Scottish Longitudinal Study
Development & Support Unit

Current Projects

Project Title:

Health and social mobility between 1991 and 2001

Project Number:



Gillian Raab (University of St Andrews)
Paul Boyle (University of St Andrews)
Lin Hattersley (GROS)

Start Date:

Approved on 21-02-2007


The ONS LS data have been widely used to investigate health- related social mobility. Most relevant to this application is the work of Bartley and Plewis (1997) which investigated the relationships between limiting long term illness (LLTI) in 1991 and related it to changes in social class for men in the labour force and captured in the censuses of 1971,81 and 91. They found differentials in health for those who moved between social classes in 1971 and 1981, but the net effect on the marginal relationship between social class and health at each of the three years was very small.

Boyle and Norman (unpublished personal communication) used the LS data for 1971 and 1991 to investigate socio-spatial mobility and health. In their analyses all LS members present at these two censuses were used and the deprivation quintile of the ward of residence at the two time points was related to LLTI in 1991. In contrast to the individual analyses they found that the marginal gradient of health by deprivation had widened between 1971 and 1991.

The present study will allow us to investigate three possible reasons for the discrepancies between the individual and area-level analyses.

  • Is the difference between the two analyses due to transitions out of the workforce? This is plausible because of the very well established link between unemployment and health.
  • The age distributions of subjects differed between the two analyses. Could this have explained the differences in terms of entries and exits to the populations. relationship between long-term illness and social class. We will use head of household social class for young people who are not yet in the workforce.
  • The Bartley and Plewis analyses only used men. Are patterns different for women?

In addition, and as a starting point, we will replicate the LS results from the SLS data. This will be a sub-analysis restricted to men in the workforce at the two periods’. Because LLTI was measured at both 1991 and 2001 we will be able to relate both of these to socio-economic transitions. Finally we will look at the effect of population change on socio-economic differentials in health. Unlike Bartley and Plewis and Boyle and Norman, we will be able to consider the mobility patterns of those with illness in 1991 and compare this with health status in 2001. In the previous analyses the health outcome was only measured at the end of the period.

The study will address the following questions:

  • How is long term illness (LTI), at the 2001 census, related to social mobility between the 1991 and 2001 censuses.?
  • How does the relationship between LTI and social class differ between the 1991 and 2001 censuses?
  • What are the components of the change (or lack of change) found in 2 in terms of entries and exits to the population, entries and exits to the work-force and transitions between social classes?
  • How does the relationship between LTI and social mobility vary by age and sex?
  • How do these results compare with past work using the ONS LS.


Bartley M and Plewis I (1997) Does Health-selective mobility account for socio-economic differences in Health? Evidence from the England and Wales 1971 to 1991. Journal of Health and Social Behaviour 38, 376-86.

Related Outputs (viewable on CALLS Hub):

  • Sorry, no related items were available.

Explore the variables held in the SLS data dictionary.

Recent News

Upcoming Events

Events feed unavailable

Latest SLS Tweets