The Japanese Occupation and Pacific War (in numbers)

Numbers quoted by L. de Jong for the Japanese occupation of the Dutch East Indies:

• Prisoners of war
  • Numbers: 42,233
  • Deaths: 8,200
  • % deaths: one in five
• Internees
  • Numbers: 100.000
  • Deaths: 16,800
  • % deaths: one in six
• Outside of the camps
  • Numbers: 220.000
  • Deaths: ?
  • % deaths: ?
• Javanese
  • Numbers: 50 million
  • Deaths: 2.5 million
  • % deaths: one in twenty

Number of Dutch prisoners of war:

In 1978, using the data preserved in the archives of the Royal Dutch Navy and the Royal Dutch East Indies Army (KNIL), Major drs. H.L. Zwitzer came to the conclusion that a total of 42,233 soldiers in the Dutch East Indies had been taken captive:

Royal Dutch Navy: 3,847 soldiers

KNIL: 36,869 soldiers

KNIL auxiliary troops: 1,517 soldiers

8,200 (19.4%) of the total number of 42,233 European prisoners of war died:

Royal Dutch Navy: 648 deaths (16.8%)

KNIL and auxiliary troops: 7.552 deaths (19.6%)

The average death rate of POWs of all Allied nationalities in the Pacific War was 27%. The American mortality rate was 34%, the Australian 33%, and the British 32%. The Dutch mortality rate was below 20%. (Gavan Daws, Gevangenen van de Japanners; Krijgsgevangenen in de Pacific gedurende de Tweede Wereldoorlog (‘Prisoners of the Japanese; Prisoners of War in the Pacific during World War II’, Baarn 1996), p. 409.)

According to L. de Jong (Het Koninkrijk der Nederlanden in de Tweede Wereldoorlog (‘The Kingdom of the Netherlands During the Second World War’), vol. 11b, pp. 710-714), 5,500 Allied prisoners of war were put to work on the Pakan Baru railway, 700 of whom died (one in eight). In addition, some 22,000 so-called romushas (Asian forced labourers) were deployed there, about 17,000 of whom died (four out of five). One in four to five of the approximately 2,000 prisoners of war who were taken to the Palembang region for building airfields did not survive.

Number of Dutch civilian internees:

Data on the numbers of interned civilians in the Dutch East Indies are very incomplete, because almost all Japanese records concerning the internment have been lost, as well as many camp archives. We can therefore only estimate the total number of internees. In the course of the war, the Japanese reported to the International Red Cross that there were approximately 98,000 internees. After the war, the Dutch government stated that there had been approximately 110,000 internees. At the end of her thesis De Japanse interneringskampen voor burgers gedurende de Tweede Wereldoorlog (‘The Japanese internment camps for civilians during the Second World War’, Groningen 1963), Dr. D. van Velden compiled the available data. On the basis of this material, L. de Jong quoted the following figures in Het Koninkrijk der Nederlanden in de Tweede Wereldoorlog (‘The Kingdom of the Netherlands During the Second World War’), vol. 11b, pp. 348 and 359:

Java: 80,000 Dutch internees
Sumatra: 12,000 Dutch internees
The Great East: 3,900 Dutch internees
Borneo: 500 internees
Total: 96,400 internees.

The data collected by Van Velden are most likely incomplete, which is why De Jong thought it wise to assume that there were approximately 100,000 internees (p. 754). In his book Mannen van 10 jaar en ouder (‘Men aged 10 and above’, Franeker 1995, p. 82), Zwitzer arrived at a similar estimate, based on the number of post-war claims of former internees for Japanese benefits. The number of 100,000 Dutch civilian internees thus seems to be the most reliable estimate.

There is less agreement on the number of Dutch civilian internees who died in the camps. Estimates vary between 10,580 and 16,800 victims. In her thesis (De Japanse interneringskampen, page 368, note 1), Van Velden noted that The Netherlands Red Cross estimated the number of deaths at 13% (which would mean 13,000 dead). Zwitzer arrives at a figure of 10,580 killed by correcting the number of 13,000 based on the pre-war mortality rate (Mannen van 10 jaar en ouder, p. 83). Based on Van Velden’s material, De Jong deduced that 13,120 people died (13.6%). De Jong also mentions that in the course of the war, the Japanese reported a death toll of 16,800 (17%) to the International Red Cross. De Jong concurs and arrives at a mortality rate of one in six (Het Koninkrijk der Nederlanden in de Tweede Wereldoorlog, vol. 11b, pp. 753-754).

Internments in the Outlying Districts and Java (De Jong, vol. 11b from page 348).

The Great East: Close to 3,900 Dutch nationals and about 240 people of other nationalities. Almost all of them were moved to Celebes in the course of 1942, with the exception of a group of about 500 people that remained on Ambon.

Dutch Borneo: About 500 internees. A group of almost 300 of them was taken to British Borneo in July 1942, where there were already about 260 (mainly British) internees.

Sumatra: About 12,000 Dutch nationals and over 1,200 people of other nationalities (including about 700 British).

Java: Van Velden: About 80,000 Dutch nationals, about 700 British, about 100 Americans and about 1,800 people of other nationalities (about 29,000 men, about 25,000 women, and about 29,000 children)

Indonesians and (Indo-)Europeans:

There are only rough estimates of the number of Indonesians who perished during the Second World War. L. de Jong notes that of the 50 million Javanese and Madurai, approximately 2.5 million are estimated to have died during the Japanese occupation, mainly from starvation (Koninkrijk, 11b, p. 572). In the early 1950s, the Indonesian government estimated the total death toll at 4 million (S. Sato, War, p. 155).

During the occupation, the Japanese military police arrested approximately 15,000 persons (Indonesian Europeans, Indonesians and Chinese), 5,000 of whom were executed and 7,000 died in captivity (D.M.G. Koch, ‘De Japanse bezetting van Indonesië’ (‘The Japanese occupation of Indonesia’), in: Onderdrukking en verzet; Nederland in oorlogstijd (‘Oppression and resistance; The Netherlands in wartime’), vol. IV, Arnhem/Amsterdam [1954], p. 584).

It is not known how many non-interned Indonesian Europeans died of hunger and other hardships. Estimates of the number of Indonesian Europeans who eventually managed to remain outside of the camps vary widely from 120,000 to 200,000. Historian Hans Meijer estimated the number of ‘outmates’ (those who had not been interned) at approximately 125,000 (In Indië geworteld (‘Rooted in the Dutch East Indies’), Amsterdam 2004, page 226).

By 1944, the European population of Bandung who had stayed outside the camps had fallen to a total of 15,222 (excluding foreigners): 10,881 women and 4,341 men, only 329 of whom were between the ages of 20 and 40, compared to 3,860 women and 2,852 boys under the age of 15. In the area surrounding Bandung lived another 2,901 people. By March 1945, the European population of Bandung consisted of 4,523 men and 10,952 women, 15,475 people in all, while the number of people in the surrounding area had risen to 4,051. These were the persons registered with the Badan Oeroesan Golongan Indo (BOGI, ‘Board for the Representation of the Interests of the Indos’). (NIOD, archive 400, number 2232, report by R.B. Quack on RAPWI work in and around Bandung, 1946, page 6.)

As a result of the Japanese occupation of the Dutch East Indies, an estimated number of more than 21,000 ‘Europeans’ died (8,200 prisoners of war, about 13,000 civilian internees, and an unknown number of ‘outmates’). Before the war, the European community in the Dutch East Indies numbered about 290,000 people, about 260,000 of whom had the Dutch nationality. This would mean that the European mortality rate was higher than 7%.

Romushas:

It is not exactly clear how many Indonesians were forced to work as romushas (Asian forced labourers)) during the Japanese occupation. According to the Japanese statistics that were preserved, in November 1944 the number of romushas on Java was about 2.6 million, almost one million of whom were employed on a temporary basis. Since these temporary workers would have had to be relieved by others at some, the total number of deployed Javanese ‘work soldiers’ during the Japanese occupation was probably much higher than 2.6 million. In 1951, the Indonesian government estimated the total number of romushas at 4.1 million, most of whom had only been deployed for a relatively short time and had not been deported from the islands on which they lived (L. de Jong, Koninkrijk, 11b, p. 530). However, Japanese historian Shigeru Sato calculated that the total number of temporarily mobilised romushas on Java must have been nearer 10 million, based on the estimate that in the twenty months between January 1944 and August 1945 (the period of the most intensive recruitment of romushas) each temporary ‘work soldier’ was deployed for an average of about two months. If one adds the permanent romushas (about 1.6 million), the total number of deployed Javanese romushas comes close to the total ‘mobilisable’ labour force on Java, which the Japanese army estimated at 12.5 million (Shigeru Sato, War, nationalism and peasants; Java under the Japanese occupation 1942-1945, St Leonards 1994, pp. 157-158). This estimate is in line with the findings of Japanese historian Aiko Kurasawa that “almost all healthy men of working age except those who were invalids and those with leading positions in village society were mobilised as romusha at least once during the occupation” (Aiko Kurasawa, Mobilisation and control: a study of social change in rural Java, 1942-1945, Cornell University 1988, p. 259).

In some publications, the term romusha is (wrongly) used only for those Javanese who were deported overseas. According to the most common estimates, about 300,000 Javanese labourers were shipped to areas outside Java (L. de Jong, Koninkrijk, 11b, pp. 535-536; Henk Hovinga, ‘Einde van een vergeten drama’ (‘End of a forgotten tragedy’), in: Elly Touwen-Bouwsma and Petra Groen, Tussen Banzai en Bersiap; De afwikkeling van de Tweede Wereldoorlog in Nederlands-Indië (‘Between Banzai and Bersiap; Winding up the Second World War in the Dutch East Indies’), The Hague 1996, p. 74). According to Shigeru Sato, however, the number of Javanese sent overseas was between 160,000 and 200,000 (War, Nationalism and Peasants, pp. 157-158). 6,000 to 7,000 labourers from Bali and about 4,000 from Lombok were taken to other islands, not to mention the recruitment of labourers to work on the island itself (L. de Jong, Koninkrijk, 11b, p. 535, note 7; Remco Raben, ‘Arbeid voor Groot-Azië’ (‘Labour for Greater Asia’), in: G. Aalders et al. (eds.), Oorlogsdocumentatie ‘40-’45; Negende jaarboek van het Rijksinstituut voor Oorlogsdocumentatie (‘War Documentation 1940-1945’; Ninth yearbook of the National Institute for War Documentation’), Zutphen/Amsterdam 1998, p. 101).

The number of romushas who died as a result of the hardship is also impossible to determine precisely. The Japanese kept very poor records and burned many of them after the capitulation. A common estimate is that only about 77,000 of the approximately 300,000 Javanese romushas who were transported overseas survived – a mortality rate of 74.3% (Henk Hovinga, ‘Einde van een vergeten drama’ (‘End of a forgotten tragedy), p. 74). The number of 77,000 survivors quoted by Hovinga, however, only concerned the Javanese who ended up in the reception camps of the Nederlandsch Bureau voor Documentatie en Repatrieering van Indonesiërs (Neboduri, ‘Dutch Bureau for Documentation and Repatriation of Indonesians’), while still unknown numbers of forced labourers were still roaming around, settled on the spot, or tried to return to Java on their own. Based on some scattered data, Remco Raben estimates the mortality rate in the Javanese romusha community in the Outlying Districts at around 50%, “and much higher on the ‘railways of death” (Remco Raben, ‘Arbeid voor Groot-Azië’, pp. 102-103). Shigeru Sato again takes a different view: he estimates the number of romushas who were sent overseas and survived to be about 135,000 (War, nationalism and peasants, p. 160).

According to L. de Jong, about 85% of the Balinese sent overseas died. De Jong also mentioned that of the approximately 100,000 romushas who worked on the railway line and in the coal mines of the Bantam residence on the western tip of Java in November 1944, only about 10,000 were still there at the time of the Japanese capitulation (L. de Jong, Koninkrijk, 11b, p. 535 note 7 and pp. 532-533). That would mean that 90,000 Javanese may have perished in Bantam. Further information on the number of romushas killed in Java is lacking. The same goes for data on the deceased local labourers in Borneo, Celebes, Sumatra, and New Guinea who were put to work on their own islands. Henk Hovinga therefore thinks it likely that a total of considerably more than 300,000 Indonesians perished in the performance of their romusha duties.

Apart from Indonesians, the Japanese also recruited large numbers of Chinese, Malays and Tamils in Singapore and Malaysia. Most of them had to work on the Burma-Siam railway, together with the Burmese and Thais who had been recruited. It is likely that a total of 190,000 romushas from the British colonial territories worked on the railway. Reports and witness statements indicate an 80% mortality rate among these forced labourers. That would bring the number to 152,000 deaths. Together with the Indonesian romushas who perished, the total death toll in South-East Asia would then be around 450,000, according to Hovinga (Henk Hovinga, ‘Einde van een vergeten drama’, pp. 136-137). De Jong states that the approximately 4,000 romushas found in Thailand after the Japanese surrender were the survivors of groups that together numbered at least 10,000 Javanese (L. de Jong, Koninkrijk, 11b, p. 535). According to Japanese data, 6,173 Javanese were deployed on the Burma-Siam railway, but according to a Dutch officer (a former prisoner of war), who conducted research into this matter in August and September 1945, the number was at least 7,408 and probably much higher (NIOD archive 400, number 565).

Estimates of the total number of romushas deployed on the Burma-Siam railway range from 200,000 to 250,000 or more. As for the mortality rate, researchers agree that it was at least 50% and probably much higher. In some places along the railway line, the mortality rate may have been as high as 80 or 90%. This means that estimates of the total number of deceased romushas range from about 100,000 to about 225,000.

As estimates vary considerably, it is not possible to come up with an even remotely exact mortality rate. For instance, the mortality rate of Javanese romushas sent overseas is set at 74.3% by Hovinga, while Sato suspects it to have been between 15 and 40%. The mortality rate for the total number of Indonesian romushas deployed could perhaps be put at about 7.5% - assuming that there were 4 million forced labourers of whom 300,000 died - but it must be stressed that this percentage is based on very rough estimates as well.

Heihos:

Heihos were Indonesian auxiliary soldiers who fought alongside the Japanese troops.

L. de Jong, Het Koninkrijk der Nederlanden in de Tweede Wereldoorlog, vol. 11b, p. 958: ‘The total number of heihos is unknown - the last commander in chief of the Sixteenth Army of the Japanese [...] Lieutenant General Yuichiro Nagano, stated shortly after Japan’s capitulation that in August 1945 there were still almost 25,000 on Java and about 2,500 on Timor, and that from Java several tens of thousands of heihos had been moved elsewhere, “at their own request”, or so he claimed, of whom about 15,000 were said to be still alive.’

Idem, page 960: ‘Van Witsen estimates that of about half of the approximately 15,000 former KNIL soldiers who became heihos on Java were killed. How many of the other tens of thousands who became heihos also lost their lives is unknown.’

Based on the figures quoted by De Jong, the number of heihos can be estimated at around 60,000.

Forced prostitutes:

Bart van Poelgeest estimated that there were between 200 and 300 European women employed in Japanese military brothels in Indonesia. In about 65 cases, it was established that the women were actually forced to work as prostitutes (‘Gedwongen prostitutie tijdens de Japanse bezetting’ (‘Forced prostitution during the Japanese occupation’), in: Wim Willems en Jaap de Moor (eds.), Het einde van Indië (‘The End of the Dutch East Indies’), The Hague 1995, pp. 186-187). Van Poelgeest does not give an estimate for the number of Indonesian forced prostitutes.

According to George Hicks, the most reliable estimates of the total number of forced prostitutes (‘comfort women’) indicate that there were 80,000 Korean women and probably about 20,000 women with Japanese or Taiwanese nationalities or originating from the Japanese occupied territories (‘Japanse legerprostitutie 1932-1945: een overzicht (‘Japanese army prostitution 1932-1945: an overview’), in: N.D.J. Barnouw et al., eds., Vijfde jaarboek van het Rijksinstituut voor Oorlogsdocumentatie (‘Fifth yearbook of the National Institute for War Documentation’), Zutphen 1994, p. 19). That would mean that the number of forced prostitutes from Indonesia was lower than 20,000.

George Hicks calculated that the ratio between the number of Japanese troops and the number of forced prostitutes can be put on average at about 50:1 (The Comfort Women, London 1995, page XIX). At the time of the Japanese capitulation, there were approximately 300,000 Japanese in the Dutch East Indies (Elly Touwen-Bouwsma and Petra Groen, eds., Tussen Banzai en Bersiap (‘Between Banzai and Bersiap’), The Hague 1996, p. 95). That would mean that about 6,000 women and girls worked in Japanese army brothels in the Dutch East Indies.

After researching the matter, the Japanese writer Fumiko Kawada concluded in 1995 that there had been 22,234 victims of sexual violence in the Dutch East Indies, a number that included mistresses and rape victims. In 1992, the Indonesian journal Tempo claimed that there had been 60,000 victims, without mentioning its sources. (Numbers quoted by Brigitte Ars, Troostmeisjes (‘Comfort Girls’), Amsterdam 2000, page 145.)

Allied soldiers killed in action:

The number of Allied soldiers killed in South-East Asia and the Pacific was over 140,000. This figure is based on the following data:

United States: approximately 93,000 deaths

United Kingdom and Commonwealth (except Australia): at least 27,444 deaths

Australia: 17,501 deaths

The Netherlands: 2,654 deaths

New Zealand: at least 211 deaths

The Filipino soldiers killed under American command, the approximately 1.4 million Chinese killed, and the approximately 12,000 members of the Soviet armed forces killed in August 1945 are not included in this number.

Hiroshima and Nagasaki:

The main problem in estimating the number of victims/deaths from the atomic bombs is that nobody knows exactly how many civilians and soldiers there were in Hiroshima and Nagasaki at the time of the nuclear explosions. The figures quoted in Volume 11b of Koninkrijk der Nederlanden in de Tweede Wereldoorlog by L. de Jong - i.e. 64,000 killed and 72,000 wounded in Hiroshima, and 39,000 killed and 25,000 wounded in Nagasaki - are most likely based on the results of various American and Japanese research reports from the second half of the 1940s and the 1950s. These results have since been cited in many authoritative books. However, these figures may be too low. In 1950, a census of the number of survivors of nuclear explosions was published for the first time in Japan. Subtracting the number of survivors from the estimated number present during the atomic bombing, it was concluded that by 1950, 200,000 people had died in Hiroshima and 140,000 in Nagasaki. However, since many survivors had not yet come forward by 1950, these estimates were far too high. In 1976, the two affected Japanese cities reported to the Secretary-General of the United Nations that the atomic bombs had caused 130,000 to 150,000 deaths in Hiroshima and 60,000 to 80,000 deaths in Nagasaki by the end of 1945. American military historian Richard Frank drew the following cautious conclusion in 1999: “The actual total of deaths due to the atomic bombs will never be known. The best approximation is that the number is huge and falls between 100.000 and 200.000.” (Richard B. Frank, Downfall; The end of the Imperial Japanese Empire, New York 1999, p. 287.)

There is also great uncertainty about the number of people who died in later years as a result of the nuclear explosions. Due to the long-term effects of the atomic bombs, the mortality rate among those who survived the explosions is clearly higher than the national average, but no reliable figures are available. Hiroshima and Nagasaki keep death lists of deceased hibakusha (victims of the atomic bomb). A hibakusha is someone who was within two kilometres of the epicentres of the atomic bombs during, or immediately after, the explosions. In August 1994, the Hiroshima list numbered 186,940 dead and the Nagasaki list 102,275. By August 2000, the Hiroshima death list contained more than 217,000 names. When a hibakusha dies, he or she is automatically put on the death list, regardless of the cause of death. It is therefore not certain that all those on these lists actually died as a result of the atomic bombs. Nevertheless, according to American historian John Dower, the total of “immediate and longer-term deaths caused by the bombing of the two cities” can be estimated at “as high as triple the familiar early estimates - in the neighbourhood, that is, of three hundred thousand or more individuals” (Michael J. Hogan, ed., Hiroshima in History and Memory, Cambridge 1996, p. 125).