Deriving trends in life expectancy by the National Statistics Socioeconomic Classification using the ONS Longitudinal Study
Johnson, B. (2011) Health Statistics Quarterly, 49(Spring), 9-52. 22 February 2011. [ONS LS]
Background Health inequalities among socio-economic groups are well documented. One of the measures used to track inequalities over time is the series ‘Trends in life expectancy by social class, 1972–2005’, on the Office for National Statistics website. In 2001 the National Statistics Socio-economic Classification (NS-SEC), replaced Registrar General’s social class (RGSC) for the purposes of official statistics. This paper describes the challenges involved in producing an analogous series of trends in life expectancy by NS-SEC to that by RGSC, the approach adopted, and publishes the first results of the new series.
Methods NS-SEC was devised in the 1990s and introduced in 2001. Like RGSC, it is an occupation-based measure. In order to produce a series of trends over more than 20 years based on NS-SEC, it is necessary to classify people according to NS-SEC based on their occupation at the 1981 and 1991 Censuses and then to measure subsequent mortality rates for different classes. The 1981 Census preceded the construction of the NS-SEC classification system by nearly 20 years, and there was no recognised way of classifying 1981 Census respondents by NS-SEC. This paper describes how an approximation to allow such a classification was derived. The ONS Longitudinal Study was used to provide the data from which mortality and survival rates by NS-SEC class could then be estimated.
Results The results are presented in terms of life expectancy at birth and at age 65 by five-year calendar periods, from 1982–86 to 2002–06. A social gradient was found using NS-SEC, similar to the one found using RGSC. For most classes for all periods studied, life expectancy improved for both males and females but inequalities persisted between classes. There was a difference of around six years for males between the most and least advantaged classes in expectation of life at birth and about four years for females in the ￼￼￼Office for National Statistics 1 period 2002–06. The estimates suggested a widening of inequalities over the study period for men, which appeared to end after 2001. For women, no overall trend could be detected, but there were no signs of any narrowing of the gap in the most recent period.
Conclusions NS-SEC can be used to provide medium-term trends in life expectancy by occupation based class, which will be capable of extension over time, although certain approximations are necessary. It is important that work should continue on investigating other means of classification, particularly for women, for example based on educational attainment and on household rather than individual-based measures.