Introduction

International migration flows to Spain have been intense in the last fifteen years. While at the start of the 1990s, the share of foreigners with respect to the total population was barely 1.5 per cent, by 2000 it was 2.3 per cent, and by 2009 it was 12 per cent. The Spanish immigrant population's rate of increase has been remarkable compared to other countries as well (Cebolla & González-Ferrer 2008: 12). While in 1990, Spain was not even among the twenty countries in the world with the highest immigration rates, by 2005 it was already tenth in the ranking (in absolute figures).

Although the Spanish statistical system had different registers and surveys that could be used to learn more about this new population, there was no single source of specific data with information on the trajectories of immigrants, previous countries where they had lived, housing, social and family networks, and so on. The Instituto Nacional de Estadística (INE, Spanish National Statistics Institute) thus decided to carry out the National Immigrant Survey (ENI) in order to fill this information gap and to satisfy the demands for data on this new population group (see INE 2009).

The literature on ethnic and migration studies played a key role in the ENI's design. Aims, definitions, operational variables, categories and classifications are the cornerstones of any production of social data, including sample surveys. In social surveys, data production is restricted by the social categories and practices specific to any society and historical point in time (Gigerenzer, Switjtink, Porter, Daston, Beatty & Krüger 1989; Desrosières 1993; Porter 1995). This relationship between tools and society applies to the statistical techniques themselves as well, as many historians of statistics have illustrated from the beginning of the Bielefeld School until today (Stigler 1999).

The ENI was based on the idea that immigrants should be defined in objective terms, that is, as people who have a particular experience of migration, rather than in subjective terms, according to self-identification with a cultural or ethnic group. In our view, it would not be suitable to choose an approach to the definition of the immigrant in Spain based on a self defined concept of ‘group membership’.