 10–11; Kimball 1991, 22–36; Sherwood 1950.
12 Because of the tension southward, not all these troops and materiel would arrive in Manchuria.
13 That US officials were taking the new situation seriously is shown by two actions on July 3. Chief of Naval Operations Harold Stark sent a message to subordinates warning that “[Japan’s] neutrality pact with Russia will be abrogated and a major military effort will be [made] against [Russia’s] maritime provinces,” probably at the end of July, “though [the] attack may be deferred until after collapse of European Russia.” He asked that principal army commanders and the British chiefs of staff be informed of this situation (PHA, 13–14:1396). Across town, Welles told the Soviet ambassador that he had information that Japan had decided to break its neutrality pact and attack Russia soon after (FRUS [1941], 1:787).
14 On the Japanese awareness that the German-Soviet war has led many in the United States to believe that Japan would now undertake “her long-cherished policy” of a northward advance, see Nomura’s diary, July 8, in PWP, 150.
15 See Ike’s summary notes, in JDW, 103–4.
16 For references and a summary, see Sagan 1989.
17 Roosevelt and Hull’s concerns were reinforced the next day by a navy report sent through the president’s Chief of Naval Operations Admiral Harold Stark. It recommended against an oil embargo, noting that “the shutting off [of] the American supply of petroleum will lead promptly to an invasion of the Netherlands East Indies.” FRUS (1941), 4:836–40.
18 For evidence that reinforces this argument, see Heinrichs 1988, 141–42, 246n68.
19 As this meeting was kept secret from State Department officials, there is nothing about it in the FRUS series.
20 Just prior to this line, Hull had sarcastically told Nomura that the stationing of large armies by both Japan and Russia along the Manchurian border must be the two countries’ way of preserving the Soviet-Japanese neutrality pact.
21 Even a careful historian like Barnhart (1987, 232–33) breezes quickly through the August–September negotiations.
22 Hull’s (1948) own recollections are a wonderful example of dissimulation on this issue. At every step of the way, his recounting of events is the exact opposite of what actually happened. He does admit that he and Roosevelt desired a peace deal in late August, and then changed their minds in early September. Yet he claims that his about-face was a result of Japan only wanting to adhere to vague statements of principles (ibid., 1024). In fact, it was Tokyo that had made potentially game-changing concessions. Hull (ibid.) states that while the negotiations in early September were proceeding, Konoe’s government “narrowed the concessions it had originally been willing to make.” It was Washington that had in actuality done just that. Hull (ibid.) asserts that the very holding of a summit “would cause China grave uneasiness” unless agreement had already been reached to protect China’s sovereignty, even though Tokyo had finally confirmed that all troops would leave China proper within two years. Overall, Hull (ibid., 1025) argues, he and Roosevelt could not make an agreement at the “expense of our principles and interests” as well as those of China, notwithstanding the fact that his diplomacy until that point had nothing to do with ideals and everything to do with realpolitik. Such consummate lying by a former secretary of state gives us prime facie evidence that he and Roosevelt were covering up something that had it become known, would have greatly damaged their reputations and that of the United States.
23 The quotation is a paraphrase of the original Halifax memorandum to London on August 18 (Heinrichs 1988, 161).
24 I found no Magic transcripts discussing the conference in the usual sources (MB and PHA), but that of course doesn’t mean they don’t exist.
25 Such concerns might explain Hull’s comment to Nomura on September 4 that he opposed Japan’s “station[ing] of troops for anti-Cominterm purposes” and “desired a complete evacuation” of such troops (Nomura’s diary, September 4, in PWP, 177–78). This was contrary to his position in early June, before Germany’s attack.
26 While this FRUS document has a September 6 date on it, it was the same one given to Grew three days earlier in Tokyo and immediately passed on the Hull.
27 On Hull’s constant berating of Nomura from early September to late October regarding the increasing presence of Japanese troops on the Soviet border, see footnote 35 below.
28 For such an argument and references, see Snyder 1991.
29 For the documents presented to Hirohito, see ibid., 152–63.
30 For references, see Sagan 1989; Taliaferro 2005.
31 On the proceedings of the five Liaison Conferences from September 11 to 25, see JDW, 167–78. For Nomura’s dispatches from early September to mid-October, see MB, 3:73–163, appendix.
32 See diary notes of Kido Koichi, lord keeper of the privy seal, in PWP, 126–28; Barnhart 1987, 253–54.
33 See especially Snyder 1991; Kupchan 1994.
34 It is important to note that in the November talks, the Japanese did not adopt the position that they were unwilling to leave China or would leave China only after twenty-five years, as is sometimes thought. The twenty-five-years figure was discussed at the November 5 Imperial Conference, but it referred only to the question of troops in northern China, not to those that would leave southern China. Moreover, this figure was to be used only for initial bargaining purposes, and only if Washington wanted to know how long Japan envisioned keeping troops in the north (JDW, 209–10). In short, by offering proposal A, the Japanese were making it perfectly clear they still wanted out of the China Incident with US help (their goal since April). The main US concern from September to November was not the Japanese troops that had occupied southern China after 1937 but rather those that might remain in the north from the pre-1937 period (see the following footnote). This subtle although critical distinction must be kept in mind to understand the US-Japanese talks of mid- to late November. In the end, Hull made “leaving China”—that is, all of China including the northern provinces and Manchuria—a key US demand. But he made this demand—one that was far more extreme than his position in early June—only after he and Roosevelt had decided against a modus vivendi on the morning of November 26.
35 Togo does not mention this, but from September 3–4 onward, US officials had continually berated Nomura about Japan’s deployment of troops against the Soviet Union. Nomura had made it clear to his superiors that this was Washington’s number one concern and the main stumbling block to any deal. See Nomura’s dispatches to Tokyo for September 4, 11, 17, 22, and 30 and October 9, 14, 16, and 27, in PWP, 177–96.
36 For references and a reworking of this argument, see Trachtenberg 2006.
37 The revisionist thesis has always had one underlying flaw: it assumes that it was obvious to the United States that Hitler would declare war on the United States should Japan attack the United States in the Pacific. By the terms of the Tripartite Pact, though, neither Germany nor Japan was under any obligation if its ally was the clear aggressor against a third party. In retrospect, it seems obvious that the Pacific War got the United States into the European war. But this could not have been anticipated with any degree of confidence prior to December 7, 1941.
38 See my discussion in the first two chapters; Fearon 1995.
39 These reports were generally sent via Brigadier General Sherman Miles, acting chief of staff of the G2, the army’s planning and information-gathering division.
40 Fear of German hegemony was reinforced by the internal release of the Joint Board’s Victory Program on September 11, 1941, which stressed that if Germany ended up controlling Eurasia, it could then prepare for the “the eventual conquest of South America and the military defeat of the United States” (Trachtenberg 2006, 118–19; see also Stoler 2000, chaps. 2 and 3).
41 See Nomura’s dispatches to Tokyo on November 20 and 22, in PWP, 205–7.
42 The military’s more relaxed attitude is shown by the notes of the Joint Board meeting held Wednesday at 11:35 a.m., presided over by Stark and Marshall. The discussion revolved solely around military preparations in the Pacific Rim (PHA, 15–16:1641–43).
43 According to Marshall, the language of this message came directly from Roosevelt himself (PHA, 1–3:1310).
44 For references and one of the better versions of this argument, supported with Chinese sources, see Sun 1993, chap. 7.
45 For the record of the Saturday briefing and Hull’s dismissive reaction to Chinese objections, see FRUS (1941), 4:635–40. See also Heinrichs 1988, 209–10.
46 The fact that Morgenthau obtained this information from his meeting with Soong on Thursday afternoon shows that the Chinese envoy still believed that the modus vivendi was on the table more than twenty-four hours after it had been kicked over. Roosevelt had not only misled him on Wednesday but also clearly still had no interest in telling him the truth even the next day, despite having already met with Japan’s diplomatic team. If Chiang’s worries had been critical to the Wednesday morning flip-flop, one might think the Chinese embassy would have been the first to be reassured.
47 Given the calculated fabrications contained in Hull’s memoirs regarding the April–September negotiations (see note 22), the very fact that he jumps on the Chinese and British letters as the cause of his flip-flop should raise red flags as to the explanation’s validity.
48 Indeed, they were upset when they found out that Hull had not presented the deal to Tokyo, since they knew they would be the front line of the war to come. See, in particular, the Australian documents on this period, detailing the efforts of Australian and British diplomats to get diplomacy back on track after November 26 (DAFP). For British and Australian efforts on November 30 to persuade Hull to buy time via diplomacy, see PHA, 19–20:3690–91.
49 See Heinrichs 1988, 213; Clark 1965, chap. 9; PHA, 19–20:4473–74. On the administration’s fear that Moscow would fall, see George Marshall’s testimony before the Pearl Harbor congressional hearings, in PHA, 1–3:1148–49.
50 Morgenthau’s diary that day also summarizes information that Soong had told him of a recent meeting between Hull, Soong, and British, Australian, and Dutch ambassadors. After the envoys were given details of the modus vivendi, Halifax mentioned he would have to consult with London. Hull “became very annoyed and said that it was up to the English to accept the proposal without any comments, which seems to me [Morgenthau] rather high-handed.” This is further proof that Hull and Roosevelt had little interest in allied opinion, and were determined to act according to their own strategic imperatives.
51 All the above quotations and summaries are taken from Morgenthau’s diary, FDRL, November 26, 1941. One source for Roosevelt’s information, aside from increasingly pessimistic reports in the New York Times (Heinrichs 1988, 213), was likely a military intelligence report written late on November 25. The report expressed new doubt about the city of Moscow’s ability to resist the German onslaught, noting that Russians officials were for the first time signaling their worry over the situation (PHA, 19–20:4473–74).
52 Stimson’s diary records his conversations with Roosevelt later that morning, after Hull had already informed Stimson that the deal would not be presented. But we have apparently no other documents revealing Roosevelt and Hull’s thinking prior to their decision to reject the modus vivendi and present Japan with ten demands that afternoon. Interestingly, Stimson himself was not told that the decision was final until he talked to Hull the next morning. Stimson immediately phoned the president, who led him to believe that Roosevelt and Hull had only presented the Japanese with a “magnificent statement.” He found out only later that this was not a “reopening” of US-Japanese talks but instead their essential termination. See Stimson’s diary notes for November 26 and 27, in PHA, 10–12:5434–35. It seems clear from this that Roosevelt and Hull were trying to make sure that no one, not even a close associate such as Stimson, could act to reverse their decision.
53 For this reason, even after severe setbacks in early 1942, US military strategy continued to focus on fighting hard enough in the Pacific such that Japan would not be able to attack Siberia. This effort knowingly diverted US resources from the European theater, but by keeping the Soviets in the war, it prevented the German domination of Eurasia long enough for the United States to complete its build up for total war. See Stoler 2000, 72–97.
54 There is no evidence, by the way, that Roosevelt knew about the impending attack on Pearl Harbor. But he certainly anticipated the strike on the Philippines.