.png)
Cold Files 006: The City That Was Never Counted. Manila, 1945
The Sentence
Everything written about the Battle of Manila rests on a single sentence, and that sentence has no footnote.
Triumph in the Philippines is the US Army's official history of the campaign, written by Robert Ross Smith and published by the Office of the Chief of Military History in 1963. It runs past seven hundred pages and footnotes almost everything: battalion strengths, river crossings timed to the hour, rounds fired by battery. On how many Filipino civilians died while the Americans retook their capital, it offers one clause. An estimated 100,000 Filipino civilians had lost their lives during the battle. Nothing underneath it.
Richard B. Meixsel, a historian at James Madison University, went into the manuscript files to find where the number came from. Smith's draft had said 5,000. Someone changed it to 100,000. Major General Robert S. Beightler, who commanded the 37th Infantry Division through the battle, caught it in review and objected in writing that the figure should be less than 10,000. Smith did not yield. His answer, in July 1957, was one sentence: my figure is that provided by the Philippine government. No document supporting it has been found.
That is the entire documentary basis. Meixsel's conclusion is flat: every published claim that one hundred thousand civilians died at Manila traces back to Triumph in the Philippines, and there are no exceptions. His essay is not peer reviewed, which is worth saying plainly; what carries it is the archive it cites. He also notes who did not use the number. Hartendorp, Agoncillo, Friend and Steinberg all had direct access to Manila after the battle, and gave figures from thousands to tens of thousands.
This file is not an argument that fewer people died. It is an argument about something worse. In the official record the American dead are known to the man: 1,010 killed, 5,565 wounded. The Japanese dead were counted at 16,665. The Filipinos who lived in the city were estimated once, in a clause, and repeated for sixty years.
The Open City
Manila in 1944 held roughly 1.1 million people in a metropolitan area of about fourteen and a half square miles. It had already been through this once.
On 26 December 1941, Douglas MacArthur declared Manila an open city, without the characteristics of a military objective. Banners reading OPEN CITY and NO SHOOTING were hung at the city hall, and the declaration went out by radio and newspaper.
The Japanese bombed it the next day. Five waves came in around midday on 27 December, killing about forty civilians. A bomb hit Fort Santiago, the Intendencia was heavily damaged, and the Church of Santo Domingo burned with roughly twenty thousand volumes inside it.
In February 1945 nobody declared Manila anything. Not the Japanese, who held it. Not the Americans, who came to take it. A million people stayed where they were because no one had arranged anywhere else for them to go.
Two Armies, One City
General Tomoyuki Yamashita did not want to fight in Manila and said why. The official history quotes him: the population is approximately one million, therefore it is impossible to feed them, and the buildings are very inflammable. His order to General Shizuo Yokoyama was to destroy the bridges and withdraw as the Americans arrived.
What he would not do was make it formal. The Filipino historian Benito Legarda Jr., himself a survivor, records that Yamashita denied the pleas of Filipino officials to declare Manila an open city, as MacArthur had in 1941. He wanted his army out. That is not the same as wanting the city spared, and the distinction would matter at his trial.
Rear Admiral Sanji Iwabuchi decided otherwise. He commanded the Manila Naval Defense Force, nearly 17,000 men, about 12,500 of them naval personnel and 4,500 army. He had captained the battleship Kirishima until it was sunk off Guadalcanal in November 1942, a loss his biographers return to when they explain why he chose to die in place.
On paper Iwabuchi answered to Yokoyama's Shimbu Group. Smith records that the control authority actually transferred was so limited as to contain the seeds of many disagreements, and treats what followed not as insubordination but as compromises between commanders who did not control one another. Connaughton, Pimlott and Anderson put it harder: Iwabuchi ignored repeated orders to withdraw, and the naval staff in Tokyo backed him. Either way, seventeen thousand men fortified a city of a million and prepared to be killed inside it.
The Order
The International Military Tribunal for the Far East quoted a captured document in its judgment on conventional war crimes, and named its provenance: the file of orders issued by the Manila Navy Defence Force between December 1944 and February 1945.
Be careful to make no mistake in the time of exploding and burning when the enemy invades. When killing Filipinos, assemble them together in one place as far as possible thereby saving ammunition and labor.
What followed is documented unevenly, and the numbers do not agree. The Tribunal described the German Club, where Japanese troops surrounded the building with a barricade of inflammable material, poured gasoline over it and set it alight. It described St Paul's College, where roughly 250 people were confined with explosives concealed in the chandeliers, and machine guns and grenades turned on the building after the charges detonated. Other accounts put the St Paul's dead at 360, or above 600. De La Salle College on 12 February, where sixteen Christian Brothers died with the families sheltering there, carries another set again.
American war crimes investigators working in 1945 gathered thousands of pages of sworn testimony and documented 27 major atrocities in Manila, out of roughly 300 across the Philippines. They recorded that the list was by no means complete. Nobody ever completed it.
The Guns
MacArthur refused to bomb the city from the air. The official history records only that Allied air forces flew very few strikes inside the city limits before he forbade them, and the Combat Studies Institute that he repeatedly denied requests from subordinate units, to spare Filipino civilians. The wordings usually quoted from him here cannot be traced to a primary source, so they are not quoted.
The air ban held. Artillery was another matter. In the second week of February the restrictions came off, and the official history explains the logic without embarrassment: if the city were to be secured without the destruction of the 37th and the 1st Cavalry Divisions, no further effort could be made to save the buildings, and everything holding up progress would be pounded.
The Philippine General Hospital shows what that meant. The Japanese had fortified it with Filipino patients inside. The 140th Field Artillery fired 2,091 rounds of 105mm high explosive into it, and the 82nd Chemical Mortar Battalion added 1,101 rounds of high explosive and 264 of white phosphorus. More than 2,000 civilians came out afterward, and the 148th Infantry led another 5,000 to safety that night. Both halves of that are true at once and neither cancels the other.
You will often read that American shelling caused forty percent of civilian deaths. That chain runs to two places. One is a 2019 National WWII Museum essay by Dan Olmsted, which does not contain the figure at all. The other is an unpublished conference paper. Nobody has produced a method. What exists instead are three apportionments collected by Legarda, all pointing the other way: General Ramon Farolan at 60 percent of the killing done by the Japanese, the Memorare Manila 1945 Foundation's survivor questionnaire at 70, and Legarda's own count at 85, from a list of about 250 Spanish nationals. Only the last is reproducible, and it is 250 people in a city of a million.
Intramuros
The American preparation ran from 17 to 23 February. The final barrage opened at 0730 on 23 February and lifted at 0830. Eight inch and 240mm howitzers, 155s and 105s, tank destroyers firing point blank, chemical mortars. The Combat Studies Institute counts 7,487 high explosive rounds into Intramuros in that hour. The infantry went in at 0830, the 129th crossing the Pasig in engineer boats, the 145th against the east wall.
That same day the Japanese released about 3,000 of the civilians they had been holding inside the walls. Almost all of them were women and children. They had murdered most of the men in the group first.
Some three thousand civilian men had been herded into Fort Santiago, on the testimony of Antonio Gisbert, one of the fifty or so who came out. American troops found the dead where they had been left. Fighting there ended at noon on 25 February.
The last stand was in the reinforced concrete government buildings south of the walls, which infantry could not enter. Howitzers and tank destroyers were run up to the facades and fired into them point blank. Iwabuchi is believed to have killed himself around 26 February, most likely in the Agriculture Building. The Finance Building fell on 3 March and the battle was declared over.
What Was Left
The business district was gone entirely. Eighty percent of the southern residential district, seventy five percent of the factories and seventy percent of the utilities were destroyed, and James Scott counts more than 11,000 buildings levelled. Manila itself was never separately costed.
The Trial
Yamashita was charged on 25 September 1945 and tried in Manila from 29 October to 7 December before a US military commission of five officers, none of them lawyers, appointed by General Wilhelm Styer under regulations MacArthur had issued. The 123 specifications covered 62,278 civilians tortured or murdered, 144 American personnel killed and 488 women raped.
The Supreme Court declined to intervene on 4 February 1946, six votes to two. Justice Frank Murphy's dissent has outlived the verdict: the recorded annals of warfare and the established principles of international law afford not the slightest precedent for such a charge. The immutable rights of the individual, he wrote, belong not alone to the members of those nations that excel on the battlefield. Rutledge dissented on the evidence, which included hearsay and unsworn depositions against a defence given three weeks to prepare.
Yamashita was hanged on 23 February 1946, one year to the day after the guns opened on Intramuros.
The case created the doctrine that carries his name. A commander can be criminally liable for what his troops do, on a failure to control them, without proof that he ordered it or knew of it. It was later codified in Additional Protocol I and the statutes of the modern tribunals. Allan A. Ryan, a former US war crimes prosecutor, titled his study Yamashita's Ghost: the general, he argues, was executed for what his soldiers did without his approval. Set against that: the commission sat under MacArthur's regulations and his delegated authority, and MacArthur confirmed the sentence, over the destruction of a city he had commanded the recapture of.
No American faced charges over the shelling of Manila. The question was never formally put.
The Forgetting
John Lee Candelaria argues the forgetting was structural rather than accidental. Postwar Philippine governments commemorated Bataan and Corregidor, stories of Filipino and American soldiers fighting side by side, useful in the Cold War. A month in which enormous numbers of civilians died under two armies fits no such story. Writing in 2024, he noted there is still no state sanctioned memorial site dedicated to the Battle of Manila.
Gregory Kupsky records that Manilenos resist the word liberation. It has appeared, he writes, only intermittently in official ceremonies, usually giving way to The Battle for Manila.
The monument that does exist was built by the bereaved. Memorare Manila 1945 was unveiled on 18 February 1995, the fiftieth anniversary, in the Plazuela de Santa Isabel inside Intramuros. Peter de Guzman sculpted it: the victims gathered around a cloaked, faceless mother.
The same foundation ran a survivors' questionnaire and has worked through the war crimes trial records to compile the dead by name. The list is still not finished.
That is the whole file in one comparison. One thousand and ten American dead, known individually. One hundred thousand Filipino dead, asserted once without a source and repeated ever since. And a list of names, still unfinished, kept by volunteers who buried them.
Nicholas Evan Sarantakes of the US Naval War College published a study of the battle in 2025 and called it a poisoned victory. His publisher bills it the third bloodiest battle of the Second World War, behind Leningrad and Berlin, a ranking that depends on accepting the number nobody has verified. Which is the point. Manila is the largest atrocity of the Pacific war the record cannot describe, because the institutions that could have counted did not think it needed doing.
NEXT FILE: Kenya, 1952 to 1960. A million people moved into fortified villages, an official death toll that collapsed under scrutiny, and the colonial files flown out of Nairobi at independence.
Sources consulted
Robert Ross Smith, Triumph in the Philippines, OCMH, 1963
Richard B. Meixsel, Did 100,000 Civilians Die in the Battle of Manila?
Thomas M. Huber, The Battle of Manila, Combat Studies Institute
International Military Tribunal for the Far East, war crimes judgment
In re Yamashita, 327 U.S. 1, 1946, and the Murphy dissent
Benito Legarda Jr. and Ricardo Trota Jose, battleofmanila.org
John Lee Candelaria, Forgetting the Battle of Manila, 2024
Gregory Kupsky, Origins, Ohio State University
Dan Olmsted and James M. Scott, National WWII Museum
Nicholas Evan Sarantakes, The Battle of Manila, 2025
Allan A. Ryan, Yamashita's Ghost, 2012
Connaughton, Pimlott and Anderson, The Battle for Manila, 1995
VERA Files and the Inquirer on the Memorare Foundation
Cold Files runs twice a week: deep reads on the wars nobody covers, sourced to the document. For the daily brief and the archive, follow Project Leaflet on Telegram at t.me/TheLeaflet.
Comments