« The "Creative Class" | Gene Expression Front Page | Why Mars? »
January 18, 2004

Intelligence and social mobility

There has long been debate among sociologists and psychologists over the extent to which social status reflects personal abilities such as intelligence and ambition. Roughly speaking, ‘lefties’ believe that social status is largely inherited, and has little to do with personal merit, while ‘righties’ believe that in modern open societies, individuals usually reach the social level they deserve on their personal merits.

There are two recent British studies that tend to support the ‘meritocratic’ model:

Peter Saunders: ‘Reflections on the meritocracy debate in Britain: a response to Richard Breen and John Goldthorpe’ British Journal of Sociology 53, (2002), 559-74.

[This is the latest in a series of papers by Saunders and colleagues, but can be profitably read on its own.]

Daniel Nettle: ‘Intelligence and class mobility in the British population’, British Journal of Psychology, 94 (2003), 551-61.

Both studies use the very large sample of the National Child Development Study, which follows up the fortunes of an entire cohort of children born in Britain in one week in 1958 (and therefore now in their mid-40s). Saunders and Nettle both show that intelligence (measured by IQ tests in childhood) is the most important single factor predicting social mobility and status in adulthood.

Oddly, Nettle does not cite any of the previous work by Saunders. Maybe psychologists don’t read sociologists? And neither of them cites Cyril Burt’s controversial 1961 paper ‘Intelligence and social mobility’. Burt reached similar conclusions, but his empirical data are generally considered unreliable if not fraudulent.

A correlation between IQ and social mobility does not in itself say anything about the genetic or environmental origin of IQ differences between social classes. If social status reflects individual merits, then the same pattern of stability and change would be observed if the parent/offspring correlation is the same, whether the reasons for the correlation are genetic, environmental, or any combination of the two.

However, it is interesting to note that levels of social mobility seem to be broadly similar across long periods of time. I have read several studies of social mobility in Britain from the mid-C19 onwards, and mobility varies much less than you might expect. If social classes (based on occupation or income) are divided into three broad levels, then between a third and a half of men end up in a different level from that of their fathers, and have done as far back as the records go. The idea that in the ‘old days’ there was little social mobility is a complete myth.

Social mobility also seems to be much the same in all industrialised countries. Sociogists used to believe (based more on assumptions than evidence) that social mobility must be higher in ‘meritocratic’ America than in ‘hierarchical’ Europe, but the differences in reality are not very great.

This relative (though not perfect) constancy might be thought to give some support to a genetic interpretation of social mobility. On a multifactorial genetic interpretation it is what you would expect (as Burt argued) whereas on an environmental interpretation it is a bit of a puzzle.

Posted by David B at 05:32 AM