Jarvis H. Housing, labour markets and household structure: questioning the role of secondary data analysis in sustaining the polarization debate. Regional Studies 1997; 31 (5): 521-531.

This paper presents a critique of the polarization thesis as it is popularly conceived in Britain. The apparent phenomenon of an increasing divergence of the population into extremes of multi-earner and no-earner households is typically demonstrated using secondary data sources. Consequently, polarization is explained in terms of household employment with regard to earners in a way which overlooks sources of division which emanate from within a variety of household structures. It is posited that the appearance of a polarization of household prospects is manifested, at least in part, by the conflation of social and demographic regional profiles, which combines the observation of two discrete causal mechanisms into one. To illustrate this point, data is presented h-om the 1991 Census of Population Sample of Anonymized Records (SARs) which identifies a pattern of disaggregated household employment compositions (by gender and hours worked) for a specific population of 'nuclear family' households. It is suggested that once the regional effects of flexible employment practices and sectoral restructuring are viewed through the lens of detailed household gender divisions of labour, beyond a one-earner versus multi-earner division, to focus on who is earning, how securely and with what effect on household practices, then questions can be asked about the social implications of global economic restructuring.