Computer vision for removing blind spots in a migrant registration system

Abstract

In this paper we discuss the extension of established methodology with computer vision to make it possible to almost literally see into the blind spots that using established methods on large serial collections only leave. We argue that this method overcomes the dangers of implicit selection that are commonly designated as ‘cherry picking’, or selecting the ‘most important’ files. Furthermore, in using combined traditional and DH-methodology it becomes visible what is and what is not in the collection as a whole. As such it is a a replacement for traditional leafing through an archive. As a method of source criticism, this gives many more possibilities than would have been possible with established methodologies only. We illustrate our findings on our project Migrant, Mobilities and Connection on Dutch-Australian emigration 1950-1992.

The main research question in the project is which factors determined the whole migration experience and what the relation was between policy, civil society and individual agency. Australia was, together with Canada, the main destination of Dutch emigrants after 1945, receiving around 160,000 migrants from 1950 to 1992.

The point of departure is a registration system that was kept by the Dutch migration authorities, based at the Dutch consulates. It consists of 50,000 cards (100,000 images scanned) and contains data about the interactions between migrants and the migration officers from 1950-1992. The cards themselves contain a wealth of information that is not readily available as the writing on the cards is a mixed of manuscript and typescript that are distributed unequally over the cards.

Before starting to answer the main question we had to determine first how to study Dutch-Australian emigrants with this extensive registration system that is hermetic by its size and composition. Traditionally, historians would tackle a collection like this by taking a sample from the cards and additionally study the most interesting cases.
However, case selection is difficult as it is impossible to read a hundred thousand images or even leaf through them. Moreover, it is not clear how cases fit into the registration system and whether there are hidden features of the system influencing the size of files.

Thus, a combination of a large archive collection of mostly undifferentiated material and methodologies not devised for distant reading, leads to blind spots for the historian and asks for additional methods to inspect the whole collection. The computer vision method we adopted measured the amount of writing on the cards. Viewed over the whole registration system, this gives a distribution of the information over the cards.Combining this with traditional sampling we were able to identify distinguishable groups of migrants (eg. by as religion, marital status or age).

In the paper we will elaborate on the (non-)possibilities of relating these groups to the ‘information distribution’ as a whole and on the (non-)possibilities of distinguishing changes in policies and executive strategies of the Dutch migration authorities by using this combined methods.

Full text

1
Computer vision for removing blind spots in a migrant
registration system
Rik Hoekstra rik.hoekstra@di.huc.knaw.nl (Huygens ING-KNAW Amsterdam) and Marijke van Faassen
marijke.van.faassen@huygens.knaw.nl (Huygens ING-KNAW Amsterdam)
Paper ESSHC-conference organized by Leiden University 26 March 2021 (online)
Introduction
Nineteenth and especially twentieth-century archives are full of index card registration systems that
contain a wealth of partially structured information. If only their size defies reading them all and
even leafing through them. It is even more complicated to use the structured information on them.
We had to face these problems a few years ago, when we started research using a card index system
on Dutch migrants to Australia. The cards are part of a registration system that was formed between
1948 and 1992 by the Dutch migration authorities. These were the years in which the Dutch
government together with civil society organizations facilitated emigration to overseas destinations
as part of the Dutch labour market policies. Australia was one of the main destinations of Dutch
emigrants after 1945, receiving around 160,000 migrants till 1992.
Post World War II Dutch emigration is both understudied and well-documented, as The Netherlands
are mainly thought of as an immigration country. Still, there were almost half a million emigrants in
mainly the 1950s and 1960s that merit research in their own right but also offer interesting
possibilities for comparison with immigrants.
In his 1950 New Year's address, the social democratic Dutch Prime minister Drees spoke about the
necessity for emigration. He said that notwithstanding the high Dutch population density the
country had 'hitherto succeeded in maintaining sufficient employment rates ... but that task [would]
be made easier when part of the population ... [would] dare to seek its future in larger areas [of the
world].'
1
This speech expressed the feelings in both Dutch government and bureaucracy that got its
expression in the organization and stimulation of emigration in the years to come. It also coincided
with the outcome of polls of a year before, in which one third of the Dutch population expressed its
wish to emigrate if this would be possible (Van Faassen 2014). With countries like Australia and
Canada migration treaties or informal agreements were drafted about quota and subsidizing. Official
emigration policies were in force until 1992.
In the public debate in the Netherlands and also in the memories of emigrants themselves this
emigration policies are often seen as an anonymous force that almost manipulated parts of the
population into migrating from the Netherlands (e.g. Fels 2016, p.69). This raises the question what
1
“Ondanks onze grote bevolkingsdichtheid zijn wij er tot nog toe in geslaagd een ruime werkgelegenheid te handhaven en
een redelijke voedselvoorziening te waarborgen, maar die taak wordt verlicht wanneer een deel van ons volk, zoals het ook
deed in vroeger eeuwen, het aandurft zijn toekomst te zoeken in groter gebieden, waar op den duur ook ruimer armslag te
vinden zal zijn.” Willem Drees, Nieuwjaarstoespraak 1950 NA-Archief Drees, 171/2(69?). Also Elseviers Weekblad 7-1-1950
NA-Archief Drees, 366

2
the balance was between personal decisions and policy, both national and international. This
problem was already discussed in 2013, in Schrover’s and Moloney’s interesting study on Gender,
Migration and Categorization. They state that “ the literature on migration patterns, networks and
ties focuses on the choices and deliberations of migrants and less on how states create frameworks
in which choices are made” (p.19).
They also observed that scholars tend to follow the categorizations that policy makers use, often as a
result of the source material that is available and organized according to these categorizations (p.9).
The categories they refer to are the four main categories of migrants used by states: postcolonial,
refugee, labour and family migrants (p.7). When we apply these categories to the Dutch emigrants -
most of them Dutch nationals - Drees referred to in his speech, they mainly belong to the categories
‘labour’ or ‘family’ migrants. When in 2014 we started the project Migrant, Mobilities and
Connection we took up the challenges of including these official categorizations into our analysis. In
recording the migrants, migration authorities did apply a number of categories that partly overlap
with the categories in Schrover and Molony‘s book. They include gender, occupation, religion and
ethnicity.
Migrant, Mobilities and Connection has a twofold aim: first digitally connect the cultural heritage of
Dutch-Australian migrants that is dispersed over many collections in many institutions in two
countries. The core of this effort are the migrant registration cards that were made by the Dutch
emigration services and travelled to Australia with the migrants where they were copied to serve as
input for the Australian immigration authorities. Subsequently they were repurposed by mainly the
social services officers that were positioned at the Dutch consulates in Australia to support the
migrants in their new home country (Van Faassen& Oprel 2020). In doing so we hope to have
established a new resource that is easily accessible for a larger public (the core business of Huygens
ING, see also Arthur et.al, 2018), but also facilitate our second aim: to start answering our main
research question: how are policy and migrant agency related with respect to the whole migration
experience?
Card index systems and methodology
This paper is meant to research the card files of migrant cards (NL_HaNA, 2.05.159). There are
51,525 cards that correspond to 100,000 images. They represent circa 180,000 migrants, or 80-90
percent of all migrants to Australia, as the cards contained data about migrant units, often but not
always families, with personal data about the migrants. The cards contain a number of data about
the migrants like personal data such as name, birthdate, occupation, place of origin and the dates
and means of migration. These were gathered before migrations. The cards contain additional
unstructured data about interactions with the consulates. Because they comprise such a large share
of the emigrants and a lot of their context, the cards enable us to connect all sorts of digital heritage
(Van Faassen & Hoekstra 2020, in press). They also make it possible to make informed samples.
From a research point of view, it is necessary to assess the cards. We have images of the cards and a
very summary index table with core data. In digitizing, the archive lost the connection between the
index and the images. Also, the original order was disturbed in many places. We cannot read all
cards because there are far too many and the cards contain a mix of typed and handwritten
information that made experts in text recognition for the most part shy away. Some typewritten

3
portions of the cards are readable by ocr, but cards are not evenly distributed and usually only
partially typed, so this does not provide us with a reliable overview.
2
We did make a general assessment of the cards with a 1 percent sample, taking each 100th card
(consisting of 2 images and possible follow-up cards as far as they were localizable), structuring and
analyzing the information on the cards. The sample confirmed that there is a wealth of information
about migrants on the cards about the Dutch-Australian migrant population. However well the
sample is taken, this results in a simplification, or a small world representation of a larger world
(McElreath, 2020, 19-46). Historians often complement this simplification by taking cases from the
collection and studying these in depth to get insight into the variation, an established method in
history. For the emigration cards, however, this posed a few different problems, especially the
selecting cases. It is very hard to select the largest cards as some historians suggested, as we have no
physical or visual access to the cards. Moreover, there may be a reason that files get big that would
lead to an unconscious selection bias, also known as cherry-picking, as it is unclear for what reason
migrants would get more attention. Notwithstanding the samples, the cards effectively were closed
for research and we had to find a solution to complement the sampling. This solution had to be
based on digital methods to be able to deal with the size of the registration system. We devised a
method with different steps (Hoekstra & Koolen, 2019).
3
We first manually reconstructed the relation between images and the index table. Then, we devised
a way to measure the information density on the cards, using a simple form of computer vision. The
mixed script on the cards may be too difficult to transcribe using a combination of OCR and HTR, but
it is possible to measure the amount of writing on the cards, using the script edges that can be
measured using software (see fig.1). In this way we do not know what is written, but how much is
written, giving a measure of the information density on the cards. The information density can then
be related to what we do know about the content of the cards.
2
Cards that are completely typed are usually copies for the file, made by the emigration attaché the moment a migrant
decided to move within Australia. The original card was sent to the consulate in the consular area of new arrival and
continued to record the migrants interaction with the emigration authorities.
3
The process is outlined in a series of Jupyter notebooks. A summary notebooks with references to more detailed analyses
is available in a Github repository
https://github.com/HoekR/MIGRANT/blob/master/results/exploring_data_integration/notebooks/Profiles.ipynb

4
fig.1 script edges on an migrant registrationcard
(source - emigrant registration cards)
What we know about the cards is analyzed in the sample analysis. We distinguished different stages
and influences on the cards, that may be depicted in the schema in fig.2
fig 2. schema of influences on migrant registration cards

5
The different stages in the scheme are all visible on the cards. As we wrote above, the front of the
cards was (primarily) filled in before migration, and the back of the cards after migration. The card
fronts are therefore better structured. There is also a time lag with an average of 2.5 years and a
median of 1 year between the date of emigration on the front of the cards and the first dates on the
back of the cards. Although the cards sometimes contain information about the travel themselves
and the first time after arrival, this is not structural. In the information distribution this translates to
a different characteristic for the card fronts compared to the card backs (fig 3).
Card fronts Card backs
fig 3 differing distribution of edge densities on registration card fronts and backs
(source - emigrant registration cards)
Because the cards contain partial reflections of the lives of the migrants, it is obvious to assume that
some properties of the migrants would determine the information on the cards. Of course, the cards
represent the perspective of the registering authorities, that is the Dutch Emigration service (NED) in
the Netherlands and consulate personnel, mainly the emigration attaches and the social work
officials in Australia. We have studied the possible relations between all the variables, ranging from
age to family composition, religion, place of residence in the Netherlands primary to migration and
the migrant scheme (some examples in fig 4)
.
Fig 4 Some examples of analyses of migrant properties and and card densities, that were
inconclusive
4
(source: emigration registration cards)
4
For more analysis see the notebooks in note 3 above.

6
Only the place of residence (fig 5) and the migrant scheme showed conclusive influence on the
information distribution. This allows for one conclusion, but also a further research question. The
conclusion is, that in selecting the largest files (that is the cards with the highest information density)
from the registration card files, will not introduce a selection bias for variables such as age, family
composition or religion. If we want to study migrant lives, the distribution of information density
does not reflect any sub groups among the migrants. Of course, there is a selection bias because
cards with a lot of information reflect the most eventful lives, for whatever reason, but this
constitutes a point of further study and it is a good idea to compare them with (a selection of) less
information dense files.
Connecting policy to the registration system
The further research questions that this raised, is why the place of residence and the migration
scheme did have a marked influence on the cards (fig 5). To be able to assess that, we have to
consider the intention of the instrument of the cards for the authorities. It was a type of monitoring
device that was used at two different stages of emigration. The first was to record and streamline
the emigration process itself. That stage, however, ended with the arrival of migrants in Australia
and was primarily recorded on the card fronts. The second stage that was recorded on the card
backs (and possible follow-up card) started well after arrival. The question is why the Dutch
authorities would care about the Dutch who had left. This is only obvious for the strict consular
activities in which official intervention of either a consul or the ambassador was required, such as
passport prolongation and remigration. But the range of activities employed by the consulates was
much broader and had to do with the well-being of the migrants. This reveals that there was an
active policy by the authorities aimed at making emigration a success.
5
5
For a similar assessment of the Old Bailey accounts as a public policy resource see Devereaux, 1996 and Shoemaker,
2008, who characteriezed the Old Bailey Proceedings not so much as a impartial record of crime proceedings but "as a
deliberate intervention, on the side of the authorities and the property-owning classes, in the century-long debate over
how to respond to the apparently ever- rising tide of criminality in London" (p. 580)

7
fig 5. Map of the origin of Dutch-Australian emigrants (1951-1992)
(source - emigrant registration cards)
The two influences of the migration scheme mentioned above reflect different sides of the policy.
The variation place of residence suggests that there was a conscious policy on the part of the
authorities to stimulate migration in parts of the country. Although this is known from
historiography, the focus has been primarily on the agrarian sector, as postwar emigration policy is
understood as a solution for the ‘small farmers problem’ in the Netherlands.
6
However, as
emigration was meant to be complementary to the industrialization policies (Van Faassen, 2014),
this calls for further research, because it seems likely that there also was a relation with postwar
changes in the industrial situation of the Netherlands and the closing of the Limburg coal mines in
the early 1960s.
6
This might also be an explanation for the underrepresentation of emigrant units from Zeeland and the Northern Dutch
provinces.

8
The effects of the different migrant schemes is much more subtle. Most migrants travelled under a
migration scheme, that is an agreement of the Dutch and Australian governments that subsidized
the passage on the migration ships or planes. The most important was the Netherlands-Australian
Migration Agreement, that was officially operative from 1951, but there were more schemes.
7
The
schemas implied an involvement of both the Dutch and the Australian authorities that found an
expression in many areas.
In the NAMA case, Australian co-subsidizing of the passage required migrants to work in Australian
government service for two years. The schema also included migrant selection with both Australian
and Dutch involvement (Schrover and Van Faassen, 2010). On the other side, it also implied that the
Dutch authorities wanted to make migration a success. Return migration was always sizable, but the
contemporary files in the archives were marked secret (NL-HaNA, 2.15.68, inv.nrs. 883-884; 1408)
and to take away reasons to return, the Dutch authorities invested in social officers that resided at
the consulates, partly in response to bottom up pressure from the civil society organizations, who
had the majority in the Dutch emigration governance system (Van Faassen, 2014, Ch.2). They
supported the emigrants by intervening in their affairs, providing assistance in all sorts of social
matters. They also were the ones who (predominantly) filled out the backs of the cards. In our
sample, we have classified the types of events they noted in categories. Below we will discuss the
event types ‘finance’ and ‘housing’.
fig 6. financial events/year
(source - emigrant registration cards)
An example of a direct influence of a migration schema is visible in the financial events (fig.6). From
the graph, it would seem that there were many migrants with financial issues in the mid-1960. Upon
closer inspection, however, it appears that these mostly stem from migrants that migrated under the
7
For an overview of the different schemes see resources.huygens.knaw.nl/emigratie/gids/instelling/3251792704, also van
Faassen, Polder en emigratie, 165-6.

9
so called program for youth (Jongeren Programma, JP) who could only stay for one or two years in
Australia and had to save with the consulates for their return fare. The sums they saved were
notated in succession on the registration cards. Archival research to find an explanation for this
phenomenon revealed that this temporary migration, embedded in youth programs, was a
deliberate policy of both governments, aiming at increasing the ’emigratability’ of the Dutch
population (when departure figures went down after 1956) by introducing young people to
perspectives abroad for a longer period of time. After their return to the Netherlands, the youth
could function as ‘goodwill ambassadors’ for emigration, as they were expected to supply
emigration supporting information (Van Faassen & Hoekstra, 2015).
8
fig. 7 housing events per year
(source - emigrant registration cards)
One of the other prominent problems in the Dutch-Australian migration was housing. There is a
wealth of studies on the first years after arrival especially from the migrant perspective, when
emigrants were housed in camps (or, formally, reception centers) like Bonegilla, Scheyville, Wacol
etc. or hostels and families were separated from the moment the man had to start working, for
instance to fulfill the NAMA-scheme requirements (Peters, 2001, Ch.4, Walcker-Birkhead, 1988, 190-
206; Eysbertse 1997) In contrast with the graph on finance, the housing graph doesn’t show specific
patterns that prompt analysis. In general, it follows the pattern of the departure peaks and tops,
which is rather obvious. However, analysing the housing events per migration scheme show more
variation (fig 8).
8
As follow-up of the Youth Programs (1959-1967) the Working Holiday Schemes were introduced (1981- ).NL-HaNa,
2.15.68, inv.nr. 1351-1353 en 1355-1357. To stimulate migration after taking part in the Youth programs a Secondary
Assistance Scheme was introduced.

10
NAMA
LP
other
JP
NESS
NGAS
social assistance
30%
39%
34%
27%
17%
32%
employment
25%
23%
18%
24%
12%
18%
financial
13%
8%
17%
33%
4%
26%
consular tasks
13%
20%
15%
9%
54%
10%
housing
8%
5%
5%
1%
4%
7%
administration
7%
4%
8%
5%
8%
2%
health
3%
1%
2%
0%
3%
unknown
1%
1%
0%
2%
2%
fig 8 event-per-scheme in percents
(source - emigrant registration cards)
The first thing that stands out is that the Youth Program (JP) showed the least housing events, in
contrast with the two largest and simultaneously runned schemes NAMA and the Netherlands
Government Agency Scheme (NGAS), which have the highest percentages of housing events.
9
This is
still rather obvious as migrants under the Youth programs usually travelled alone, or at least without
a family. For the other schemes (which are often affiliated with churches and their social networks) it
is known that most of the time sponsorships of private persons (including housing) were required.
Close reading of the housing events on the cards of the sample however reveals an intriguing shift
around 1960-1961. Complaints about the migrant camps and hostels then make way for questions
about ‘Building Societies’ and ‘Housing Committees’. Together with the downward trend in housing
events/problems after 1960 in the graph in fig. 7, this can be interpreted as a possible starting point
for further analysis on supposed governmental policy on housing. Archival research in both countries
indeed reveals the other side of the ‘camp -stories’ in migrant studies.
In 1948-1949 the Australian Government organised a conference on housing for ‘Australians and
migrants’ in which a policy was formulated to improve the supply of building materials and secure a
greater output of houses. One of the suggestions to the Australian states was to order more
‘prefabricated houses’ (NAA, A445/202/3/34). Archival research in the Netherlands shows a twofold
response to this idea. The Utrecht Building Company Bredero-De Vries, who already developed the
9
NGAS was a Dutch unilateral scheme for Dutch unskilled migrants, ineligible under NAMA terms, but in general suitable
for emigration. The NGAS also guaranteed work and accomodation (Walker Birckhead 1988).

11
idea of ‘prefabs’ during the Second World War (Clark, 2002, 24-25) took the momentum (1950) to
establish an Australian holding to build prefab-houses for Dutch migrants who were employed at the
Snowy Mountains Hydro Electric Schemes, which ultimately resulted in the now famous Dutch
Australian multinational Lendlease Corporation, led by Dick Dusseldorp (Schlesinger, 2018, Clark,
2002, 5, 23-25).
10
This initiative was supported by the Dutch government who developed in the early
1950s a policy of financing and sending prefab houses to overseas emigration destinations, in
addition to their policy to solve their own domestic housing issues. In researching the specific files it
becomes clear that this policy was based on reports from Dutch emigration attaches abroad. Thus
we can conclude that the Duch emigration attaches in Australia ‘translated’ the complaints recorded
on the emigrant cards in a more general policy issue on housing, leading the Dutch government to
react with ‘new’ policy. (NL-HaNA, 2.15.68, inv.nr. 842, esp. 1385).
Further research on the housing policy leads to the conclusion that the questions about the Building
Societies (and Housing Committees) on the cards from the 1960’s onwards must in turn be
understood as a reflection of a follow up of the ‘prefab phase’ of the housing-policy-abroad,
developed by the Dutch government from 1954 and implemented from 1959 till 1975. In order to
create a new incentive for emigration to Australia, the Dutch Government designed a policy in four
consecutive steps in which money was provided by the Development Loan Fund, Australian Banks
and later by Dutch institutional investors. They lended money to so-called Dutch(-Australian)
Building Societies, which in turn made it possible for Dutch migrants to borrow money on low
interest rates to buy newly builded houses (NL-HaNa, 2.15.68, inv.nrs 1262-1265). There even seems
to be a closer connection with the Dutch initiated real estate business in Australia, although the
exact relations require further investigation. Lendlease director Dick Dusseldorp started Civil & Civic
as a foreign branch of the Bredero's Bouwbedrijf holding, with which it always kept warm ties. In the
late 1950s Civil & Civic was involved in constructing residential areas in Sydney, in which director
Dick Dusseldorp's intention "was to go beyond the ‘standard’ spec-built house of the period to
produce repeatable model houses of superior quality " (Harfield & Prior, 2010, 1). One of the leads
to research the relation between the Dutch government and Dusseldorp Civil & Civic is the fact that
the moment the Bredero’s Building company in the Netherlands did not want to invest large sums of
money in Dusseldorp’s Australian Civil & Civic anymore, Dusseldorp succeeded in convincing the
Dutch National Investment Bank to give him a loan. These difficulties in financing ever larger building
projects led Dusseldorp to find innovative ways of financing and he started the Lend Lease Company,
that proved a very successful concept and in due course took over Civil & Civic in 1961.
The development of the Australian housing market and the involvement of the Dutch authorities in
building activities in Australia were reflected on the emigration cards. First of all, the nature of
housing activities of the consulates became less intensive as is visible in the graph. In some cases,
however, the emigrant cards reveal other echo’s of this housing policy in a rather propagandistic
way: we find cards with remarks that the specific migrant was ‘chosen’ by the Dutch government for
photoshoots in front of his new house during state visits of the Dutch Commissioner for Emigration
to Australia.
10
Reports Huygens ING interns Alison Hurley and Rajae el Morabet Belhaj (both Leiden University). Bredero also built
prefab houses in the Netherlands. (Clarck,2002, 25).

12
Conclusion
The inaccessibility of the migrant registration cards for analysis leads to blind spots in our research.
The counting of script edges on the cards provided the researcher with a crude measure of the
information distribution. The idea was to combine the distribution of information density with a
more traditional sampling and analysis of the contents of the cards and in this way get insight in the
determining factors for the information distribution. This relation proved very hard to establish. The
analysis could only advance after the realization that the registration cards were an instrument of
the registering authorities and their own policies were much more influential than the
characteristics of the migrants. The involvement of policies of emigration support in the use of the
cards provided an additional perspective that enabled us to analyse the distribution of information.
While computer vision was instrumental in measuring the information distribution, further analysis
is impossible without close reading policy files. This illustrates that for historical research computer
assisted methods can provide an important extension of the methodology, but that they are most
effective if combined with established methods.
The whole assessment of the information in the registration cards was established to prevent
selection biases in further sampling migrant lives (and connecting them to dispersed cultural
heritage materials). On basis of that assessment we now know that further sampling will not give a
selection bias towards specific groups of migrants, but we should include the policies of the
emigration authorities in our analysis. Therefore, this underlines Schrover and Moloneys’ statement
to create awareness of the interaction between policy and migrants choices. We can even go one
step ahead: only the whole assessment of information revealed that states do use other, more
hidden or implicit forms of categorization, like the on economic principles based migration schemes.
Thus researchers should also take into account that there are micro forces such as seemingly
‘neutral migration schemes’ that can influence migrant’s life course experiences. Furthermore, in
combining close reading (sample) and distant reading (edging) of the cards we are able to conclude
that the card system was not only a one way monitoring device, but that it was a constant form of
systemic interaction with input and feedback loops between migrants experiences abroad and
policymaking in the homeland.
References
Arthur, P., Ensor, J., Van Faassen, M., Hoekstra, R., Peters, N. (2018),”Migrating people, migrating data. Digital
approaches to migrant heritage”, Journal of the Japanese Association for Digital Humanities, 3-1, 98-113
https://doi.org/10.17928/jjadh.3.1_98 (accessed 12-03-2021)
Clark, L. (2002), Finding a Common Interest. The Story of Dick Dusseldorp and Lend Lease, Cambridge University
Press, Cambridge.
Devereaux, S. (1996). The City and the Sessions Paper:" Public Justice" in London, 1770-1800. Journal of British
Studies, 35(4), 466-503.
Eysbertse, D. and M. (1997, 2006), Where waters meet. Bonegilla, the Dutch migrant experience in Australia.
Melbourne

13
Van Faassen, M. (2014). Polder en emigratie.: Het Nederlandse emigratiebestel in internationaal perspectief
1945-1967. The Hague, Huygens ING
Van Faassen, M. & Hoekstra, R. (in press), ‘Storytelling, Identity and Digitizing Heritage’ in Wergin, C. and
Affeldt, S. Digitising Heritage. TransOceanic Encounters between Australia and Europe. KEMTE, Heidelberg
University Publishing.
Van Faassen, M. & Hoekstra, R, (unpublished paper CGM seminar, 2015)
‘Tijdelijke migratie als bewust beleid. De Nederlands-Australische Youth Programs en Working Holiday
Schemes.’
van Faassen, M., & Oprel, M. (2020). ‘Paper trails to private lives. The performative power of card indexes
through time and space’. In: Nijenhuis, I., van Faassen, M., Sluijter, R., Gijsenbergh, J., & de Jong, W. (Eds.).
(2020). Information and Power in History: Towards a Global Approach. Routledge. 254-274.
Harfield, S., & Prior, J. H. (2010). A bright new suburbia? GJ Dusseldorp and the development of the Kingsdene
Estate. In Green Fields, Brown Fields, New Fields: Australasian Urban History, Planning History Conference.
Faculty of Architecture, Building and Planning, University of Melbourne, Australia.
Hoekstra, R., & Koolen, M. (2019). Data scopes for digital history research. Historical Methods: A Journal of
Quantitative and Interdisciplinary History, 52(2), 79-94.
Peters, N. (2001), Milk and Honey - but No Gold. Postwar Migration to Western Australia, 1945-1964.
University of Western Australia Press.
Schlesinger, L (2018), ‘Lendlease’s Dick Dusseldorp: from Nazi labour camp to listed giant’, Financial Review
(september 2018). https://www.afr.com/property/residential/lendleases-dick-dusseldorp-from-nazi-labour-
camp-to-listed-giant-20180702-h1252g.
Schrover, M. and Moloney, D.M. (2013), Gender, Migration and Categorization. Making distinctions between
migrants in western countries 1945-2010. Amsterdam University Press, Amsterdam
Schrover, M. & Van Faassen, M. (2010) ‘Invisibility and selectivity. Introduction to the special issue on Dutch
overseas migration in the nineteenth and twentieth century’, TSEG 7 (2010) nr. 2 Theme – ‘Invisibility and
selectivity. Special issue on Dutch overseas migration in the nineteenth and twentieth century’ edited by M.
Schrover and M. van Faassen, 3-31.
Shoemaker, R.B. (2008) The Old Bailey proceedings and the representation of crime and criminal justice in
eighteenth-century London. Journal of British Studies, 47 (3). pp. 559-580. https://doi.org/10.1086/587722
available at http://eprints.whiterose.ac.uk/4697/
Walcker-Birckhead, W. (1988, unpublished thesis ANU), Dutch identity and assimilation in Australia: an
interpretative approach.