Social mobility: up or down?

One of the most worrying issues in many developed countries is the legacy we leave to our future generations, such as many citizens believing their children will be in a worse financial situation than their own: only 9% of the French, 19% of the Japanese and 24% of the Spanish think their children will be better off.1 But is such pessimism warranted? This Dossier examines one of the key mechanisms to be able to face the future with optimism: intergenerational social mobility.

Social mobility can be defined as the relationship between the economic or social situation of parents and their children.2 Broadly speaking, this relationship can be analysed in two ways, each providing us with different information and to some extent being complementary. One alternative is social mobility in absolute terms, which measures whether the earnings received by children are larger or smaller than those received by their parents at the same age (adjusted for cost of living differences). The second alternative looks at social mobility in relative terms: i.e. the relationship between the children’s position in a certain country’s income distribution and the position occupied by their parents at the same age.

Absolute social mobility measures the likelihood of a child’s economic level being higher than his or her parents (in euros, for instance). Differences highlighted by this measure may therefore be due to changes in public policies or economic structure that directly affect the capacity of parents or their children to progress economically. However, this measure is also influenced by the economic growth occurring between generations. An increase in GDP per capita growth tends to improve absolute social mobility and vice versa.

A relative measurement of social mobility, however, focuses on differences in position on the «social ladder» and is therefore not affected by changes in GDP growth (at least not directly). It could be the case that relative social mobility has increased but, because of a relatively prolonged slowdown in GDP growth, such as the one observed the past few years in many developed countries, absolute social mobility has actually decreased. Both measures therefore provide complementary information and must be taken into account and analysed carefully.

Moving from theory to practice, we can start by examining indicators of absolute social mobility. In general, in most developed economies a large number of citizens have improved their economic situation compared with their parents over the past few years. In the US, for example, several studies have shown that between 65% and 85% of individuals today have higher incomes than their parents.3 This general statement is not without its opponents, however. For instance, a recent study by Chetty et al. (2016)4 has brought into question the existence of the «American Dream». According to the authors, in 2012 just 50% of Americans aged 30 were better off economically than their parents.5,6 This reduction in absolute mobility over the past few decades in the US is particularly acute in the Rust Belt (mostly the Mid-West states such as Michigan and Illinois). The data for Europe, however, suggest the American Dream is still alive and well. According to Eurofound (2017),7 in all countries most individuals are better off economically in absolute terms compared with their parents.

But what about the social ladder? As we have already mentioned, in this case we must look at relative mobility and the most widely-used measure is intergenerational earnings elasticity; i.e. the relationship between the income distribution of parents and children. Several examples illustrate this. There is 100% elasticity when the distribution of the parents and children is completely the same: for instance, a parent with 10% higher earnings than average when the child also has 10% higher earnings than average. On the other hand, there is 40% elasticity if the parent has 10% higher earnings than average but the child has 4% higher earnings than average. In short, the greater the intergenerational earnings elasticity, the lower the social mobility.8 An overview shows significant differences across countries (see the first chart). The Nordic countries and Canada have the highest social mobility while it is clearly lower in the US, Italy and UK.

Intergenerational earnings elasticity may be the most widely-used measure but it is not without its limitations. Most particularly because intergenerational elasticity may alter at different points in the income distribution and such information is not provided by this measure. Other relative measures focusing on different points of income distribution are therefore becoming increasingly popular. One example would be the likelihood of a child moving to the highest income quintile when the parent is in the lowest quintile. Using this kind of measure, Chetty has observed that, in the US, the mobility of individuals now entering the labour market is the same as for those born in the 1970s. As can be seen in the second chart, a child’s chances of reaching the top quintile from any point in the income distribution has remained stable over the past decades.9 In other words, the probability of going up a rung on the social ladder seems to be almost the same as in previous generations. However, Chetty himself stresses that the income distribution among US citizens has altered compared with previous decades and the marginal distributions of income have widened; i.e. inequality has increased. That is why, although the likelihood of changing social class has remained relatively the same, the benefits, or costs, associated with this change are different.

In most European countries, however, there is an upward trend in relative social mobility indicators such as social fluidity10 according to Eurofound (2017). Nevertheless, it is important that, for the latest cohorts, these indicators have stabilised in some countries such as Germany and Spain and have even declined slightly in France.

The social ladder therefore seems to be in working order. Nevertheless, we sometimes perceive the situation differently. In a recent study, Alessina et al. (2017)11 have shown that actual social mobility differs quite considerably from perceived mobility. Americans tend to overestimate the probability of moving from the lowest income quintile to the highest whereas this likelihood is actually lower. Europeans, however, believe that social mobility in their countries is lower than is actually the case. We therefore have our homework cut out for us: not only to continue improving social mobility but also to bring perception closer to reality.

Josep Mestres Domènech

CaixaBank Research

1. Pew Research Center, «2017 Global Attitudes Survey».

2. Alternatively, occupational status mobility and class mobility have been measured, among others. See Torche, F. (2013), «How do we characteristically measure and analyze intergenerational mobility?», Stanford Center on Poverty and Inequality, vol. 29, pp. 2-3.

3. See Reeves, R. and Halikias, D. (2016), «On the new Chetty-bomb that only half of Americans are better off than their parents», Brookings.

4. See Chetty, R. et al. (2017), «The fading American dream: trends in absolute income mobility since 1940», Science 356(6336), pp. 398-406.

5. Along similar lines, the study by Piketty, T., Saez, E. and Zucman, G. (2018), «Distributional national accounts: methods and estimates for the United States», The Quarterly Journal of Economics, 1–57, shows that the income of the bottom 50% of adults in the US has remained constant over the past 50 years.

6. Winship (2017) moderates these results on finding that mobility is closer to 68% once adjusted for different family sizes and using earnings measured at the age of 40.

7. Measured in occupational terms. Eurofound (2017), «Social mobility in the EU», Publications Office of the European Union.

8. Ideally, the comparison would be between the lifetime earnings (or permanent income) of both generations but this information is rarely available or contains measurement errors, so a more frequent comparison is made between the mean income over several years or at a specific age, such as 40.

9. See Chetty, R. et al. (2014), «Is the United States still a land of opportunity? Recent trends in intergenerational mobility», American Economic Review: Papers and Proceedings, 104 (5).

10. There is greater social fluidity when there is less association between the origin (defined as the father’s occupation) and destination (defined as the son’s occupation).

11. See Alessina, A., Stantcheva, S. and Teso, E. (2017), «Intergenerational mobility and preferences for redistribution», NBER Working Paper 23027.