The current flurry of interest in social mobility seems to have lost sight of part of the motivation that sociologists (for, let's face it, sociologists have been consistently interested in social mobility for much longer than some other disciplines one could mention) had for studying it. Yes, estimating the probability that you can be found in a particular social class given that you started off in another class is useful. And ratios of such conditional probabilities (and even odds) can tell us important things about the "openness" of the society we live in. But there is another way of looking at things.
Unless you have a prospective cohort study to hand you will almost certainly have to make do with a cross-sectional survey with retrospective reports on the social origins of the respondents. Rather naturally this leads you to do the conditioning the other way around and ask: given you are now observed in this class what is the probability that you come from that class? For a previous generation of sociologists this opened their eyes to a different kind of concern. As well as mobility telling you about openness they were interested in the demographic homogeneity or heterogeneity of the social classes. If you read John Goldthorpe's Social Mobility and Class Structure in Modern Britain you will find that a lot of the analysis, discussion and speculation is precisely about this.
One of the questions addressed in that book was whether the working class was becoming more demographically proletarian in its class origins. As rates of mobility out of the working class increased and in as far as the middle-classes were successful in making sure their progeny did not fall out of the middle-class nest it seemed to follow that the working class would increasing consist of people who themselves had working-class origins. What followed from that was speculation about the implications of this for the "classness" of the working class, the implications for class solidarity and for working-class politics.
But how have things actually turned out? The figure below reveals a rather different story.
As in my last post I'm restricted to talking about men because of the lack of useful data about the class mobility on women in the 1970s. On the other hand I can bring the story pretty much up to date by pinching some QLFS data from Lindsey Macmillan and Luke Sibeta's Social Mobility Commission Technical Annex: Quantitative Analysis of Downward Mobility which I recommend for your reading pleasure. As an aside, though the QLFS collects data on social mobility, ONS perversely do not include father's or mother's NS-SEC in the most easily accessible public release version of the data (it isn't possible to construct the NS-SECs from the, much more disclosive, 2 digit SOC codes that they do release).
The figure shows the proportion of men aged 30-59 observed in the working-class (NS-SECs 6 and 7) at three time points: 1972-5 (the pooled England & Wales, Scottish and Northern Irish mobility surveys), 2009-10 (Wave 1 of Understanding Society) and 2014-18 (pooled QLFS) by their class origins ( father or head of household's social class when the respondent was 14 years of age). The pattern these data reveal is very straightforward. Compared to 1972-5 rather than becoming more homogeneously proletarian in class origins the working-class seems to have become more heterogeneous. Take for example NS-SEC 7 routine occupations. In 1972-5 around two-thirds of the men found in this class had fathers who themselves had either routine or semi-routine occupations. In 2009-10 this had dropped to about 55% and by 2014-18 it was less than half. In the seventies less than 5% of men in NS-SEC 7 had father's in professional and managerial occupations, but by 2014-18 just short of one-fifth had so!
If we can trust these numbers then it is no exaggeration to say that the "working-class" ain't what it used to be. Of course those who concern themselves with electoral politics have long noted the decline of the working-class. But what they had in mind was usually its numerical decline not the increased heterogeneity of its class origins. Long ago Frank Parkin turned the conundrum of working-class conservatism on its head by arguing that conservatism should be thought of as the natural political disposition of the working class unless its members are exposed to countervailing institutions, such as trade unions or countervailing experiences such as some kind of demographic or cultural class continuity with previous generations. Maybe it is time for psephologists to take these ideas more seriously.
Footnote: The numerical results look rather solid - large samples, reputable surveys etc. But the sceptic in me does worry a bit about measurement error. Mobility is a difference variable and we all know that difference variables accumulate error like nobody's business. In particular the differences between the 2009-10 and 2014-18 surveys are suspiciously large (though conveniently supportive of my narrative).
Posts
