What can we learn from the ANDB database about refugees during the World Wars?

By Mayra Murkens, Nicolas Sy and Jasmijn Janssen

Introduction

Amsterdam has been a hot spot for diamond workers for centuries, although its importance waned during the first half of the twentieth century. Many of those diamond workers unified in the Dutch Diamond Union (Algemene Nederlandse Diamantbewerkersbond, ANDB), which secured solid wages and an 8 hour workday for the workers. Although the union’s name refers to the Dutch nationality, the majority of its members (~73%) had a registered birthplace beyond the borders of the nation. The database of the ANDB has therefore great potential to look into migration patterns during the operation of the union. As such, it may also tell us more about migration influxes during the World Wars. Belgian refugees sought refuge in the neutral Netherlands during the First World War, while we may expect an influx in Jewish refugees shortly prior to the Second World War. Since the diamond industry was a predominantly Jewish enterprise, the ANDB serves as a perfect test case for looking into refugees shortly prior to the Second World War in particular. Our aim is therefore to endeavor into an initial exploration of the ANDB database, with a focus on its potential to provide insights into migration patterns, in particular people relocating during disruptive periods where we may expect high numbers of refugees. We focus our initial exploration on two elements. First we want to know where the ANDB members originally came from during those periods, and see whether there were differences between the national background of the members over time. Second, we want to know whether there was an increase in members from a certain place over time, which could be related to fleeing disruptive circumstances at their place of origin.

Possibilities & Limitations

The ANDB contains the records of its union members for the period 1893-1958. This creates a rich dataset, with information on the nationality of the member, the registration date, the date of departure if they left, and potential relatives among the members. However, we do need to account for the potential limitations in using this database. First of all, it only contains the members of the ANDB, which is not necessarily the same as having the information of all diamond workers in the city. Although membership of a union was widely spread, we cannot assume that all diamond workers in Amsterdam were indeed members. Diamond workers whose relatives or friends were members of the union, were potentially more likely to join the union as well, since their network may have convinced them to join. Moreover, we will use registration dates as our point of entry and thus as our time variable. However, the registration date of the membership might not reflect the arrival in Amsterdam per se. It could very well be that people arrived in the city and worked as a diamond worker for a while, before joining the union. Any effects we may see of refugees could therefore potentially underestimate actual increases in migration. Moreover, migrant diamond workers who did not have any relatives or acquaintances who were union members were also less likely to join the ANDB, which may have occurred more frequently for migrant workers. It is therefore unfortunate that we could not generate the script necessary to look into these networks. We also use place of birth as an indication to see whether someone migrated or not. We aggregated this by nation, since we are most interested in international migration patterns related to crises in other European countries. Looking at the exact place names that were listed for migrant workers could be interesting when looking at immigration patterns within the Netherlands and Belgium. Yet, another argument for using countries, is the ambiguity of place names. For instance, when looking at the entire timespan of the dataset, the town of Deurne shows up as one of the highest ranking birth places just a few spots below Amsterdam and Antwerp. There is a Deurne in the Netherlands and one in Belgium. Even though the one in Belgium is the most likely fit -as it is located close to Antwerp- it is still ambiguous whether all workers came from this town. What is more, place of birth could however reflect a quite distant history of an individual, since the members may have migrated already on many occasions before their arrival in Amsterdam. For example, the individuals who were born in Belgium, could have been living just across the border in the Netherlands for quite some time already. Yet, the only information we which could indicate migration, is the place of birth, meaning that this is the information we have to work with. However, this does warrant us that we should be cautious with interpreting results, especially when we look at migrants who were born only on the other side of the border.

Data manipulation

First of all it was a challenge to combine the birth year, registration year, and birthplace in the script in such a way that the graph visualized where diamond workers were born for a selected span of registration years. In earlier versions the script linked birth years to birth places instead of registration years to birth places. In the final version of the script the number of times a birth place occurred was visualized for the selected window of registration years. Furthermore, as was mentioned in the possibilities and limitations section, we started out by visualizing the cities, towns and villages where workers were born. This resulted, however in a long list of place names that provided little insights or possibilities for analysis. We therefore coded the script in such a way that the countries of birth were listed, which was significantly more helpful to answer our research question.

The origins of diamond workers during World War I

Before the start of World War I, the members of the ANDB came either from within the Netherlands, or from Belgium. During World War I, still 40% of the members came from Belgium. This was a relatively smaller share compared to the decade before WW I, although in absolute numbers, there were almost three times as many members (n=413) joining the union during the period of the war, compared to the entire decade before (n=111). After the war, the number of new Belgian diamond workers continued to increase, both in absolute numbers as relatively. The increase in Belgian union members might thus have been the result of an influx of Belgian refugees, yet this seems rather unlikely given the continuing increase of Belgian union members in the following decade.

Refugees during World War II

The variety in origin of the ANDB members changed drastically in the build-up to World War II. During the 1920s, already some members were born in Eastern Europe, yet these numbers increased drastically during the 1930s. Specifically diamond workers from Poland started to join the ANDB during the period from 1933 until 1940, but also other Eastern-Europen and Central-European countries saw some diamond workers move to Amsterdam. During the German occupation of the Netherlands, the amount of the newly registered diamond members from the Eastern-European countries was lower, and the Poles were only the third largest national group of new members.

The fact that there are very few German diamond workers is striking. Perhaps there was no substantial diamond industry in Germany, which could explain the very low numbers of German workers, or the diamond industry had even better working conditions than Amsterdam. Without any prior knowledge of the diamond industry, this seems to be the most likely explanation to us, since there were no members of the ANDB originating from Germany in any of the researched periods. The potential explanation that German diamond workers chose different destinations because of the close proximity of the Netherlands is therefore unlikely. Another possibility is that German diamond workers did come to Amsterdam but that they did not join the union.

Conclusion

The vast majority of diamond workers who were members of the ANDB came originally from Belgium. Although we may have expected a unique increase in Belgian members during the German occupation of Belgium during WWI, the increase in Belgian workers in this period rather seems to reflect a steady trend in attracting Belgian diamond workers. There does not seem to be a correlation between the amount of Belgian workers and the disruptive period of WWI, based on these exploratory descriptive statistics. However, the opposite was the case during WWII, especially during the build-up towards the war. Especially diamond workers from Eastern-European nations, Poland in particular, joined the ANDB. They did so mostly during the build-up towards the war, which can indicate that a certain tenseness regarding the geopolitical situation was being felt, especially amongst a group which we know was most likely predominantly Jewish. The fact that the amount of members from these countries was smaller during the period of German occupation in Amsterdam, can be explained by two most likely reasons. First, the Netherlands were not a safe haven anymore once they became occupied territory. Jewish diamond workers who still had the opportunity to flee, may have chosen different destinations. That is to say, if there would have been Jewish diamond workers that still had the opportunity of fleeing, and were not yet incarcerated, or worse. Thus, we conclude that we mainly find refugees in the period of World War II, and not so much during World War I. We believe this is mainly related to the strong Jewish ties of the ANDB, which naturally were an important element during the Second World War. When the Netherlands were still regarded as relatively safe, prior to them being under German occupation, yet when unrest was already increasing in Europe, union members from a wider international variety joined the ANDB. Unfortunately, the role of religion cannot be fully substantiated, since the database does not include information on religion. However, the industry is known for its strong ties to the Jewish population, which increases the likelihood of these relations. Other researchers could use these findings as a starting point for further research. In particular the findings for the Second World War appear to be interesting. With large numbers of members coming from the Eastern-European countries during this period, we have found an indication that the Netherlands may have been viewed as a particular safe haven. The fact that the Netherlands had been neutral during World War I may have fueled this belief. A comparison between the ADB and the ANDB for this period may give more insight into the mechanisms of migration. Since Belgium had been occupied already during World War I, migrants might have opted more often for the Netherlands during the build-up of WW II. Ultimately, this could teach us more about potential motivations of migrants in times when they face disruption in their place of birth.