We have entered the exciting world of “big data”. Currently, however, social scientists are divided about its potential for rigorous research. On the one hand, it promises the prospect of assembling vast data bases offering the prospects of detecting distinctive patterns across time and space and with the capacity to provide granular findings. On the other hand, however, it can be argued that through this very process of abstracting data, context and specificity can be subsumed into superficial accounts which flatter only to deceive.
Clark's high profile and much publicized book is therefore a fascinating test case to reflect on these issues. His premise is a simple one. By examining the patterning of distinctive kinds of surnames and their association with markers of status in different times and places, it is possible to offer new light on the extent and character of social mobility. And so it is that this book ranges across nine nations (Sweden, the United States, England, India, China, Taiwan, Japan, Korea and Chile), goes back in time as far as 1170 (in the English case), and comes up to the present day. The kind of data which is marshalled to extract surnames include student records (such as for Oxford and Cambridge Universities), professional bodies (such as the American Medical Association or of doctors in Bengal), registration data (including death records, as well as population registers such as in the Swedish case), probate records, censuses, the Domesday Book, or political and military records. How can this remarkably wide-ranging assemblage be anything other than a bravura performance?
Clark decides that he has discovered a “social law”, that “there is a universal constant of intergenerational correlation of 0.75 from which deviations are rare and predictable” (p. 12). It has been a long time since it has been thought that laws could be detected in historical analysis, and in this respect the book harks back to an older positivist style of thinking. His fundamental tool for detecting this law is also simple. By extracting unusual surnames which can be shown to occupy high (or low) status at a certain moment, Clark examines how the status of such surnames changes over time. He shows that there is a slow but steady “regression to the mean”. This trend can be interpreted both to indicate a remarkable stability in elite (or “underclass”) reproduction, but it can also be interpreted as indicating that there is always a tendency towards elite (or “underclass”) dilution as well. The glass is either half full or half empty, depending on taste.
The book contains some lovely vignettes. For my taste, the account of the changing distribution of Norman names (which arrived in England after the Conquest) amongst Oxbridge graduates over succeeding centuries is very nicely done. In 1170, such names were over-represented amongst Oxbridge graduates by sixteen times, showing in concrete terms how the Normans dominated this elite educational institution. In the following decades the proportion of Norman names steadily declines, but very slowly. Norman names are still four times over-represented 300 years later. Amazingly, even in the 1980s, Norman names are slightly over-represented. There is still a Norman elite afterlife 900 years after the Conquest.
However, whilst recognizing the book's ambition and verve, ultimately, it can only be seen as a failure. Most fundamentally, Clark's intellectual range does not stray away from his home discipline of economics (and is selective even within that discipline). To say that “[d]iscussion about the mechanisms that drive the inheritance of social mobility has been limited” (p. 126) can only be regarded as an embarrassing admission of ignorance of the range and quality of sociological and historical research which has addressed this very question in recent decades. This is not simply academic point scoring. It matters because research in these disciplines has presented alternative arguments and theorizations which are considerably more sophisticated than Clark's. This can be seen in at least four ways.
Firstly, structural approaches to social mobility are often seen to be more powerful than those which focus on individual attributes (as here). By placing social mobility within the context of structural shifts in the division of labour, analysis need not rely on vignettes but can provide accounts of mobility which are able to examine what proportions of social groups are, for example, self-recruiting, open to the upwardly mobile etc. From Clark's account, however, we learn nothing about how “closed” different elite sectors are. More broadly, a structural approach would also allow Clark's own measures of status, such as belonging to specific professions, to be put in better historical context. The relative standing of doctors, university students, and such like is not historically invariant as Clark implies but itself needs to be contextualized. Oxbridge undergraduates are not an equivalent kind of elite in the twelfth and twentieth centuries.
Secondly, building on this point, sociological and historical accounts have argued that actually there are substantial shifts in the extent of absolute mobility, at least in modern times. It would have been interesting for Clark to have engaged with such an argument (if only to criticize it) but in fact he talks entirely past it. This point is also important in the context of Clark's reliance on male names alone. Whilst recognizing the pragmatic defence of this (given that gender relations are historically variant, and that there are differences in the extent to which, and the means by which, women are associated with elite positions), it becomes more difficult to read off from Clark's findings to make more general claims about universal laws.
Thirdly, Clark's conceptual terminology is very loose. Rather than framing mobility within an analysis of social classes (as in Goldthorpe's class-structural approach) or in a clearly defined occupational hierarchy (as with the Blau-Duncan or the CAMSIS approaches), or within a “social-space” perspective (as with Bourdieu), generic terms such as elites and underclasses are bandied around as if they are self-evident groupings. It is assumed that status can be measured hierarchically on a scale, with no recognition of the argument that it might be better rendered as mobility between groups which are not always hierarchically ordered.
Finally, and most controversially, there is Clark's biological account of patterns of social mobility. As he admits, he has no direct evidence of any biological factor which might allow him to prove this. Instead, his argument is dependent on the logic of his statistical approach where he (understandably) differentiates underlying factors (“competences”) from empirical observations which may be dependent on luck, contingency, etc. Clark then assumes a biological basis to such competences, but this is entirely gratuitous. They could equally well be economic, cultural, or social-capital, in Bourdieu's terms, for instance.
In the latter chapters of the book Clark throws caution to the winds. In a remarkably cavalier way he seizes on examples of Christian, Jewish, and gypsy/traveller experience to argue that the persistence of advantage or disadvantage do not disprove his “social law”. He thus argues that endogamy amongst the Jewish population, combined with high population growth, explains why their relative high status persists and does not regress to the mean. This interpretation, which sums up Jewish history into a three-page synopsis based on a handful of sources, entirely evades how the relationship between Jewish people and other social groups is organized, and notably the role of anti-Semitism. Even more bizarrely, a photograph of two travellers is used to report that “they do not look like people of Indian descent”. It is to be hoped that readers are not offended by this kind of cavalier treatment which appears dismissive of the role of racist forces in history.
Ultimately then, Clark's ambition and confidence proves to be his undoing. Because he puts all his eggs in one basket – of identifying his “social law” – it follows that the book as a whole fails if this law is not convincing. It would be a great shame to disparage the work that has gone into this book. With the remarkable wealth of data at his disposal, Clark could have attempted a more modest but surely more valuable project of elaborating his data sources more contextually, so that they are used to explicate mobility processes in particular nations and times, rather than being yoked to one grand purpose. It would have been fascinating to see much more detail on any of the data bases he has assembled and sensitive reflections on social mobility in any particular context. In fact, we rarely get details of the overall size and descriptive features (most common names at different times, for instance). Perhaps if the big data had been examined with greater attention to detail, a more satisfying book would have been written which could have more clearly identified the power of potential new data sources.
