From Total War to Korea — Roosevelt, Truman and Eisenhower
Firestorms over Japanese cities, the atomic bombs, the expansion of the Korean War, and the question of what the president actually authorized
Editorial illustration.
Firestorms over Japanese cities, the atomic bombs, the expansion of the Korean War, and the question of what the president actually authorized
When we discuss presidential responsibility for the use of military force, the years 1945–1953 are among the most difficult periods to compare fairly. Franklin D. Roosevelt died in the middle of a world war, when America's strategic bombing forces had already become a vast industrial machine. Within a matter of months, Harry S. Truman inherited that war, authorized the use of a new weapon, accepted Japan's surrender and, five years later, committed the United States to a new war in Korea. Dwight D. Eisenhower then inherited a conflict he had not started, but continued military pressure before the armistice and seriously considered the possibility of nuclear use in discussions at the presidential level.
If we reduced this history to the question “How many people did Roosevelt, Truman and Eisenhower kill?”, we would arrive at a number that concealed more than it explained. Tokyo is not Hiroshima, and a single air raid is not the entire Korean War. The number of people who died in a war is not automatically the number killed by American forces; nor does a state's responsibility for an operation necessarily prove that the president personally selected a particular target. This article therefore examines the political decision first, then its execution through the chain of command, and finally the categories and reliability of the figures for those who died.
In the research record, P1 denotes a documented presidential authorization to use or expand force; P2 denotes a broader political or campaign framework; and P3 denotes operational execution by military commanders. These levels may overlap, but they are not interchangeable. We also distinguish deaths directly resulting from U.S. attacks from the overall losses of a multi-actor war, indirect consequences, and historical estimates that cannot be independently verified.
Roosevelt's air war did not begin with Tokyo
By the time Roosevelt died on April 12, 1945, the United States had already spent years participating in an air war on a scale surpassing earlier conflicts. In Europe, American air forces emphasized daylight attacks on industrial, transport and military targets, while Britain's RAF conducted much of its night campaign as area bombing of cities. In practice, the line between a factory, a railway junction and the neighborhood where workers lived was often very thin. It would therefore be misleading to turn the total Allied deaths in Hamburg, Dresden or Germany as a whole into Roosevelt's personal American death toll.
In the Pacific, the final months of the war presented an even more direct problem. American B-29 bombers initially carried out mainly high-altitude attacks, but winds, cloud cover and poor accuracy reduced their effectiveness. General Curtis LeMay, who took command of XXI Bomber Command in January 1945, changed the tactics: lower-altitude night flights, heavier bomb loads and the mass use of incendiary weapons against densely built urban districts. The National Museum of the U.S. Air Force describes precisely this shift from attacks on individual industrial facilities to large-scale incendiary raids. [1]
That did not mean that there was no military production in these cities. Japan's wartime economy dispersed some manufacturing among smaller workshops and residential areas, physically intertwining industrial and civilian spaces. But the same urban pattern also made incendiary weapons capable of inflicting catastrophic damage on civilians. When the target is no longer just a particular industrial facility but an entire area, the category of consequences changes as well.
March 9–10, 1945: Tokyo
The most striking example came on the night of March 9–10, 1945. More than 300 B-29 bombers flew against Tokyo. The United States Strategic Bombing Survey summary records 1,667 tons of bombs dropped and approximately 15 square miles of densely populated urban area burned; the National Museum of the U.S. Air Force gives 334 bombers and almost 16 square miles destroyed. The difference between these rounded descriptions of area does not refer to two separate attacks. [2] [1]
Estimates of the dead are not entirely uniform. American historical sources frequently give a range of approximately 90,000–100,000 deaths, while more than a million people were left homeless. For our purposes, that range is more useful than choosing one deceptively exact figure. The attack was so lethal that the losses inflicted in a single night were comparable to early estimates for each of the atomic attacks considered separately. [3]
The question of responsibility requires a clear distinction. American responsibility for the attack is not in doubt: American forces carried it out. But the reviewed documentation does not provide equally direct evidence that Roosevelt personally ordered this specific raid or set its time, altitude and type of bombs. The tactical shift is documented in connection with LeMay and U.S. air force leadership. Roosevelt bore political responsibility for the state's strategy in a total war; holding the presidency alone, however, does not prove that he issued a specific order for Operation Meetinghouse.
We therefore classify this case as Roosevelt's broader wartime framework (P2) and American operational execution (P3), rather than a proven personal Roosevelt order for a specific target (P1). The deaths caused by the attack nevertheless remain deaths directly resulting from American force; the absence of proof that the president personally selected the target does not diminish them.
The campaign continued after Tokyo — but the president changed
The March raid on Tokyo was not an isolated event. Mass incendiary attacks followed against Nagoya, Osaka, Kobe and other cities. The United States Strategic Bombing Survey describes the rapid expansion of attacks on urban areas and reports that many Japanese cities were subjected to systematic bombing before the campaign ended. [2]
Here the presidential boundary becomes chronologically important. Roosevelt died on April 12. But the B-29s, the Mariana Islands bases, LeMay, targeting systems and military doctrine did not disappear. Truman did not begin the incendiary campaign, but he inherited it, and it continued during his presidency. To attribute all of Japan's incendiary bombing to Roosevelt alone would ignore the change of president; to attribute it all to Truman would erase the fact that the decisive tactical shift had already occurred before he took office.
The example shows why presidential terms do not divide military campaigns into independent units. Military institutions, aircraft, commands and tactics do not stop when the president changes. A new president inherits a war machine already in motion while also bearing responsibility for new decisions taken within it.
Truman becomes president and learns about the Manhattan Project
Harry Truman assumed the presidency on April 12, 1945, without the usual transition period. The war in Europe had not yet formally ended, and the war against Japan continued. As vice president, he had only limited knowledge of the secret Manhattan Project; after taking office, he was briefed more fully on the development of atomic weapons.
On July 16, 1945, the Trinity test demonstrated that a nuclear device worked. Nine days later, the military chain of command had prepared an execution directive for the first atomic bombing after approximately August 3, weather permitting. Hiroshima, Kokura, Niigata and Nagasaki were among the targets; the directive also envisaged using additional bombs as they became available. Declassified documents show the process involving General Leslie Groves, General George Marshall, the War Department and General Carl Spaatz. [4]
The distinction between political authorization and the form of the surviving documents matters. The Truman Presidential Library points out that no surviving written order is known in which Truman explicitly and directly ordered the atomic bombing of Japan. Yet his political decisions, written notes, military execution orders and public statements do survive. The absence of one signed order is therefore not evidence that presidential authorization was absent. [5]
In the decision record, the use of atomic weapons is accordingly P1 — authorization at the presidential level. The choice of the operational moment and the execution of the missions belong to the military chain of command (P3).
Hiroshima: one bomb, but not one category of victim
On August 6, 1945, the B-29 Enola Gay dropped the uranium bomb Little Boy over Hiroshima. The City of Hiroshima now estimates that approximately 350,000 people were present in the city. They included not only ordinary civilian residents but also military personnel, mobilized workers, people from surrounding areas, Koreans and Taiwanese from Japan's then-colonies, other foreigners and some American prisoners of war. Approximately 140,000 people are estimated to have died by the end of December 1945, and the city explicitly stresses that an exact number cannot be determined. [6]
We therefore do not label the 140,000 in this series as “140,000 civilians.” It is an estimate of all deaths resulting from the bombing through the end of 1945 among a mixed population. Nor is it a count of those who died at the moment of the explosion. Some victims died later from burns, injuries and acute radiation effects.
That same day, Truman publicly accepted responsibility for the decision to use the new weapon and warned of further destruction. His public statement is important evidence that the president owned the decision, but it must be read as a wartime statement by his administration, not as an independent historical assessment of what was necessary or what alternatives existed. [7]
Nagasaki: the second bomb within an existing execution framework
On August 9, the B-29 Bockscar dropped the plutonium bomb Fat Man over Nagasaki. The city's official website gives 73,884 dead and 74,909 injured in its estimates through the end of December 1945. Here, too, the municipal authorities emphasize that these are historical estimates and that a complete individual register of all deaths cannot be compiled. [8]
For presidential attribution, it is crucial that Nagasaki did not require a wholly new presidential decision-making process on August 9. The July military directive had already provided for additional bombs to be used as they became ready. The second attack thus took place within an already authorized framework. We should neither describe it as an accidental operational decision with no presidential connection nor claim that Truman is documented as having personally selected Nagasaki again on August 9.
Combining the two cities' official historical estimates gives approximately 214,000 deaths by the end of 1945. This arithmetic is useful only with the correct label: it does not mean 214,000 immediate deaths or 214,000 confirmed civilian deaths. It combines two municipal historical estimates covering different categories of people and deaths occurring weeks or months after the explosions.
Did the atomic bombs alone end the war?
This part of the story remains politically and historiographically contested; a single slogan cannot resolve it. Hiroshima was bombed on August 6. The Soviet Union declared war on Japan on August 8; Soviet forces entered Manchuria on August 9, the same day that Nagasaki was bombed. A crisis followed within the Japanese leadership, and it was neither simple nor consensual. Emperor Hirohito intervened to break the deadlock; an attempted military coup followed, and the formal surrender was signed on September 2. [9] [10]
It is reasonable to conclude that the atomic bombs had a powerful effect on the process leading to surrender. It is not fair, however, to erase Soviet entry into the war, conventional destruction, the naval blockade, the internal political crisis or the different responses of Japanese decision-makers to the same events. Historians still disagree about which factor was decisive and how the war might have ended under a different scenario.
For that reason, this series accepts neither “the bombs were the only reason for surrender” nor the opposite claim that “the bombs played no significant role.” Both are more categorical than the documents support.
After Nagasaki, the rule changes
On August 10, after both bombings and as Japan moved toward surrender, General Leslie Groves reported on preparations for the next weapon. General George Marshall wrote on his memorandum that another atomic bomb must not be used against Japan without the president's express authorization. This appears in document no. 82 in the collection published by the National Security Archive. The note helps explain how control over any further use changed after Nagasaki. [4]
That does not settle the moral debate, but it does allow us to describe decision-making more precisely. Hiroshima and Nagasaki fell within an already authorized execution program. A further attack would have required a new step by the president.
Truman's responsibility: what is documented and what is interpretation
In Truman's case, a strong documentary conclusion is possible. The president accepted or confirmed the political decision to use atomic weapons against Japan and repeatedly defended it afterward as a way to end the war quickly and avert greater losses in an invasion. That is the documented position of Truman and his administration.
A different claim — that atomic use was the only way the war could have ended — is not a documented fact of the same kind. Even in 1945, different proposals and concerns existed within the American system, ranging from a demonstration of the weapon to changes in surrender terms. This does not prove that any alternative would have worked. It shows that the decision was not made in the absence of arguments for alternatives.
Presidential responsibility therefore needs no mythology. We need not claim that Truman personally chose every technical detail of a mission to recognize that the decision to use atomic bombs was made at the highest presidential level.
Five years later: Korea
On June 25, 1950, North Korean forces invaded the Republic of Korea across the 38th parallel. United Nations documents and American archives clearly identify the North Korean invasion as the beginning of open war. The Security Council condemned the invasion and called on member states to assist South Korea. On June 27, Truman announced that he had ordered American air and naval forces to provide support to the South Korean forces. [11] [12]
The distinction between the beginning of the war and American entry into it is decisive. Truman did not order North Korea's invasion. He did, however, authorize American military intervention. The precise chronology also matters when examining the expansion of air operations: the Joint Chiefs of Staff (JCS) directive of June 29, 1950 already authorized MacArthur to attack specifically defined military targets in North Korea, subject to conditions and restrictions near the Chinese and Soviet borders. A memorandum dated June 30 confirms presidential approval of air attacks north of the 38th parallel; the expansion was also publicly announced that day. These dates therefore refer to distinct documented stages of decision-making and communication, not to a first directive issued only on June 30. [13] [14] [15]
In our record, American entry into the conflict is a documented presidential decision (P1). The expansion of military action against targets in the north also has a documented presidential connection, although restrictions, specific targets and execution were determined through the command structure. This is not evidence that Truman personally selected every bombing target.
Korea's legal basis was not a formal declaration of war
Truman did not ask Congress for a formal declaration of war. The administration presented American involvement in the context of UN Security Council action and presidential executive and commander-in-chief authority. The National Archives explicitly notes that Truman sent American forces without a formal congressional declaration of war and that the conflict was often officially described as a “police action.” [16]
This was an important constitutional turning point, but the article does not purport to settle the decades-long legal debate over the scope of presidential war powers. For our purposes, a more precise statement suffices: American military intervention was authorized by the president within the UN framework, without a formal declaration of war by Congress. Subsequent funding and congressional involvement further complicated the political and legal picture, but did not change the fact that the initial order came from the executive branch.
From defending South Korea to war north of the 38th parallel
The initial public objective of the American intervention was to repel the North Korean attack and restore the situation before the invasion. Following the landing at Inchon in September 1950 and the collapse of much of the North Korean force, a second question arose: should UN forces remain south of the 38th parallel, or carry the war northward?
An archival analysis by the National Archives shows that advancing north was not simply an arbitrary move by General Douglas MacArthur. The Truman administration had already discussed the possibility of forcible reunification during the summer, and Truman authorized a plan for operations north of the 38th parallel, subject to restrictions intended to reduce the risk of Chinese or Soviet intervention. [17]
This means that the presidential objective changed. An intervention that began as assistance in defending South Korea became a campaign to destroy North Korean forces north of the prewar border and potentially bring about political reunification of the peninsula under South Korean rule. This was a new presidential decision, not merely a continuation of the initial order.
Chinese intervention: causation without oversimplification
As UN forces approached the Yalu River, China sent substantial combat forces into Korea. China's decision was its own military decision, and Chinese attacks cannot therefore be attributed to Truman as DIRECT-US. The historical context nevertheless matters. Chinese leaders had previously warned of the consequences of advancing northward, and the American administration itself recognized the risk of Chinese intervention.
The causal formulation is therefore twofold: China independently decided to send troops into Korea; the UN advance north of the 38th parallel and toward China's border was an important immediate context for that decision. Neither observation needs to be erased in order to preserve the other.
When Chinese intervention reversed the situation on the battlefield, the Truman administration returned to a more limited objective. The dispute with MacArthur over expanding the war into China became so intense that Truman dismissed the general in April 1951. The National Archives connects the decision to MacArthur's public demands for a wider war and his insubordination toward civilian leadership. [16]
This is another important aspect of presidential responsibility: a president decides not only when to expand the use of force. A president can also refuse further escalation.
The air war over Korea
American air forces and other UN forces attacked roads, warehouses, industry, military positions and, later, other infrastructure in Korea. Figures for the tonnage dropped must be read together with their sources. In a comparison with Vietnam, the Naval History and Heritage Command gives approximately 653,000 tons of conventional bombs for the Korean War as a whole. Historian John W. Dower, citing Bruce Cumings, gives 635,000 tons of bombs plus an additional 32,557 tons of napalm; in the same note, he explicitly cautions that estimates vary between sources. Because these figures are not defined identically and were not derived from a single shared table examined here, we neither combine them nor present them as one range of statistical certainty. [18] [19]
For USAF units, Peter Grier's historical review The Remembered War gives 720,980 combat sorties and 476,000 tons of ordnance dropped. These figures concern a particular branch of the armed forces and are not directly comparable with statistics covering all coalition forces or using different definitions of bombs and ordnance. The quantity of weapons dropped does not, by itself, tell us how many people died. [20]
The published overviews differ in the kinds of weapons, the military units covered and the accounting methods used. The difference between 476,000, 635,000 and 653,000 tons is therefore not a basis for calculating three separate estimates of deaths. We must first understand what was counted.
How many people died in the bombing of North Korea?
Documentation about the fate of people in Korea is far less complete than records of sorties and bomb loads. In a brief report published in 2001, the Wilson Center summarizes Soviet reports that relayed North Korean figures, including 282,000 people killed in bombing attacks. The same publication notes the lack of a reliable statistical census, missing people, refugees and widely varying estimates of overall losses. The figure is presented in the context of reports on civilian losses, but it remains a historical estimate attributed to its sources: it is neither an independently verified register of individual deaths nor proof of how losses should be allocated among the forces carrying out attacks. [21]
We do not add this estimate to other Korean War losses, since their categories may overlap. Nor do we divide it between Truman and Eisenhower without evidence supporting such a temporal and causal allocation.
The total death toll of the Korean War cannot be attributed to Truman or Eisenhower as the direct result of American attacks. It was a multi-actor war: people died in ground combat, American and other air attacks, repression, massacres, disease and deprivation. We must distinguish actions documented as having been carried out by American forces from cases in which the evidence does not allow a confident conclusion.
The atomic bomb returns to presidential discussion
On November 30, 1950, after China entered the war, Truman said at a press conference that all weapons were under consideration. The exchange caused such concern that the White House clarified the same day that only the president could authorize use of the atomic bomb and that no such authorization had been issued. [22]
Nuclear weapons were not used in Korea. The imagined effects of a nuclear attack that did not occur therefore do not belong in the record of actual deaths. Discussion of possible use nevertheless matters for understanding the options and constraints facing the president.
Eisenhower inherits a stalemated war in January 1953
Dwight D. Eisenhower did not inherit peace. When he became president on January 20, 1953, the war had already fluctuated for more than two years around roughly the same front, while armistice negotiations dragged on, chiefly over prisoners of war and political conditions.
As a candidate, Eisenhower had promised to go to Korea. After his election, he treated the conflict as a problem that could not be resolved simply by continuing small local attacks. The Eisenhower Presidential Library describes his policy as a combination of diplomacy and military pressure. [23]
Nuclear weapons also became a subject of serious presidential discussion under Eisenhower. FRUS records a meeting at which he considered the possibility of using atomic weapons to achieve a military effect in Korea. Other documents show that the administration wanted the adversary to believe that broader escalation was possible. Nevertheless, nuclear weapons were not used. [24]
Caution is needed with claims that “the nuclear threat ended the Korean War.” Eisenhower's administration later emphasized strength and willingness to escalate, while other historical sources also point to Stalin's death in March 1953, changes in Soviet and Chinese policy, exhaustion and negotiations that had already been under way for a long time. No single available document justifies the conclusion that one factor alone caused the armistice.
Dams: when water and crops also become targets
American air pressure continued in the spring of 1953. On May 13, American aircraft attacked the Toksan dam north of Pyongyang. U.S. Air Force history describes damaged bridges and roads, more than a thousand flooded buildings, an airfield at Sunan rendered unusable and flooded rice fields. Roughly three days later, an attack followed against another irrigation dam. Different official historical sources spell its name Chasan (in the Pacific Air Forces history) and Chosan (in the history of the 58th Wing). On the basis of these sources, we do not assume they were two separate events. [25] [26]
In a dispatch dated May 14, 1953, UN forces commander Mark Clark included approximately twenty irrigation dams among proposed future attacks, not all of which had yet taken place. The document estimated that breaching them would flood some 422,000 acres of land (approximately 171,000 hectares) and damage or destroy a quarter of a million tons of rice. These were projected or anticipated effects, not confirmed results of attacks against twenty dams. [27]
We cannot use these projections to calculate how many people subsequently died from lack of food. But the document shows that the consequences for irrigated fields and crops were not overlooked as accidental damage: military commanders included them in their thinking about pressure on the adversary. It is therefore legitimate to ask the moral question of what it means deliberately to direct warfare against a system supplying food, even though the available figures do not justify a new death toll.
Presidential attribution must remain at the level supported by evidence. The attacks occurred during Eisenhower's term and within the broader framework of military pressure (P2), while their planning and execution are chiefly documented at the military command level (P3). The documents examined do not establish a personal Eisenhower order for either particular dam as P1.
July 27, 1953: armistice, not a peace treaty
The armistice was signed at Panmunjom on July 27, 1953. The historical FRUS record states that attacks continued for several hours after the signing and that the agreement came into effect at 10 p.m. local time. It was an armistice, not a peace treaty. The documents do not support the simplification that fighting stopped the moment the agreement was signed. [28] [29]
Eisenhower inherited the war and achieved an armistice during the first months of his presidency, but he did not bring it about single-handedly. Negotiations, military conditions, the change of leadership after Stalin's death, decisions in Beijing and Pyongyang, Seoul's position and American pressure all contributed to the outcome. To claim that a single presidential measure alone ended the war would go beyond the available evidence.
Three presidencies, different levels of responsibility
Roosevelt led the country in a total war during which American forces carried out devastating incendiary attacks on Japanese cities. Tokyo was a direct American attack with a high death toll. The documents examined, however, do not prove that Roosevelt personally ordered a particular operational target. That distinguishes his broader political responsibility from an operational decision.
Truman authorized the use of atomic weapons against Japan at the presidential level as well as American military entry into Korea; the expansion of operations against specified targets in the north is also documented. The distinction between initial authorization and subsequent execution remains important, as does the difference between deaths directly caused by American force and the wider consequences of a multi-actor war.
Eisenhower inherited the war. His administration combined diplomacy, continued bombing and consideration of nuclear escalation that was never carried out. Dam attacks are documented as actions by American forces, not as proof that he personally selected each target. We do not need to deny his contribution to the armistice in order to examine the human consequences of the pressure applied before it.
What the figures tell us — and what they do not
For the Tokyo attack on the night of March 9–10, 1945, we use approximately 90,000–100,000 deaths as a historical estimate for one American operation. For Hiroshima, we use the city's estimate of approximately 140,000 deaths by the end of December 1945, and for Nagasaki 73,884 deaths within the same time boundary, again according to the city's historical estimate. Combining the two city estimates gives an approximate arithmetic total of 214,000, but not a census of immediate deaths or of civilians alone.
For the bombing of North Korea, the figure of 282,000 remains an estimate attributed to reports cited by the Wilson Center. It does not prove that all deaths occurred under identical circumstances, that they can be divided among individual presidents or that every person who died was individually recorded. Bomb tonnage is not a death toll. Figures for the overall Korean War losses are not the same as figures for deaths directly caused by American attacks; we do not automatically attribute losses caused by other belligerents to the United States. Nuclear operations that were considered but never carried out are not a category of actual victims.
Conclusion: between an order and a human death
It is difficult to look at the history of aerial warfare through a city on a map without also seeing the people who lived there. Yet concern for their fate must be accompanied by precision, not substituted for evidence. In Tokyo, we see how a change in operational tactics could kill tens of thousands in a single night; in the atomic attacks, the political authorization of a new weapon and many deaths that followed later; in Korea, a sequence of decisions that cannot be reduced to one president's name or one total figure.
Responsibility can be understood only if we do not confuse the president's decision, operational execution, and the category and reliability of the data on those who died. This distinction does not excuse any loss of life. It does, however, permit a more serious judgment: who had the power to decide, who actually decided, and how much we still do not know about the consequences.
Next article: Vietnam, Laos and Cambodia — Kennedy, Johnson and Nixon. It will extend the question to covert operations, presidential restrictions on target selection, and bombing in countries that were not formally the main theaters of the war.
Sources and further reading
- National Museum of the U.S. Air Force, “Fire Bomb Raids” (museum historical overview). Source
- United States Strategic Bombing Survey, Summary Report (Pacific War), Air University Press reprint, 1987, section on the attacks beginning March 9, 1945. Source
- Jeremy Prichard, “Part 4 — A Case Study in Modernization: Challenge #3 Operational Effectiveness,” Kirtland AFB, 2022. Source
- William Burr (ed.), The Atomic Bomb and the End of World War II: A Collection of Primary Sources, National Security Archive, 2020, especially the July directives and document 82. Source
- Harry S. Truman Presidential Library, “Decision to Drop the Bomb” (museum overview and account of the documentary chain). Source
- City of Hiroshima, “How many people died because of the atomic bombing?” (official estimate through December 1945). Source
- Harry S. Truman, “Statement by the President Announcing the Use of the A-Bomb at Hiroshima,” August 6, 1945. Source
- City of Nagasaki, “Nagasaki, the Atomic Wasteland” (official figures from a 1950 study). Source
- The National WWII Museum, “The End of World War II 1945.” Source
- The National WWII Museum, “To Bear the Unbearable: Japan’s Surrender, Part II,” 2020. Source
- U.S. National Archives, “US Enters the Korean Conflict” (educational dossier and primary documents). Source
- FRUS, 1950, Korea, vol. VII, document 119: Truman statement, June 27, 1950. Source
- FRUS, 1950, Korea, vol. VII, document 161: JCS to MacArthur, June 29, 1950, 6:59 p.m. Source
- FRUS, 1950, Korea, vol. VII, document 175: memorandum of conversation, June 30, 1950. Source
- FRUS, 1950, Korea, vol. VII, document 174: editorial note on public announcement, June 30, 1950. Source
- National Archives, “National Archives to Display Truman Document Commemorating U.S. Participation in Korean War,” press release, May 3, 2000. Source
- James I. Matray, “Revisiting Korea: Exposing Myths of the Forgotten War,” Prologue, Summer 2002, National Archives. Source
- Naval History and Heritage Command, “H-017-2: Operation Rolling Thunder Overview” (comparative tonnage statement). Source
- John W. Dower, “The San Francisco System: Past, Present, Future in U.S.–Japan–China Relations,” Asia-Pacific Journal, 2014, note 13 (summary of Bruce Cumings’s figures). Source
- Peter Grier, “The Remembered War,” Air & Space Forces Magazine, July 1, 2003. Source
- Wilson Center / Cold War International History Project, “New Evidence on North Korean War Losses,” August 1, 2001. Source
- Harry S. Truman, “The President’s News Conference,” November 30, 1950. Source
- Dwight D. Eisenhower Presidential Library, “Korean War” (collection and short historical overview). Source
- FRUS, 1952–1954, National Security Affairs, vol. II/1, document 53: NSC discussion, March 31, 1953. Source
- U.S. Air Force, “58th SOW History” (history of the 58th Wing and dam attacks). Source
- William T. Y’Blood, “The Korean Air War,” Pacific Air Forces (discussion of Chasan and Toksan). Source
- FRUS, 1952–1954, Korea, vol. XV/1, document 519: Clark telegram to JCS, May 14, 1953. Source
- FRUS, 1952–1954, Korea, vol. XV/2, document 723: editorial note on the armistice, July 27, 1953. Source
- Dwight D. Eisenhower Presidential Library, “Presidential Years” (overview of the presidency). Source