There are ‘lies, damned lies, and statistics’, as the old adage goes, yet the importance of statistics for shaping policy at national and European level is increasing. More and more data on populations is collected and is of ever more salience in policy debates. As new member states join the EU, they are called upon to contribute to Eurostat – the European statistical office of the European Union, which provides statistics at European level that enable comparisons between countries and regions. As its website states: ‘This is a key task. Democratic societies do not function properly without a solid basis of reliable and objective statistics.’ With the increasing value placed on data collection, and the growth of migration control as a policy topic, as part of work package 10 of bEU, Barriers to EU citizenship: insiders and outsiders, we are exploring how migrants are captured in datasets. We are looking at both national datasets, in the form of national labour force surveys (LFS) of the partners in the project (the UK, the Netherlands, Ireland, Spain and Croatia), as well as Eurostat harmonised data sources: the European Union Labour Force Survey and European Union Statistics on Income and Living Conditions.
One of the problems in exploring migrants in datasets stems from how they are conceptualised, which may differ across and within countries. In the UK, for example, migrants can be defined in at least three different ways: by place of birth (i.e. foreign-born), nationality (i.e. foreign citizens), and length of stay in the UK. As a Migration Observatory briefing on the topic sets out, this results in numerous conceptual difficulties and implications. Further, the datasets were not developed for research purposes, or for the purposes of exploring migrants in the population; thus using them to draw conclusions about migrants can be problematic. National LFS are sample surveys, and only cover ‘households’. Therefore, certain groups are excluded: those living in communal establishments such as students in dormitories or people living in temporary accommodation are not covered. There can also be issues with small sample size and difficulties capturing more recent migrants who do not have the language ability to respond to surveys. Thus the LFS tends to underestimate proportions of migrants who in some ways differ from the established population, that is, categories which are most likely to be vulnerable in the labour market are less likely to be included (Jayaweera & Anderson, 2008). At the European level, Eurostat also acknowledges that there are many caveats regarding the reliability and comparability of the data. Indeed, the migration data reported by the individual countries in Eurostat are not completely comparable (neither between countries nor over time).
The measurement of numbers, then, could be seen as an ‘illusion of bureaucratic control’ (Appadurai, 1996 in Sussman, 2004). Look only to the recently reported net migration statistics in the UK, which far exceeded the Tory party’s pledge to curb numbers. According to the Migration Observatory, net migration only relatively recently became the yardstick by which migration is measured, entering into public discourse in 2010 when limiting net migration became a Conservative party policy. Yet this has become a tool by which migration and migrants come to be defined, as well as a symbol of state control. As scholars have identified (Hacking, 1982), data about populations, accurate or inaccurate, is “seldom effective in controlling or altering the populations of study in the way intended” but rather through their categorization “render rigid new conceptualizations of the human being” (in Sussman, 2004: 102).
Compare, for example, the new European member Croatia with the Netherlands, a country with a long history of immigration and colonial ties, and a large migrant population. In the Netherlands, populations are aggregated into autochthone (native Dutch, officially defined as persons with both parents born in the Netherlands, regardless of where he/she was born) and various sub-categories of allochthone (non-native, officially defined as persons with at least one parent born abroad) populations. The category of allochthones is split into western and non-western allochthones. Western allochthones are those who, or whose parents, come from Europe (excluding Turkey), North America, Oceania, Indonesia or Japan. Statistics Netherlands explains that allochthones coming from Indonesia and Japan are classified as “Western” due to their socio-economic and socio-cultural position. The category encompasses persons born in the Dutch East Indies and employees of Japanese companies and their family members. Western-allochthones can be further split into persons coming from one of the A12 countries and the rest of the western-countries. Non-western allochthones are those who, or whose parents, come from Africa, Latin America, Asia (excl. Indonesia and Japan) and Turkey.
In contrast, Croatia, as a new EU state, finds itself in the position of being more of a transit country in terms of migration, and collects next to no data on ‘migrants’. The majority of foreign nationals in Croatia are from one of the neighbouring ex-Yugoslav countries (46% of all foreign citizens in Croatia are citizens of either Bosnia and Herzegovina, Serbia, Slovenia or Montenegro), and as such, they or their parents were effectively nationals prior to the dissolution of the common state in the early 1990s. This prevalence of neighbouring-country migration is indicative of the relatively low interest in Croatia as a migration destination. There are nonetheless a number of other foreign nationals residing in Croatia, but, according to our partners, as yet Croatian public bodies have not seen the need to monitor foreign nationals’ economic activity and benefit take-up, and ‘migrants’ are thus mainly absent from the data. It will be interesting to see whether/how this changes with Croatia’s membership of the EU.
Thus the ‘lie’ of the statistic, as is known, but on occasion disregarded, is that statistical processes are not necessarily the neutral and benign form of enumeration they can be taken to be (Sussman, 2004), but can contribute to processes of ‘othering’ and normalised ideas of in/exclusion. One needs to look behind the numbers, at the framing of concepts embedded in statistical systems, and remember the caveats that warn of unproblematically accepting the numbers game. We will be publishing a report later in the year on migrants in datasets, which will be publicly available on the bEU citizens website: http://beucitizen.eu/
Sussman, C. (2004) ‘The Colonial Afterlife of Political Arithmetic: Swift, Demography, and Mobile Populations’ Cultural Critique, 56, Winter 2004, pp. 96-126
Jayaweera, H. and Anderson, B. (2008) Migrant Workers and Vulnerable Employment: A Review of Existing Data. TUC Commission on Vulnerable Employment