Nazi Registration of Jews in 1941: Retrieved Amsterdam List of Jews
In January 1941, the Nazis ordered all persons living in the Netherlands who had one or more Jewish grandparents to be registered. Amsterdam residents had to pick up their registration form in alphabetical order by last name between 10 and 18 February at special appointed locations, and return the completed questionnaire between 10 March and 7 April (Stuldreher 2007: 80–81); questions were included about the personal religious denomination of individuals, their spouses and their grandparents. The completed returned questionnaires were compared to the information in the Amsterdam population registry, and the registry cards of Jewish persons were marked with a ‘J’ and signalled with a clip (Stuldreher 2007: 79; Tammes 2009). An overview of the van de Bevolkingsregisters (1942) dated October 1, 1941 mentions 86,291 Amsterdam residents having one or more Jewish grandparents, among them 79,497 persons classified as ‘full Jews’ (e.g. Presser 1965, part 1: 62).Footnote 1 Those ‘full Jews’ were persecuted by the Nazis; in this article they are referred to as Jews.
Dutch historians generally believe that practically all Jews complied with the order to register (De Jong 1969–1991 part 4: 874–875; Herzberg 1978: 50; Presser 1965, part 1: 62–63). Most Jews were known in their social network as being Jewish or of Jewish origin. At that time Jews could hardly have anticipated the life-threatening measures that would be taken by the Nazis in the years to come and therefore had little reason to refuse registration, particularly as citizens’ religious denomination was commonly recorded in municipal registries and reported at each decennial census.
After the liberation the marked Jewish registry cards were replaced by unmarked cards (Croes and Tammes 2006: 34), and in the post-war corrected registry one cannot trace assimilated Jews, i.e. persons who had three or four Jewish grandparents but had abandoned Judaism. Croes and Tammes (2006) recovered an undated registration list of Amsterdam Jews that mentions a total of 77,238 persons.Footnote 2 This list contains the name, date and place of birth, marital status, address, religious affiliation, nationality and occupation of each Jew. As the youngest person on the list was born on May 7, it might be assumed that this list was being finished in the second week of May 1941 (e.g. Tammes 2011). Although this retrieved list is missing 2259 Jews when compared to the statistical overview of the Van de Bevolkingsregisters (1942), it is the only source on the Dutch capital containing information on tens of thousands of individual Jews. The question is whether these 2259 persons are a specific group that might be underrepresented on the retrieved registration list.
Among those missing persons could be the group of about 400 young adult men who were caught in a roundup on February 22, 1941 as a response to Jewish resistance, in a harassment by the uniformed commando group of the Dutch National-Socialist Party (NSB) and German patrol groups that resulted in the death of a NSB member. All 400 men were deported to Buchenwald at the end of February, and those who were still alive in May 1941 were transferred to Mauthausen (Presser 1965, part 1: 86–88). These Jews, however, are registered on the retrieved Amsterdam list as we will see later on when deceased are split up by place of death and year of death. They probably had already picked up their registration form the week before they were caught, and this form might have been returned by a household or family member.
To identify possible missing or underrepresented groups, the retrieved Amsterdam listFootnote 3 is compared with the overview of the van de Bevolkingsregisters (1942) and a statistical overview of Jews in Amsterdam given by Veffer (1942) on the following four characteristics: gender, age, nationality and religious affiliation. According to the van de Bevolkingsregisters (1942: 22–23), 52.2 % of Amsterdam Jews were females. Based on the maiden name of married women and the first name, 51.5 % of the Jews on the retrieved list were women. Females are slightly underrepresented on the retrieved Amsterdam list. The age distribution based on the retrieved Amsterdam list hardly differs from the overview given by Veffer (1942); the biggest difference is 0.3 % in the 50–59 age group for females and 0.4 % in the 10–19 age group for males. Based on the nationality given on the retrieved Amsterdam list, 86.5 % had the Dutch nationality—very close to the 86.8 % Veffer (1942) calculated. The list also counts 8.7 % German, 2.7 % stateless, 1.3 % Polish and 0.7 % other Jewish nationals. Those percentages are again very close to Veffer’s calculations.
Although all persons on the retrieved Amsterdam list had three or four Jewish grandparents, they themselves might have not longer belonged to an Israelite congregation. Based on the religious denomination given on the retrieved Amsterdam list, 91.8 % belonged to an Israelite congregation, the same percentage given by van de Bevolkingsregisters (1942: 22–23). The retrieved list counts 7.4 % religiously unaffiliated persons and 0.7 % belonging to another congregation, nearly all converted to Christianity, as confirmed by the van de Bevolkingsregisters (1942). Although the retrieved Amsterdam list is missing 2259 Jews compared to the overviews of the van de Bevolkingsregisters (1942), it is not biased with respect to gender, age, nationality or religious denomination. This retrieved Amsterdam list is therefore an excellent source for answering the research questions.
Lists of Victims of the Holocaust
A listing of all Jews who were deported from the Netherlands and perished without a grave is published in the book In memoriam-Lezecher (IM) (1995) as a means of honouring the memory of those who did not have a proper burial. This memoir contains the names, date and place of birth, and date and place of death of over 101,000 Jews. These data were gathered by the Red Cross, the Dutch Institute for War Documentation, and the Dutch Ministry of Defence and Ministry of Foreign Affairs, and checked against the population registries. The adjustments published in two addenda in 1997 and in 2000 are included in the digitised version of IM used in this study.
Not all Jews who perished during World War II are mentioned in IM though. Jews who died in the Dutch concentration and transit camps Westerbork and Vught, and those who perished outside a Nazi camp or who had a grave are not mentioned in IM. For this reason, lists of Jews who died in Westerbork or Vught and buried Jews mentioned in other death lists are put in a different victimisation database (WB+), counting in total more than 1200 Jews.Footnote 4
I also had access to a database containing data extracted from the website Digital Monument to the Jewish Community in the Netherlands (DMJ).Footnote 5 DMJ is an Internet monument dedicated to preserving the memory of all those who were persecuted as Jews during the Nazi occupation of the Netherlands and did not survive. Since its launch in 2005 this website is continuously updated. The DMJ database contains information such as first, last and maiden name, date of birth, and last official place of residence on more than 104,000 Jews, including those who died in Westerbork and Vught and other Dutch locations, as well as recent corrections to IM.
The victims mentioned in those databases overlap significantly. However, as the listings’ initial purposes to gather and publish information on victimisation of the Holocaust differ, the databases might also differ slightly in terms of the victims listed, and none of the lists will be complete. To minimise underreporting of specific groups of victims and to avoid ‘political’ biases (Aronson et al. 2013: 290; Brunborg et al. 2003: 236–237), this study uses all three databases to determine who among Amsterdam Jews fell victim to the Holocaust.
Lists of Survivors
Amsterdam Jews not matched to the lists of victims or reported as missing on the victim lists were further investigated by matching them to other post-war information (e.g. Brunborg et al. 2003). Ideally, these Jews would be checked against the names on the Dutch post-war population registry to determine whether they were alive after the liberation. However, this registry is not computerised and is not easy accessible due to privacy legislation. In addition, not all survivors returned to their place of residence or to the Netherlands. Instead of the registry, this study uses a database of Jewish Holocaust survivors who had lived in the Netherlands during the Nazi occupation. Lists of returnees and camp survivors had been put in a database, counting 22,692 Jewish survivors.Footnote 6 Some survivors were registered on multiple lists. After identifying multiple registered persons using surname/initials, given name and/or date of birth, the database counted 16,704 unique persons. In addition to surname, last name and date of birth, this database contains information on place of birth and former or last place of residence, although it is unclear what that means exactly. However, for many persons, data are missing. For about half no former or last place of residence is given, and for more than half no place of birth is mentioned.
The 16,704 persons in the survivors database are far lower than the estimated number of survivors since, among others, the returnee registry is far from complete (Van de Vosse 1947). Hence, this database can only account for part of the Jews not matched to victimisation lists who had shown signs of life after the liberation. Matching might also be useful to determine which groups of Jews are underrepresented in the Dutch survivors database according to their socio-demographic profile.
Matching Amsterdam Jews to Lists of Victims
To compare the Jews mentioned on the retrieved Amsterdam list to Jews mentioned in the victimisation databases, this study uses a deterministic linkage approach by constructing a unique matching key (e.g. Grannis et al. 2002). To this end, only individual characteristics present in both databases can be used. Although Amsterdam Jews were asked to give their complete first and last name when registering in 1941, it might be better not to use the complete name to avoid mismatching due to possible different spelling or typos. Following up on the matching method developed by Croes and Tammes (2006), this study uses the first two characters of the first name and the first two characters of the last and maiden name of Jews registered on the Amsterdam list that were entered into a database, but excludes the prefix of the surname, like van (‘of/from’), de/het/’t (‘the’), der (‘of the’), as such prefixes are quite common. The combination of these two components is not that unique and produces many double matches; to avoid these, the complete date of birth is added. Married women also undergo a matching that includes the first two characters of the maiden name instead of the last name.
This developed matching key appeared to be very unique: only 76 combinations exist twice. This means that 152 Jews (0.2 % of all Amsterdam Jews) could not be uniquely identified on the Amsterdam list. These 152 persons were manually checked to see if they had perished during the Nazi occupation using the described victimisation databases. For 210 Jews it was impossible to construct this matching key due to a missing value on one or several key components. Moreover, 112 persons appear twice on the Amsterdam list. In total, 76,916 (77,238-210-112) individual records were matched using the constructed matching key. The matching procedure is repeated three times, first matching Amsterdam Jews to IM, then to WB+ and finally to those mentioned in DMJ who resided in Amsterdam.
The non-matched Jews were subjected to a second matching procedure using an alternative matching key. Sometimes the date of birth was not too readable on the Amsterdam list, so an alternative matching key was constructed to perform a second matching procedure: first two characters of the first name, first two characters of the last name, and a combination of two of the three date-of-birth components: day, month and year. This resulted in three alternative matching keys for each person, namely the first two characters of the first name, the first two characters of the last name, and, respectively, day and month, day and year, or month and year. The results of this second matching procedure were checked manually.
Using a deterministic linkage approach, these matching procedures resulted in 58,144 matched individual records; 95.7 % were mentioned in both IM and DMJ (Table 1). Another 3.5 % were from DMJ only. All the matched persons found in WB+ were also found in DMJ or IM; only a very small number were found in IM only. Among the matched individuals 58,070 were matched to DMJ and all of them lived in Amsterdam.
Matching Amsterdam Jews to Lists of Survivors
To compare Amsterdam Jews who were not matched to lists of victim to survivor lists, the same matching key is used as described before. Within the Dutch survivors database, 2332 of the 16,704 Jews had an incomplete key due to missing information and 410 were born after the estimated registration date of the Amsterdam list (May 1941), resulting in 13,962 Jewish survivors to be included in the matching procedure. None of the Jews reported missing were matched to the survivor database. Using the matching key, 2878 of the 18,772 Jews not matched to victim lists were matched to the survivor lists (15.3 %).
For more than half of the 2878 matched persons, no last place of residence was given and for about one-third (about one thousand) Amsterdam was their last place of residence. Since about 2600 survivors had Amsterdam as last place of residence, using the alternative matching key described earlier the Amsterdam Jews not matched to victim lists once again are matched to the remaining 1600 survivors who had lived in Amsterdam. This yielded another 216 matched individual records resulting in 3094 matched survivors, which is 16.5 %. This low percentage of matched records might be due to the incompleteness and inaccuracy of the information on the survivor lists.
Cross-tabulations and Chi-square test show that for nationality German and Polish Jews were overrepresented and Dutch Jews underrepresented, and for civil status unmarried Jews were overrepresented and divorced, widowed and married Jews were underrepresented among the Jews matched to the survivor lists. Among married Jews, intermarried Jews were under-registered. The elderly and females were underrepresented, while no difference in post-war registration exists according to social class.