We know that socio-economic advantage can be transmitted across generations, sometimes for long periods, writes Francisco H. G. Ferreira. But how much of the inequality that we observe today is inherited? And what does this mean for “social mobility” in the context of our current economic system?

No fewer than 31 presidents of Costa Rica, and 285 members of its parliament, have descended directly from Don Juan Vázquez de Coronado y Anaya, a leading Spanish conquistador of that part of Central America, who was born in Spain in 1523. Or at least, that was the count up to 1975, when Samuel Z. Stone reported this finding in his book La Dinastía de los Conquistadores.

At the other end of the distribution, although slavery was formally abolished in Brazil in 1889, when Jèrèmie Gignoux and I divided Brazil’s population in 1996 – more than a century later – into 104 groups of people based on shared inherited characteristics (such as parental education, race or place of birth), we found that the six poorest groups, which accounted for 10 per cent of the population, were made up entirely of Afro-descendants.

Socio-economic advantage – and disadvantage – are transmitted across generations, often for long periods. The forces that drive this transmission are manifold: there is, of course, genetic transmission. But more importantly, there are various mechanisms through which human capital (embodied in more education or better health) is passed down within families. This happens at home, because of differences in both parenting and economic resources, but it also happens through sorting into different neighborhoods and different schools. And this is before we even consider the transmission of real and financial capital through the inheritance of wealth.

But is all of the inequality – in incomes, say – that we observe today inherited? Does all of it come from differences among our parents, and their parents and grandparents in turn? This seems unlikely: we do observe examples of social mobility. There are grandchildren of well-off families who have fallen on hard times, and people with modest family backgrounds who have amassed great wealth. There are some real “rags to riches” stories out there. But they are the exception, rather than the rule, which is exactly why they attract attention.

If not all of the inequality we observe is inherited, how much is?

The answer to this question has far-reaching political ramifications. On the one hand, there are those who argue that in a capitalist system, inequalities do not really matter at all, since social mobility is possible. For instance, in their 1962 book Capitalism and Freedom, Milton and Rose Friedman famously argued that because the United States was (believed to be) a country full of opportunity and thus marked by considerable intergenerational mobility, the inequality that one might observe at any particular moment in time was of little consequence. Conversely, critics have pointed to the intergenerational transmission of inequality as a key flaw of contemporary capitalism.

Either way, claims of social justice in any economic system will be weakened if personal success is largely pre-determined by the circumstances people inherit. Unsurprisingly, then, sociologists and economists have long attempted to work out the share of observed inequality that is driven by inherited circumstances – using various approaches to address this question. Sociologists have historically focused on occupational mobility: by tracing the evolution of occupations from parents to children, they could infer a great deal about changes in the class structure of the population, and about its fluidity. Economists have tended to look at income or educational mobility. In both cases, the dominant approach was to treat intergenerational mobility as the opposite of intergenerational persistence, and to assess persistence essentially as a measure of the association between the incomes (or schooling) of children and their parents.

There are many ways to measure this association: different studies have used regression coefficients (regressing a person’s income on her parent’s); compared parent’s and child’s ranks in their respective income distributions; or simply reported the correlation between parental income and child’s income. Correlation coefficients turn out to be quite useful, here, because the square of this coefficient gives a measure of the share of current inequality that is “explained” by variation in parental income. So, a correlation coefficient of 0.5 or 0.6 – at the high end of those observed in practice – would imply that 25 to 36 per cent of observed inequality is inherited.

What gets passed down: can we look beyond income?

But these estimates, which have been quite influential, essentially treat parental income as the only source of intergenerational transmission. In reality, of course, we know that incomes are not the only thing we inherit from our parents or, indeed, from the past more generally. There is plenty of evidence that parental education, occupation, and wealth matter in and of themselves. So does a person’s biological sex at birth, her race or ethnicity, and where she was born, to name but a few circumstances.

So, while measures of intergenerational mobility of income are good summaries of the association between those two specific variables (parental and child’s income), they are perhaps not able to capture all the mechanisms though which advantage and disadvantage are transmitted across generations. A broader approach is needed. And this is what the empirical literature on inequality of opportunity has attempted to offer.

Strictly speaking, inequality of opportunity is not the same as inherited inequality.  The former encompasses any inequality arising from factors outside an individual’s responsibility. In principle, this could include various aspects of luck, as well as certain actions of others for which one cannot be held responsible during one’s own lifetime. In practice, however, most of the variables available in the data used to measure inequality of opportunity were inherited at birth.

Taking a broader range of circumstances into account seems to lead to higher estimates of intergenerational transmission. In a recent study of Latin America, undertaken as part of the Latin American and Caribbean Inequality Review we looked at twenty-seven household surveys for nine countries, spanning the period between 2000 and 2015. Various “circumstance” variables were considered, including parental education and occupation, sex, race or ethnicity, and place of birth. Figure 1 shows the results for the latest year available for each of these nine countries, with the bars depicting the share of inequality in household income per capita which is attributable to inherited circumstances, according to one of the methods we present in the paper (specifically, a method using conditional inference trees).

Figure 1: Share of inherited inequality in nine Latin American countries

These are very high shares indeed, ranging from 44% in Argentina (in 2014) to 62% in Guatemala (in 2011). In Brazil and Panama, too, inherited inequalities account for, respectively, 59% and 55% of the total. Overall, in seven out of the nine countries, more than half of the total observed income inequality is attributable to inherited circumstances. (In terms of correlation coefficients, such as those obtained when looking only at income or education across generations, note that these shares would correspond to values between 0.7 and 0.8 – considerably higher than those typically observed, even in Latin America.)

The takeaway is clear: while mobility across generations certainly exists, the weight of the past – at least in Latin America – is enormous. Those coming from privilege do enjoy the wind on their sails, while those attempting to climb up from positions of disadvantage face odds stacked against them. That is not only grossly unfair – it is also likely to be wasteful of talent and to lead to misallocated resources. That, however, would need to be the subject of another post…

All articles posted on this blog give the views of the author(s). They do not represent the position of LSE Inequalities, nor of the London School of Economics and Political Science.

Image credit: Bernardo Ramonfaur via Shutterstock.