Coast, Ernestina, Fanghanel, Alex, Lelievre, Eva and Randall, Sara (2013) Counting the population or describing society?: a comparison of British and French censuses. In: Chaire Quetelet 2013, 12-15th November 2013, Research Centre in Population and Societies, Catholic University of Louvain, Louvain-la-Neuve, Belgium. (Unpublished)
Household units are integral for the planning and use of resources by governments, and censuses are an important source of household data in most countries. Some academic consideration has been given to the way in which ‘household’ as a conceptual, statistical and analytical category has been used in social research (Bauman 1999; Beaman and Dillon 2010; Casimir and Tobi 2011; Randall, Coast et al. 2011) particularly in countries of the global South but there remain few examinations of the ways in which the concept of household itself has evolved in Europe, and the implications of this for social science analyses (Hoffmeyer-Zlotnik and Warner 2008). Census data are used by politicians, policy advisors, statisticians, lobby groups and national and international organisations for the planning and allocation of resources, welfare and support. Data collected at household level in population censuses are, therefore – in principle – significant for the promotion of socio-economic health and well-being of people at all levels of society. In order that census data are comparable across time and space (historically and geographically) it is important that we understand if, and how, concepts and definitions change between censuses. This paper compares the conceptualisation of the household via evolving census definitions in England and Wales with a very different statistical tradition found in France. This comparative review of the way in which ‘households’ have been defined in censuses will be used to identify changes in thinking about households and reveal what census designers think is important in the way that a household should be defined – what their preoccupations and priorities are and the extent to which definitions of households can, or should seek to, reflect reality. Insights from the analysis are then situated within the context of goals of harmonisation of statistical tools and concepts. We use two different research methods: first, a review of census documentation (including the census schedules, enumerators’ manuals, training materials, allied and associated paperwork, etc.). The household definitions from census documents are analysed longitudinally to examine how they have evolved since the 1960s and comparatively to examine similarities and differences between countries. Our second method comprises analysis of a series of in-depth interviews (approximately 30 per country) with key individuals situated at different places on the chain of data production (census designers, interviewers, statisticians, policy makers, diverse data users and academics) and were oriented around respondents’ roles in the collection and/or use of household data. Discussion particularly focused on the way in which the “household” is defined and used in censuses. Recorded interviews were transcribed verbatim and coded using Nvivo. The data and analyses are explicitly comparative, both over time (1960s-present) and space (E&W and France). We highlight significant, but subtle, changes in census household definitions over the past half century in England & Wales. Analysis and interpretation of these definitions suggests three key analytic categories, present to varying degrees, in census definitions of the household: sharing of space; sharing of food; and, administrative linkages. In our key informant interviews, several themes emerge. National statistical offices in both settings (ONS and INSEE) influence change (or its absence) in very different ways. Second, there is divergence in the extent to which definitional changes either reflect, or respond to, broader societal changes such as changing living arrangements and family forms. Finally, changing data collection modes (e.g.: online completion of 2011 census in England Wales), might interact with the ways in which households are operationalised. These parallel motivations; reflecting socio-economic changes in the organisation of daily living practices and making the definition easier to interpret by lay-people in order to get data which are as accurate as possible about households (particularly several households within one dwelling) have all influenced the evolution of the household definition over the past 50 years. Whilst it might be expected that household definitions ought to capture the changing ways we live now, the pragmatics of using the definitions dominates.
Actions (login required)
|Record administration - authorised staff only|