Intergenerational Transmission and Ethnic Inequality in Switzerland. An Empirical Approach

Philipp Bauer

This book analyzes two fundamental aspects of equality of  opportunity in the Swiss society which have been disregarded up to the present. First, it measures the intergenerational transmission of education and income in Switzerland. Second, it examines if there is ethnic inequality in intergenerational transmission, in grading pupils' performance, and in students' grade signalling. It presents a set of new insights and new empirical evidence not only for Swiss society, but also in an international context. This work contains four empirical essays. Two of them analyze intergenerational educational mobility in Switzerland, one investigates income transmission, and one tests if the Swiss educational system treats immigrants equitably.

The first article is co-authored with Regina T. Riphahn and studies the intergenerational transmission of education. The article relies on data from the 2000 Swiss census, which surveyed the entire Swiss population, and focuses on the level of secondary schooling attained by youth aged 17. The students' school attainment is related to their parents' educational outcomes. Native-born Swiss students are directly compared to the group of second generation immigrants. The paper describes the heterogeneity in education transmission across population groups. Further, three economic theories of child educational attainment are extended to the patterns of intergenerational education transmission. First, the paper investigates whether higher costs of schooling are correlated with lower educational mobility. The theory refers to Chiswick who states that investment in schooling should decline with higher costs and increase with rising benefits. Second, it applies Becker's quantity-quality hypothesis, namely that the number of children as well as educational investments in each child affect educational mobility. This theory suggests that parents may either have many children in which they invest little, or they may have fewer children of higher investments. Finally, the paper analyzes if country of origin characteristics and ethnic capital determine the intergenerational transmission of education. Ethnic capital is defined as the average skill level in the parents' generation of a child's ethnic group.

The second article is also co-authored by Regina T. Riphahn. The focus lies on the investigation of the influence of a student's age at school tracking for intergeneration mobility. The paper relies again on data from the 2000 Swiss census, on data from the administrative education statistics which provide information on the grade of tracking as of 1995, and on an own survey of cantonal education departments which provides information on the typical grade and age of tracking for the period between 1994 and 1999. The underlying hypothesis assumes that if segregation based on ability in homogeneous groups takes place at early ages or low grades, a student's ability can only be measured with substantial noise, and parental background and expectations may dominate the tracking decision. So far, the empirical analysis of this relationship had not been explored by economists. The causal effect of this institutional feature can be evaluated based on its heterogeneity across Swiss cantons.
 
The third article studies intergenerational transmission of labor market outcomes. Here, the correlation of fathers' and sons' earnings is analyzed. Unfortunately, there is no Swiss data set that includes both the earnings of fathers as well as that of their sons. The data for the sons were taken from the Swiss Household Panel (SHP) covering the years 1999 to 2003. The respondents in the Swiss Household Panel provided information on their parental background. These parental characteristics were then used to predict the fathers' incomes. The basis for the prediction of fathers' incomes are data from the Swiss Labor Force Survey (SLFS), using the years 1991-2003. Information of the fathers' characteristics was taken from a period that differs from the fathers' actual work period. For the sons, only persons aged from 25 to 55 were included in the sample. Unfortunately, the methodological approach leads to a bias in the resulting intergenerational elasticity. Nevertheless, cross-national comparison of income mobility is still possible, but should only be drawn very cautiously. Additionally, a comparison between subgroups within the Swiss population is also reliable as any biases should cancel in the comparison. In contrast to the first paper, there is no information about a person's place of birth or her father's place of birth respectively. Hence, a son's ethnic affiliation is defined by his nationality. Two different econometric approaches are used; least squares estimation and quantile regressions. Using least squares, the elasticity measures the linear correlation between a son's economic status and that of her father. However, quantile regressions measures nonlinearities regarding certain income quantiles.

Finally, the fourth article analyzes if the Swiss educational system treats immigrants equitably. In most if not in all Western societies, immigrants are concentrated in low-achieving courses of study and in lower school levels. Furthermore, immigrant applicants get fewer callbacks for interviews and also get fewer job offers compared to native Swiss. For many economists as well as for sociologists, taste-based discrimination is seen as the cause of this different treatment. This paper presents another possible explanation for the case of Switzerland. The study thus focuses on two primary questions. First, are students' grades a good indicator for skills? And second, are natives and immigrants graded equally?

The first question refers to so called audit studies in which people of different ethnicities but otherwise identical credentials apply for the same job. Such studies are meant to provide direct estimates of the prevalence of illegal or taste-based discrimination. Most studies present significantly higher invitation or hiring probabilities for natives (or for 'whites' in the case of the U.S., for example). The authors conclude that companies are not 'race-blind' when hiring people. However, one might interpret these results in different ways. If the individuals' resumes reveal their skills inadequately, a lack of information might lead companies to rely on their previous statistical experiences. This study empirically investigates whether the grades' power of explanation in signalling cognitive abilities is strong.

The second question investigates whether immigrant pupils get lower grades even if their abilities are equal. Hence, it is tested if there is ethnic inequality in the grade-skill relationship that results either from discriminatory treatment by teachers, or from institutional conditions. If this were the case, employers were mislead to prefer native-born students compared to their equally skilled immigrant counterparts. Since the grades of immigrants suggest lower skills than they have on average, their skills were underestimated in the labor market.

The article relies on the PISA 2000 survey for Switzerland, and analyses 9th grade students. At that age, grades become a very important signal for both, the students' continuation at school and for the labor market. A large majority of sixteen-year-old students starts vocational training and applies for an apprenticeship. Further, they have few other signals to display their abilities than their grades.

Abteilung/Forschungsstelle | Arbeitsmarkt- und Industrieökonomik
Jahr | 2008