CeLSIUS logo


Census logo

Introduction

A central social policy aim of successive governments has been the extension of opportunity for social advancement to all children, via universal and free education (see e.g. 2004 paper by Prime Minister's Strategy Unit). However, many studies over the past sixty years have demonstrated the tendency for children to grow up in the same social stratum as that occupied by their parents, with similar occupational status and economic means (e.g. Heath and Payne, 2000). The Longitudinal Study can be used to provide valuable background information relevant to assessment and formulation of such policies. Here we look at intergenerational continuities in social class focusing on LS members who were children in 1971 and comparing them with their parents.

Assessment of a person's social class is usually based on the nature of his/her usual employment, although this can be misleading in the case of people whose livelihood does not come from their job, such as someone bringing up young children in a family where the other partner is the main breadwinner. For further consideration of how social class is measured, see the Training Module on Socio-Economic Indicators.

The first measurement taken is the social class of the LS member's parents in 1971, for LS members who are then aged six to fifteen. If the LS member lives with two parents in different social classes, the higher of the two social classes is assigned; if the LS member lives with only one parent, only this parent's class is used. This is compared with the LS member's own social class in 2001, thirty years later. This refers to their current main job, or for those who have none it distinguishes between unemployment and other non-working states. It is also possible to assess whether children who lived with two parents in 1971 show advantages by comparison with those who lived with only one, and whether boys and girls experience different outcomes.

Further analysis could examine these associations for different population subgroups, for example by geographical area, area type or area deprivation, all possible using the LS.