Choose your ancestors carefully

Gregory Clark’s [amazon_link id=”0691162549″ target=”_blank” ]The Son Also Rises: Surnames and the History of Social Mobility[/amazon_link] is one of the most genuinely thought-provoking books I’ve read for quite a while. It’s the fruit of obviously many hours of archival and other research, looking at the lessons that can be drawn from using rare surnames to track directly what happens to status through the generations. The results are startling.

[amazon_image id=”0691162549″ link=”true” target=”_blank” size=”medium” ]The Son Also Rises: Surnames and the History of Social Mobility (The Princeton Economic History of the Western World)[/amazon_image]

Our received wisdom is that social mobility, measured by the persistence of income (usually) or perhaps wealth or education over time, was sluggish, improved in the 20th century, and has perhaps diminished again. The surname evidence indicates very low intergenerational mobility, in many countries and over many centuries. There is a slow, slow regression to the mean in status, but no evidence of any change in trend, even through major political and social changes, such as the extension of the franchise, the welfare state or even the Cultural Revolution in China. Clark writes:

“Mobility is consistent across generations. Although it may take 10 or 15 generations, social mobility may eventually erase most echoes of initial advantage or want. Counter-intuitively, the arrival of free public education in the 19th century and the reduction of nepotism in government, education and private firms have not increased social mobility. Nor is there any sign that modern economic growth has done so.”

So, to give some examples of the data sets used, the records of the Swedish Bar Association show that names associated with the nobility from the 17th century appeared six times as often in 2012 as they did in the general population. “The 18th century elite in Sweden have persisted to the present day as a relatively privileged group.” Sweden! At the present rate of decline of greater-than-average representation of their names among US doctors, “It will be 300 years before the Ashkenazi Jewish population of the United States ceases to be under-represented among physicians.”

In the UK where – notoriously – it helps to have attended Eton to get into the government, the surnames of the mediaeval elite were still eight times over-represented (compared to the average name) at Oxford and Cambridge in 1980. Someone with an unusual  ‘wealthy’ surname was about 50  times more likely to attend Oxford or Cambridge in the mid-19th century, and still 8 times more likely in 2010. “In terms of social mobility, what did the Scientific Revolution, the Enlightenment and the Industrial Revolution achieve? Very little.” Surnames associated with the invading elite after the Norman Conquest in 1066 are still heavily over-represented in Parliament and in the British armed forces.

Even in China after Communism and the Cultural Revolution, Qing elite surnames are found more frequently than the three most common Chinese surnames among elites such as professors, government officials and the chairs of company boards.

There are many examples. The bulk of the book presents the data sets and the persistence of the over-representation of certain surnames among high status groups. The rate of ‘social entropy’ is both high and astonishingly consistent across space and time, at 0.7-0.8.

Why are the surname estimates of social mobility so much lower than conventional estimates looking at the correlation of (say) income over time? Clark suggests that this is because a single indicator like income is an imperfect proxy for social status, which will depend on a number of attributes. The resulting error in measuring underlying status will bias downward the estimated persistence (or bias upward the estimate of mobility). Explaining why mobility is more or less constant and universal is harder. The book tentatively suggests that status is an inherited trait in the sense that an initial combination of talent and luck is sustained over the generations by assortative marriage. “The law of social mobility tends to produce a long arc of privilege or want for those who end up at the extremes of the status distribution.”

If correct, the conclusion seems pretty disturbing. But Clark insists that there is no reason for fatalism. Surname is not destiny. Slow social mobility does not imply a society rife with nepotism or corruption. Individuals’ efforts can act on the opportunities they get and the elements of luck that affect all lives. “Even with an inter-generational correlation of 0.75, more than two-fifths of variation in outcomes in generalized social status is still unpredictable.” The chance of moving from the bottom to the top of the pecking order in one generation is negligible, but there is plenty of scope for upward (or downward) movement.

There are both personal and policy conclusions, however. For anxious parents, the message is that much of what you’ve done for your children is genetic, so by all means give them every encouragement but don’t fret about the flash cards and expensive schools. Most of the old Etonians who thrive in modern Britain would have done just as well without costing their families the fees (£33,000 a year plus extras for seven years).

In terms of public policy, the message is avoid institutions and structures that amplify inequality. “If we cannot change the heritable advantages and disadvantages of families in the economic and social world, we should at least mitigate the consequences of these differences.” This means redistributing income via the tax system, providing universal healthcare and education, having a higher education system that avoids sorting by ability (more like the German than the US or UK systems), and so on. Sweden again. If you want to do well in modern Sweden, it helps to have had an ancestor who was ennobled in 1700, but if you don’t have that inheritance, you won’t be as far behind your country’s elite as a scion of the English working class will be behind ours.

It seems hard to quarrel with the data assembled and presented in this book. I certainly wouldn’t argue with the policy conclusions either – why would you want to amplify inequalities anyway?

Still, this is a troubling argument because it does seem to imply a necessary fatalism about our ability to shape social and economic outcomes. I’ve always been greatly encouraged by the work of [amazon_link id=”0262019132″ target=”_blank” ]Jim Heckman[/amazon_link] and others suggesting that early interventions in the lives of disadvantaged children can make a big difference to their lifetime success – partly because of experience in a local primary school making me so aware of how unequal children’s capabilities and prospects are at an early age. Clark challenges the Heckman-type results in one section, presenting some evidence that there is a large effect but temporary, and that it fades rapidly in adulthood.

So I hope [amazon_link id=”0691162549″ target=”_blank” ]The Son Also Rises[/amazon_link] stimulates a lot of further debate and research. It’s an important book, and anybody at all interested in inequality and the kind of society we have should read it. Although there are many charts and a few equations, it’s very well written, and full of fascinating historical detail as well.

9 thoughts on “Choose your ancestors carefully

  1. I just studied a macroeconomic model by Storesletten, Telmer & Yaronc (2004) “Consumption and risk sharing over the life cycle”, where they ask how important the initial conditions, i.e. who your parents are, are in determining the variation in lifetime welfare (measure by consumption). The authors find that this fixed heterogeneity accounts for 47% of the dispersion of welfare, while the variance of the present value of lifetime earnings accounts for the remaining 53%. In other words, 47% of the inequality of lifetime welfare across individuals is resolved before they start their working lives, and 53% is resolved once they start earning and facing the uncertainty of income streams. As suggested, the policy implications are that, in reducing inequality, it is equally important to place interventions in the early stages of life (school children) and to promote the role of financial markets in insuring against the variability of income, in order to give everybody the opportunity to accumulate assets and smooth their consumption level (again, welfare) over time.

    I think this paper converges with “The Son Also Rises” in the broad topic of inequality.
    However, the theoretical and practical differences between social status inequality, income inequality and consumption inequality, are not subtleties. It turns out that we could be living in societies where social status is strongly linked to wealth accumulation, but it is not the same. Furthermore, the gap in consumption is not commensurate with the existent large wealth dispersion, but it is actually a lot smaller (as has been empirically shown in many papers, such as Blundell & Preston, 1998). So, at the end, touching the philosophical ground, we could ask what is it that we should be really caring about? In my opinion, the three of them are important, but guaranteeing fair living standards (not a minimum survival level) for everybody should be a priority. Thereafter, policies should make sure that the social gap between family groups or income groups does not become harmful, say, in the form of social fragmentation or other issues that erode social cohesion, such as a race of rats issue.

    Thanks for the review Diane! it was very interesting and insightful! I hope I can get to read the book sooner than later.

    • A very thoughtful response – thankyou. I agree with your suggestion that we should give careful thought to the separate components of status, and tailor policy correspondingly.

  2. I’m sorry but I am rather puzzled by this:

    There are both personal and policy conclusions, however. For anxious parents, the message is that much of what you’ve done for your children is genetic, so by all means give them every encouragement but don’t fret about the flash cards and expensive schools. Most of the old Etonians who thrive in modern Britain would have done just as well without costing their families the fees (£33,000 a year plus extras for seven years).

    So basically economic, social and cultural capital is largely over-ridden by marriage between so called intelligent people? And all governments can do is mitigate the after effects of this natural divide? This sounds so sociologically naive I really don’t know what to say. Someone should tell the Sutton Trust and JRF to stop all their research and simply read Clark’s book.

    • Clark indicates that there is still more than two-fifths of the variation in outcomes to play for with interventions of the kind I – and presumably you – would favour. But it seems equally sensible to advocate also policies that mitigate the results of unequal status. Is that so puzzling?

      • Hi Diane,

        I am all for dealing with equality of outcome vis-à-vis equality of opportunity. In fact my point is that my from perspective, which you rightly identify with elements of your own, inequality is basically the crux of social mobility problem.

        My problem is with the quoted paragraph is that the variation is pushed to the side and there emerges, whether intended or unintended, a certain form genetic determinism. I haven’t read the book yet but it is not like Clark does not have some form on this stuff. I found A Farewell to Alms also troubling on this topic.

        Basically, I think inequality and discrimination is the main problem and not simply something stuck onto the natural play of genetic inheritance. The irony is of course is that Clark and I both end up at the same solutions anyway but I still find his arguments disturbing.

        • Disturbing, I agree. Plausible? Maybe – that outcomes are about 50:50 nature/nurture (ok, 60:40 in this book) has other people in support. You started out with puzzling, though…. not sure I like it, but it seems a clear argument. And I do think it important to consider this kind of evidence even if it doesn’t change your views about what to do.

          • Hi Diane,

            Sorry, I really am not trying to pick a fight here and should have made myself a little clearer. I was and still am ‘puzzled’ by the conclusions because I don’t think the surname data can really support the idea of inheritance being used here.

            The democratic opening up society is barely 50 to 60 years old in a lot of countries. and so called ‘reduction of nepotism in government, education and private firms’ is something I would take with a pinch of salt. Basically we haven’t seen what post-war reforms vis-à-vis the congealed nature of social mobility beforehand have really achieved/can achieve. African Americans, for example, have barely had 50 years of formal political equality. These reforms have not been around for 10 to 15 generations.

            I am quite willing to change my mind upon reading the book but from your review Clark seems to make the argument that groups who have low social mobility are basically there mainly due to inherited traits. For example, the inability of certain minority groups such as poor African Americans to become upwardly mobile is largely due the stock of the people rather than the historic forms of discrimination and inequality they have and continue to face. I find that really hard to sustain with what I take to the be the evidence on show.

          • If that’s how the argument comes across, I misrepresented it badly. He absolutely acknowledges that initial social conditions make a big difference. The surname data is about the (slow) speed of social mobility, and I think it’s very persuasive. The results are consistent across many countries and time periods, including China before and after the Cultural Revolution for example. So as I’ve obviously done a bad job with my summary, you’d maybe better read the book! I do think recognising how slowly social status changes has important policy conclusions – as this interview with Clark in Prospect suggests:

Comments are closed.