Abstract: This paper investigates whether parental occupations of middle school classmates causally influence high school track choice—a critical margin for intergenerational mobility. Using administrative data from Italy, I compare students whose parents hold the same occupation and exploit within-school, within-cohort variation in classroom composition arising from quasi-random class assignment. A one-standard-deviation increase in exposure to classmates with high prestige parental occupations raises academic track enrollment by approximately 2 percentage points, while equivalent exposure to classmates with low prestige parental occupations reduces it by a smaller magnitude. I then disentangle this effect from the role of classmates' academic achievement by instrumenting the latter using partially overlapping groups of classmates. The estimated effect of classmates' parental occupations persists, and I find evidence that classmates’ achievement does not mediate it, consistent with non-academic channels such as information transmission, aspiration formation, and social-status considerations. These findings suggest that classroom composition policies can shape educational sorting beyond their effects on student achievement.

Presented at: Norwegian School of Economics internal seminar; University of Padova internal seminar; BRUCCHI LUCHINO workshop; LEER conference; ESPE conference; EALE conference; QED jamboree; Trondheim University internal seminar; PD-VR-VE PhD workshop 

Abstract: We examine whether delaying selection into secondary school streams can reduce socioeconomic inequality in cognitive aging. We link data on individuals aged 50+ from 14 European countries in the Survey of Health, Ageing and Retirement in Europe (SHARE) to a newly compiled database on reforms that postponed the age of first tracking. Exploiting within-country, within-cohort variation in tracking age, we find that later tracking significantly reduces socioeconomic status (SES)-related disparities in late-life cognition. A one-year increase in tracking age relative to the sample median reduces the SES gradient in word recall by around 8%. Furthermore, our analysis shows that delayed tracking narrows SES gaps in completed years of education and in completion of at least vocational upper-secondary school. It also improves access to white-collar, high-prestige and less physically demanding first jobs. These findings suggest that education policy should be incorporated into broader strategies aimed at reducing health disparities throughout the lifespan.

Presented at: Norwegian School of Economics ERS; University of Padova internal seminar; Gateway to Global Aging meeting; Boston Gateway to Global Aging’s Education, Aging, and Global Health Conference; EUHEA (by co-author); LIVES Thematic Workshops Lausanne (by co-author).

Abstract: We investigate the effect of postponing minimum retirement age on middle-aged workers’ depression. Using pension reforms in several European countries and data from the SHARE survey, we find that depression increases with a longer work horizon, but only among workers in occupations with a relatively high risk of automation. We explain our results with the higher job insecurity associated with occupations that are more exposed to automation, and rule out alternatives, including lifetime wealth effects and the differential exposure of occupations to the business cycle. 

Presented at: Netspar Pension Day; ROA Maastricht Seminar; LABORatorio Revelli Workshop.

Abstract: Medical expenditure risk is a major source of financial uncertainty and plays a central role in shaping households’ saving and consumption behaviour. In the United States, this background risk changes sharply at age 65, when individuals become eligible for Medicare, the federal public health insurance programme for older adults. Using microdata from the Survey of Consumer Expectations (SCE), this study examines how access to Medicare affects households’ saving, consumption, and debt responses to hypothetical income shocks. The analysis employs a fuzzy regression discontinuity design and shows that Medicare enrolment significantly alters households’ responses to positive income shocks, leading to a stronger saving response. In contrast, responses to negative income shocks remain largely unchanged.

Presented at: Padova internal postdoc&PhD seminar (by co-author); Padova Workshop on Macroeconomic Expectations. 

Abstract: We show that providing students with information on the academic and labour market returns of international mobility increased the share of students declaring an intention to apply to the Erasmus programme by 5 percentage points relative to the group receiving no additional information. Negative answers also increased, suggesting that the information reduces indecision rather than producing one-directional persuasion. AEARCTR-0017923