| Galton Institute Home Page | December 1994 Newsletter Contents | Newsletter Index |
Studies of the relationship between health and social class can be traced back to the pioneering work of Farr, the first Registrar-General in the 1840s. Farr’s own work on mortality combined occupations into categories (for example industrial workers, agricultural workers etc) but this practice was not the norm and it was not until Stevenson’s work in the 1920s that a clear picture emerged of the relationship between social class and mortality. The trend was of increasing mortality from social class I to V. However, not all diseases showed this relationship; for instance deaths from cirrhosis of the liver and appendicitis were highest in social classes I and II and lowest in class V.
It was hoped that the coming of the Welfare State in the post war era would lead to the diminution or disappearance of health inequalities. This belief was supported by the overall decline in mortality and the increase in life expectancy. But social class differences I mortality have, if anything, been increasing. So although there have been changes in disease profiles, with declines in deaths due to tuberculosis and infectious diseases, these have been replaced by deaths due to cancer, circulatory disorders, accidents and violence. The latter causes have not affected all sectors of society equally. Thus an unskilled male worker in the age range 25-44 years is four times more likely to die from lung cancer, stomach cancer and ischaemic heart disease than an equivalent professional male worker.
Data on morbidity differences between social classes was not available on a national level before the second World War. The Survey of Sickness and subsequent surveys paint a similar picture with more days off work, more reported sickness and malaise in the lower social classes.
Social class differences have also been reported for blood group frequencies, height, weight and IQ. In the case of height the mean differences between male adults in social class I and V is about 5-6cm. Given that people migrate between social classes, one would expect the mean height differences to disappear unless the movement was selective. For height there is good evidence to show that people who migrate up the social classes are, on average, taller than the class they are leaving while those migrating downwards are, on average, shorter. Thus this social migration or, to give it its proper name, social mobility is a form of selective migration. The effects of social mobility were discussed in relation to health and IQ.
The final part of the lecture focused on social mobility and IQ and reviewed the often cited paper on this topic by Cyril Burt, published in 1961. Recent reanalysis of the 1961 paper by the speaker suggests that Burt deliberately concealed information (on sample size, where the occupational class data came from and when the information was collected). The speaker showed that Burt provided very poor descriptions of his sources of data and the techniques used. Many of the ambiguities and uncertainties surrounding the paper could have been clarified had the editor of the Journal demanded publication of the raw data with clear and precise accounts of how modifications were carried out. It also demonstrates how easy it is for an editor, or someone very closely involved in the editorial process (as Burt was), to get an inadequate paper published. The speaker concluded that Burt was deliberately deceptive and that there are good reasons to believe that the data were fabricated.