[Validity of the occupation as an indicator of social class, according to the British Registrar General classification]

Measuring social class is important for evaluating its influence on health status and on the access to health services. This paper is aimed to assess the construct validity of the classification of social class based on the individual's occupation.

Data come from the Barcelona Health Interview Survey of 1986. In the survey, 2,205 households and 6,894 individuals participated (82% of the eligible households and 84% of the eligible individuals). Information was gathered on, among other issues, the following socioeconomic variables: annual family income, lack of some household services, property of the household, education level and occupation of the individuals, as well as their occupational situation and labor relationship. Data about the value of the household was collected from the City Council census (catastro). Social class was assigned using the individual's occupation or, if none, the head of the household's occupation in the following groups: I, II, III, IVa, IVb, V, and "Not Classified". The association between social class and socioeconomic variables was analyzed using: ANOVA for the comparisons of continuous variables, Chi-squared test for categorical variables, Spearman correlation coefficients and discriminant analysis.

A total of 3,357 individuals reported an occupation. "Employed in administrative services" was the most common (14.3%). Social class was "Not classified" in 6.4% of the individuals. All socio-economic variables showed statistically significant differences, following a predicted pattern: better indicators for more favoured social classes. Social class showed a moderate to high correlation with education level (r = 0.57) and somewhat lower with the other variables. Variance in socioeconomic variables explained by social class was higher than the 95% (p < 0.001).

The pattern of relationships between socio-economics variables was intense, monotone and consistent, suggesting that the occupation is a valid and feasible of social class. Routinely including occupation in health information systems should allow to monitor inequalities in health in Spain.