Switzerland: High intergenerational income mobility, despite low educational mobility
Switzerland: High intergenerational income mobility, despite low educational mobility
The notion that every child should have the same chance to succeed is a common denominator across political parties. Equal opportunities are not only ethically desirable, but also important for economic growth. Economic growth may suffer if children are hampered in achieving their economic potential – a phenomenon that has been referred to as “lost Einsteins” (Bell et al. 2017).
One important facet of equal opportunities is intergenerational income mobility. How much does children’s income depend on their parents’ income? Despite its importance, few studies have reliably estimated intergenerational income mobility – mainly because of demanding data requirements. To ensure robust estimates, a large sample of longitudinal income data and information on parent-child relationships is needed. In recent years, some researchers have gained access to large administrative data sets, leading to a resurgence in the empirical literature on intergenerational income mobility, for example Chetty et al. (2014) in the US, Heidrich (2017) in Sweden, Corak (2017) in Canada, and Acciari et al. (2019) in Italy.
Income mobility in Switzerland
In a new study (Chuard and Grassi 2020), we document intergenerational income mobility in Switzerland for the first time with administrative data. Switzerland is an interesting case for studying intergenerational income mobility, mainly because its educational system is quite unique: over 70% of children opt for vocational training and education (VET) after finishing compulsory schooling. At the same time, Switzerland’s labour market differs from most European labour markets because it is much more flexible. Also, government support of early childhood policies is rather modest compared to Nordic countries. These two factors could depress intergenerational income mobility.
Our data contain the universe of labour incomes since 1982 and administrative family linkages from the civil register. We also merge survey data on education and other individual characteristics for roughly one-third of the population. Our baseline estimate includes the cohorts from 1967 to 1984 and comprises roughly 850,000 child observations.
For our main estimate, we largely follow Chetty et al. (2014) and estimate the slope of a regression of a child’s income rank on a family’s income rank – the so-called rank-rank slope measure (RRS). Specifically, we measure child income as the mean income between the ages of 30 to 33 and parental income as the mean income when a child is aged 15 to 20. Next, we transform these mean incomes into cohort-specific percentile ranks.
Regressing child income rank on family income rank yields a rank-rank slope of 0.15. This is depicted in Figure 1, in which the blue points show the average child rank for two binned parent ranks, while the red line shows the predicted child rank of the rank-rank regression. The flatter the line, the higher the intergenerational income mobility. Conspicuously, the relationship is almost linear, as previous studies have found. This linearity confirms that the rank-rank slope is an insightful and parsimonious measure across parent income distribution.
Figure 1 Rank-rank relationship
A rank-rank slope of 0.15 indicates that children with parents in the highest income percentile on average end up 15 ranks higher than children with parents in the lowest percentile. Translated into monetary units, this rank difference corresponds to roughly 12,000 Swiss francs (roughly $12,000) at the ages of 30 to 33, which is around twice the median monthly salary in Switzerland.
Compared to other countries, the rank-rank slope for Switzerland is low and income mobility is accordingly high. For instance, Chetty et al. (2014) estimated a rank-rank slope of 0.34 for the US. Acciari et al. (2019) found a rank-rank slope of 0.25 for Italy, while Connolly et al. (2019) calculated a slope of 0.22 for Canada. Even for Sweden, possibly the most comparable country among this list, the estimated slope by Heidrich (2017) was higher, at 0.18, than that for Switzerland.
Another interesting statistic to describe mobility is the ‘American Dream’ measure, which indicates the share of children with parents in the bottom quintile of the income distribution that make it to the top quintile. This number can be inferred from the quintile transition matrix shown in Table 1 (at the intersection of child quintile 5 and parent quintile 1). In Switzerland, 12.9% of children achieve the American Dream. This is much larger than the share in the US (7.5%) or Italy (10%), but lower when compared to Sweden (15.7%).
Table 1 National quintile matrix
Low educational, high income mobility
These high mobility estimates are rather surprising. As mentioned, Switzerland’s labour market is among the most flexible in Europe: companies face relatively few constraints for hiring or dismissing personnel, wage setting, and work schedules.
A country’s educational system is among the primary suspects for causing high income mobility. This is the case mainly because education is an investment in human capital and labour income the return on that investment. Thus, one would intuitively expect high educational mobility in a country with high income mobility.
In Switzerland, children opt to pursue either VET or high school education mostly at the age of 16. The Gymnasium provides graduates with access to almost all university programs. Importantly, most children opt for the VET track (over 70%) and only a small share (ca. 20%) attend the Gymnasium.
We determined whether high educational mobility coincides with high income mobility estimates by plotting the share of Gymnasium students versus VET trainees by parent income rank (Figure 2). The chosen educational track depends highly on family income rank: up until family income rank 50, only around 10% of children attended Gymnasium or earned a master’s degree. This finding is puzzling: Despite high income mobility, whether children obtain a university degree seems to depend highly on parental income.
Figure 2 Child education and family income
What might explain this ‘high income mobility, low educational mobility’ conundrum?
One explanation might be that measuring educational mobility in terms of university versus non-university qualification is too narrow a concept in the case of Switzerland. Even after completing an apprenticeship, many children pursue some kind of non-university tertiary education. In Figure 2, the green points indicate the share of children with some sort of tertiary education. Even among the lowest parent income ranks, almost 40% of children end up with some sort of tertiary education. This achievement is likely to increase human capital and consequently labour income.
Another, tightly linked explanation is that the Gymnasium/university wage premium is lower in Switzerland than the comparable college premium in other countries. A simple regression analysis shows that Gymnasium graduate wages are only around 18% or 2.5 percentile ranks higher than those of their VET peers at the ages of 30 to 33.
Although we are unable to causally demonstrate that the VET system boosts intergenerational income mobility in Switzerland, several good arguments support such an effect. Apprenticeships incur almost no costs for parents. Some children even contribute small amounts to their family’s income during vocational training. Furthermore, VET is not the exception, but the rule. Even in the highest parental income ranks, almost half of children opt for the vocational track. Thus, VET is not stigmatised as in other countries, but widely accepted across the entire parent income distribution. The high acceptance and low costs of VET combined with many post-apprenticeship higher education options make children’s accumulation of human capital less dependent on parental income.
Of course, more research on the causal effect of VET on upward mobility is needed. If such research confirms a positive effect, countries with low intergenerational mobility should consider strengthening their VET system as a policy option.
Acciari, P, A Polo and G L Violante (2019), “‘And yet it moves’: Intergenerational mobility in Italy”, NBER Working Paper 25732 (see also the Vox column here).
Bell, A, R Chetty, X Jaravel, N Pektova and J Van Reenen (2017), “Who Becomes an Inventor in America? The Importance of Exposure to Innovation”, The Equality of Opportunity Project (see also the Vox column here).
Chetty, R, N Hendren, P Kline and E Saez (2014a), “Where is the land of opportunity? The geography of intergenerational mobility in the United States”, The Quarterly Journal of Economics 129(4): 1553–1623 (see also the Vox column here).
Chuard, P and V Grassi (2020), “Switzer-Land of Opportunity: Intergenerational Income Mobility in the Land of Vocational Education”, University of St. Gallen, School of Economics and Political Science.
Connolly, M, M Corak and C Haeck (2019), “Intergenerational Mobility Between and Within Canada and the United States”, Journal of Labour Economics 37(S2): S595-S641.
Heidrich, S (2017), “Intergenerational mobility in Sweden: A regional perspective”, Journal of Population Economics 30(4): 1241–1280.