Detailed employment history and subsequent health outcomes: an analysis using the ONS Longitudinal Study and the recently linked claimant Count data
Dermot O'Reilly and Michael Rosato, Queen's University Belfast
[Project number 40012]
Some geographically or economically grounded analyses of health outcomes in the United Kingdom (UK) stress a link between a declining or moribund local economy and an increased prevalence of adverse health outcomes, a recurring theme in analyses which comment on the decline of the traditional UK industrial base (Beatty et al; Rosato and O'Reilly). This study will utilise the recent linkage of the ONS Longitudinal Study and Claimant Count data from the Department of Work and Pensions to study the relationship between detailed employment history (from 1994 to 2004) and indicators of health such as mortality levels (from 1994 to 2004), or those derived from the 2001 Census returns (the responses to the limiting long term illness and general health questions, and economic inactivity due to chronic ill health from the economic activity question). The analysis will examine this primary concern in relation to demographic factors (such as age, sex, marital status and ethnicity); indicators of personal or household affluence (housing tenure, car access); area socio-economic indicators (such as the Carstairs Index, or ONS Area Classification indices) and geographic factors such as Standard Region of England and Wales.
The analysis would be carried out with Stata, using logistic regression techniques for the analysis of employment history and the Census based indicators, and both Cox proportional hazards models for the analysis of employment history and subsequent mortality, and (possibly) analysis of survival after unemployment. The actual definition and derivation of unemployment status would be dependent on the granularity of the Claimant Count data but we anticipate that the detail will allow much flexibility in this.
The longitudinal study (LS), based on a continuing, representative, one percent sample of the population of England and Wales, is a high quality multi-cohort research database with ongoing linkage of routinely collected vital statistics data with the census returns from 1971 to 2001. This recent inclusion of the Claimant Count data, with its attendant detail, offers a wonderful opportunity to examine the effects of employment history and health outcomes, in relation to socio-demographic and socio-economic circumstances.
BeattyC, Fothergill S, MacMillan R. A theory of employment, unemployment and Sickness. Regional Studies 2000; 34:617-630
Rosato M, O'Reilly D. Should uptake of state benefits be used as indicators
of need and disadvantage? Health and Social Care in the community (2006)