Written by Noah Carl.
Michael was born into a poor family in England in 1791. One of four children, he didn’t get enough to eat because his father, a blacksmith, was often ill and couldn’t always work. He received only a rudimentary education at Sunday school, but proceeded to educate himself after becoming a bookbinder’s apprentice at 14.
Michael Faraday, to use his full name, went on to revolutionise the field of electromagnetism and became one of the most famous scientists of the 19th century. Einstein reportedly kept a portrait of him in his office.
Interestingly, the man always remained true to his humble, Christian origins. He turned down a knighthood on religious grounds, preferring to remain “plain Michael Faraday to the last”. During the Crimean War, he refused to help build chemical weapons because he considered them unethical. And he declined to be buried at Westminster Abbey alongside Sir Isaac Newton and other luminaries. “I have always loved science more than money,” he once told a publisher.
Faraday is one of several historical figures who achieved extraordinary success despite a very modest upbringing. Other examples include Benjamin Franklin (who only attended school for two years), Andrew Carnegie (who grew up in a one-room house) and Abraham Lincoln (who also grew up in a one-room house). Indeed, all these men illustrate the principle that being born into great wealth isn’t necessary to get ahead in life.
Hold your horses. While being born into great wealth may not be strictly necessary, some will cry, it surely matters a great deal. Individuals born into great wealth usually do get ahead; those born into modest circumstances usually don’t. And a handful of exceptions like Michael Faraday can scarcely overturn this rule. Family wealth actually is what explains success, isn’t it?
Back in 2023, Gregory Clark published a landmark paper ‘The inheritance of social status: England, 1600 to 2022’. Analysing a database of more than 400,000 people, he showed that correlations in social status for different pairs of relatives (fathers and sons, brothers etc.) can be predicted almost perfectly from their genetic relatedness—assuming strong assortative mating on the underlying traits. The chart below plots the relationship for house values (one of his measures of social status).
As Clark noted, his findings are consistent with a simple model of inheritance devised by Sir Ronald Fisher1, in which the similarity among relatives is largely due to shared, additive genetic factors. However, he was careful to point out “there is no proof here that additive genetic transmission causes social status”.
Naturally, not everyone was convinced. In fact, some people were clearly very irked by the paper. (The fact that it found its way into a prestigious journal no doubt enraged them further.) A group of academics penned a response accusing Clark of “hereditarian fallacies”, despite his statement that there’s “no proof” for Fisher’s model. Instead, they suggested, the patterns in the chart above are explained by the transmission of wealth from parents to children. And they presented some tentative evidence supporting their interpretation.2
Clark has since hit back with a rejoinder to his critics. Regarding the claim that his earlier findings are explained by wealth transmission, he makes several points.
The amount of wealth a child inherits is determined not only by the amount their parents bequeath but also by the number of siblings they have to share it with—the more siblings, the less each one inherits. What’s more, before about 1880, family size was largely random. (Couples didn’t practice contraception, and child mortality was only moderately higher among the poor.) Hence variation in family size produced a kind of natural experiment. Children from large families just happened to inherit less wealth, while those from small families just happened to inherit more.
If Clark’s critics were right, the lucky children from small families should have achieved higher social status than their peers from large families who ended up with a smaller share of the pie. But they didn’t. In fact, children with 10 or more siblings achieved the same average social status as those with two or less—as the chart below indicates.
What’s more, wealth was typically inherited down the patriline, so if Clark’s critics were right, you would expect the correlations in social status between fathers and their children to be stronger than those between mothers and their daughters. Once again, however, there was no evidence of this. Clark and his colleagues found “complete symmetry” of the patriline and matriline for measures of social status other than wealth:
Plus, there is other research showing that inherited wealth can’t explain the similarity among relatives.
Moving from 19th century England to 19th century America, Hoyt Bleakley and Joseph Ferrie studied Georgia’s 1832 Land Lottery. When the Cherokee Indians were being evicted from Georgia, the state held a lottery to allocate land for white settlers and their families. Winners received a 160 acre plot; losers received nothing. (Remarkably, 98% of eligible men registered to take part.) It was another natural experiment in which some families were randomly allocated a large sum of wealth.
Did that wealth make any difference? Not a bit. Bleakley and Ferrie were able to identify the sons of those who took part in the 1870 and 1880 US censuses. They found that sons of winners had no better adult outcomes—in terms of wealth, income or literacy—than the sons of losers. And this, they say, points to “only a limited role for family financial resources in the formation of human capital”.3
On the other hand, Micah Villarreal studied oil discoveries in the Creek Nation in the early 1900s, and found that the wealth they produced did make some difference. Creek Freedmen (Black Americans formerly enslaved by Creek Natives) who received payments for oil discovered on their land were about 6% more likely to send their children to school than their counterparts whose land did not yield oil. However, they were no more likely to own their homes (though Villarreal attributes this to the fact that many migrated to urban areas where home ownership was less common).
Her findings suggest that wealth may have a small long-run payoff for people in particularly disadvantageous circumstances, such as Black Americans circa 1900. Having said that, the only long-run outcome she was able to track was the children’s school attendance—rather than, say, their income in adulthood.4
Perhaps the single most compelling study in this context is by David Cesarini and colleagues. They exploited yet another natural experiment—in this case, the Swedish national lottery—to gauge the causal impact of wealth on child development. As in Bleakley and Ferrie’s study, the prizes were both randomly assigned and large in magnitude (about three quarters were equal to or greater than Swedish median income). Another advantage of Cesarini and colleagues’ design is that they were able to link players’ wins and losses to their children’s outcomes over 20 years later, with virtually no sample attrition.
What did they find? Lottery wealth had zero impact on children’s cognitive test scores—whether at school or military enlistment. Nor did it have any impact on their “non-cognitive skills”. As the authors note, thanks to their large sample size, their estimates could be “bounded to a tight interval around zero”, suggesting that “the effect of permanent income on children’s outcomes is small” (or, indeed, nil).
Having access to more wealth due to random factors like your sex, your number of siblings5, or whether your parents happened to win a lottery, doesn’t seem to matter much in the long run. Individuals from poor backgrounds do about as well as otherwise similar individuals from rich backgrounds. The hidden gems, the Michael Faradays of the world, usually don’t stay hidden. And the dull trust-fund kids tend to squander their advantages.
Clark’s interpretation of his findings as reflecting genetic inheritance was therefore perfectly reasonable. His critics’, on the other hand, was not.
Noah Carl is an Editor of Aporia.
Become a free or paid subscriber:
Like and comment below.
Unlike Faraday, Fisher was born into comparative luxury. He lived for a time at the magnificent Inverforth House before attending Harrow (one of Britain’s most elite schools) and later Cambridge.
In another study, Philipp Ager and colleagues found that sons from slaveholding families, which lost much of their wealth in the Civil War, had almost entirely recovered it by 1900. Hence randomly losing a large sum of wealth doesn’t seem to matter much either. However, the authors discount inherited ability as an explanation, emphasising “social networks” instead. But, so far as I can see, their findings are consistent with both.
In a similar study, Cache Ellsworth and colleagues reported modest, positive effects of wealth shocks. However, they looked at various outcomes and did not observe consistent effects. Nor did they adjust for multiple comparisons. For example, they observed a significant effect on son’s occupational income but not on their annual earnings (and observed no effect for daughters).
The randomness of family size only applies to the period to which Clark and his colleagues refer (i.e., England before the late 19th century). In the 21st century, family size is a not a random variable.
I consider helping pay for schooling to be a
generational transfer of werh.
This makes sense. Sometimes there is knowledge/wisdom in common sayings or observations. One saying I’ve always taken to heart is “Shirtsleeves to shirtsleeves in three generations!” If bequeathed wealth is so advantageous, why the adage?
The opposite side of the coin: For myself, my family was anything but wealthy or privileged. None had advanced education, nor worked in White Collar jobs. Yet here I am, a retired college grad (first) with multiple degrees, and with children who are themselves college grads and employed in occupations that provide a higher standard of living than wife and I could ever have imagined in our day.
The only thing wife and I bequeath them was a good genome, what else could we bequeath—we’re not dead yet. 😉