The distinction between western and non-western has been used by CBS as the basis for classification since the 1990s. Initially used in the preparation of the population forecast, it went on to become a commonly used classification in regular CBS publications on other topics.
Handle with care
Regardless of the classification used, care needs to be exercised when formulating statistics that distinguish between people on the basis of their own or their parents’ country of birth. In 2021, CBS held a broad meeting with external experts in the field of migration and integration on the topic of conducting research involving migrants and/or their Netherlands-born children. This resulted in the Assessment Framework for Migration and Integration Statistics (CBS, 2021b), which CBS uses for its publications on this theme. The assessment framework is based on the premise that substantive and methodological considerations should always form the basis for detailing outcomes by origin. In addition, CBS remains alert as to whether differences observed in relation to origin can actually be traced back to differences in, for example, income or educational attainment. If this is the case, it will be reported in descriptive publications.
One or two parents born abroad
By taking into account whether or not someone was born in the Netherlands and whether or not one or both of their parents were born in the Netherlands, the new classification will better reflect the actual differences in starting position that people have from birth. Research has shown, for example, that children born in the Netherlands with one foreign-born parent achieved better results at school than children whose parents were both born abroad (CBS, 2016; CBS, 2012). In part, this difference is thought to be related to the fact that, on average, children with at least one parent born in the Netherlands have better language skills and greater social capital through their parental network.
In recent years, the public debate regarding the use of the terms western and non-western origin as a basis for classification has intensified. Both the broad-based consultations held by CBS in preparing the Assessment Framework for Migration and Integration Statistics and the roundtable discussions with stakeholders (see below) revealed that the distinction between western and non-western is open to various interpretations. The terms western and non-western do not reflect the geographical position of the countries of migration. Not only that, but the diversity within these two groups is considerable while the stakeholders consulted pointed out that using two opposing terms sets up an expectation of opposites in terms of culture, economy and religion. It is also a difficult classification to understand, as it encompasses a number of counter-intuitive elements. Japan and Indonesia, for example, are classified as western, yet South Korea and Malaysia are classified as non-western, even though all four countries are in Asia.
Dialogue with stakeholders
To arrive at a new classification, CBS set up a carefully structured process and entered into a dialogue with both internal stakeholders and experts representing policymakers, the academic world and society at large. The roundtable discussions initiated by CBS (CBS, 2021c) involved a broad representation from organisations including the Netherlands Scientific Council for Government Policy (WRR), the Netherlands Institute for Social Research (SCP), the Netherlands Interdisciplinary Demographic Institute (NIDI), various government departments and universities, consulting firms and interest groups. All parties who took part in these discussions agreed that the distinction between western and non-western should no longer be used. The discussions also led to the formulation of a number of principles for a new classification.
3.1 Principles for a new classification
Based on its consultations, CBS used the following principles to arrive at a new classification for migrants and their children:
- The main categories should be as neutral as possible for use in standard publications, such as StatLine tables. It should also be clear which category individual countries belong to. The classification and the concepts used should therefore not be overly complex or academic.
- The individual categories used should add up to a total to make tables and publications as easy to understand as possible. In practice, this means that ‘Other’ categories should also be displayed. If this is not the case, people may start working on the basis of their own calculations, which can lead to confusion.
- The new classification should stand the test of time to a fair amount. Some classifications are likely to change over time and should therefore be avoided. These include a standard classification based primarily on income and life expectancy as used by the Human Development Index (HDI) (United Nations Development Programme, 2021), on the norms and values measured by the World Values Study (WVS, 2021), or on whether or not a country belongs to a particular international organisation such as the European Union (EU). For example, a statistical classification based on EU membership would have to list Poland and the Czech Republic as non-EU states for the years before 2004 and as EU states thereafter. The reverse would be true in the case of the United Kingdom for statistics before and after 2020, the year in which the UK left the European Union. This could lead to trend breaks, especially in models based on time series. It is perhaps inevitable that the new classification categories will change to some degree over time, but the aim is to ensure that they should be subject to as little change as possible. Another possible disadvantage of classifications based on substantive criteria (such as the HDI or WVS) is the risk of pre-empting a potential research question or measure of outcome.
- It is important for the new classification to consist of several levels. This means that even when the number of observations is relatively small, as is the case with survey-based statistics, it will still be possible to make distinctions by country of origin at a higher level. By the same token, statistics with a large number of observations, as is often the case with registration-based statistics, can be specified at a more detailed level. In general, the more detailed the classification by origin, the less heterogeneity there will be within individual categories. This ensures greater accuracy in reflecting the diversity within the group of migrants and their children.
- The aim is for the main classification to be used for standard publications, such as StatLine tables. However, this classification need not be used for all statistics and publications. If warranted by the research question or more appropriate to the type of study being carried out, another classification may be used. This might be a classification by wealth, migration motive or linguistic affiliation.