par Francesco Vona
A high level of income inequality is commonly regarded to be more acceptable when associated with high social mobility. Empirical evidence has however shown that unequal countries are rarely able to ensure high social mobility to their citizens. On the contrary, countries that rank high in the level of inequality are also the worst in term of social mobility[i]. The simple reason is that a given level of social immobility is amplified when rewards to individual characteristics, which are transmitted from parents to child, are larger. For instance, when the earning advantage for the high skilled is large, intergenerational inequality (that is: the correlation between parent and child incomes) increases because, on average, high skilled workers come from better family backgrounds.
Economists tend to attribute cross-country differences in social mobility to the working of the educational system and its influence on the effective skills possessed by individuals coming from different family backgrounds. In particular, several empirical studies using standardized test scores show that there exist substantial background-related differences in competences and skills at a given level of educational attainment[ii]. Among OECD countries[iii], the influence of family background on test scores achievements is particularly strong in France (the second worst country after the USA in terms of intergenerational educational inequality), Germany and the UK, while it is relatively weaker in Italy and Spain. Whereas background-related differences in the effective level of skills certainly play a major role in creating persistency in socio-economic statuses, the working of labour markets is also an important, yet neglected, source of social immobility. On the one hand, labour market institutions reduce the observed level of intergenerational inequality whereby institutions compressing wages (i.e. centralized wage bargaining, high unionization or minimum wage) are present. On the other hand, family ties constitute a labour market network that can help well-off individuals in finding good jobs and obtaining promotions.
In a recent paper (Raitano and Vona, 2014a)[iv], we assess the role played by labour market networks and individual skills in the transmission of socio-economic inequalities. We argue that high levels of intergenerational inequality can be due to: 1. formal educational attainment; 2. other (empirically unobservable) dimensions of human capital affected by family background, i.e. soft skills or better quality of education; 3. family and social ties affecting labour market outcomes and occupational sorting. Our main idea is to use intergenerational occupational mobility to distinguish between two types of association between family background and child earnings. A standard type emerges because, especially in top occupations, the well-off child should have a higher level of human capital (a glass ceiling effect) due to the fact that he attended top schools or inherited better soft skills. In contrast, the second type is associated with insurance for the children of the well-off ending up in bottom occupations (a parachute effect), who clearly display a low level of skills for a given level of education. To implement this idea, we use the 2005 module on intergenerational mobility of the EUSILC dataset and examine these two effects in eight EU countries characterized by different levels of intergenerational inequality and belonging to different welfare regimes. Our empirical analysis is motivated by the claim that returns to upward and downward social mobility could arguably stem from different sources. A glass ceiling of upward mobility is likely to depend on both network effects and unobservable skills that are positively correlated with family background. Conversely, it is hard to believe that the parachute effect can be associated with better unobservable skills; hence, in this case, family networks should be of paramount importance.
By way of an example, imagine that a child is in the first tercile group (low social position) of its distribution but that his father was in the third tercile group (high social position). This individual clearly has a good background, but his relative position signals that he has a low ability. In this case, a positive association between family background and earnings (i.e., a parachute effect) would depend on the family network rather than on unobservable skills related to the child’s background. Conversely, it is not easy to infer the true unobservable skills of individuals who maintain their positions and earn more than others while sharing the same occupation but coming from a worse background. Hence, the identification of the glass ceiling effect is more problematic.
We find that family ties can create a considerable earning advantage for Spanish and Italian workers[v]. In these two countries, the high observed intergenerational inequality is mainly explained by a parachute effect for the well-off worsening their social position. In Italy, this parachute effect is particularly high: all else equal, the child of the well-off who worsens its social position earns annually 12% more than the child of the worse-off who stays in the same position. This result is consistent with a sociological view of social mobility where families play a key role both in the allocation of workers to jobs and in determining earning increases within a job[vi]. Interestingly, this result does not hold for other immobile European countries, such as the UK and to a lesser extent France. In these cases, the earning advantage of the well-off is fully driven by a penalty for those climbing the social scale, i.e. glass ceiling effect. While this result seems consistent with the classical human capital view of intergenerational inequality (where access to elite educational institutions is highly dependent on family background), our study cannot discriminate between the two explanations because a glass ceiling at the top could also be engendered by social networks. However, since the glass ceiling effect is widespread across all countries, including more equal ones (i.e. Germany, Finland, Ireland and Denmark), this effect is most likely due to unavoidable features either of the educational system or of the cumulative process of skill formation, at least in countries where students with similar socio-economic backgrounds are sorted into the same school.[vii]
Overall, our study suggests that intergenerational transmission of inequality strongly depends on the features of the country’s labour market, especially in Mediterranean countries where family ties are extremely important in finding good jobs. Further research is required to understand which part of intergenerational inequality emerges during the educational period and which part emerges during the working career, accounting for the learning advantage possessed by high skilled individuals and thus for their steeper earning profiles. In future research[viii], we aim at decomposing the two effects in a more precise way for a cohort of Italian workers that we observe during their entire careers.
[i] See: Corak, M., 2012. How to Slide Down the ‘Great Gatsby Curve’: Inequality, Life Chances, and Public Policy in the United States. Center for American Progress, December. Available at https://milescorak.files.wordpress.com/2012/12/corakmiddleclass.pdf.
[ii] See: Fuchs T., Wößmann, L., 2007. What accounts for international differences in student performance? A re-examination using PISA data, Empirical Economics 32.
[iii] See: http://www.oecd.org/centrodemexico/medios/44582910.pdf.
[iv] Raitano, M., Vona, F., 2014a. Measuring the link between intergenerational occupational mobility and earnings: evidence from eight European countries, Journal of Economic Inequality forthcoming.
[v] The results are obtained running regressions for samples of representative individuals for each country.
[vi] See: Ganzeboom, H., Treiman, D., 2007. Ascription and achievement in comparative perspective, Russell-Sage University Working Group on Social Inequality, University of California-Los Angeles.
[vii] Mixing students from different background in the same schools tends to reduce the influence of family background on individual student achievement without having negative effects for the average student achievement in the school. See: Raitano, M., Vona, F., 2013. Peer heterogeneity, school tracking and students’ performances: evidence from PISA 2006, Applied Economics 45.
[viii] Raitano, M., Vona, F., 2014b. From the Cradle to the Grave: the impact of family background on carrier path of Italian males, mimeo.