The old folks at home: parental retirement and adult children well-being (with Andrew Clark). Revise and Resubmit. [ Abstract ]
We here use UK data and exploit the State Pension eligibility age to establish the causal effect of parental retirement on adult children’s well-being in a Fuzzy Regression Discontinuity Design analysis. Maternal retirement increases adult children’s life and income satisfaction by 0.20 standard deviations in the short run. These impacts are stronger for adult children with lower incomes, with young children of their own, and who live close to their retired parents. We emphasise the critical role of intergenerational time transfers from retired mothers in enhancing their adult children’s well-being.
Small pictures, big biases: the adverse effects of an Airbnb design intervention (with Julio Garbers). Submitted. [ Abstract ]
A 2018 Airbnb design intervention reduced the size of the host profile picture, creating a natural experiment to test whether the salience of visual cues affects racial bias in the demand for Airbnb listings. Using scraped data from Airbnb in New York City and a face classification model, we find that, unexpectedly, the new design increased the Black-White demand disparity by 3.3 percentage points, an increase of about 30% relative to the pre-intervention gap. We show that smaller images made it harder for guests to detect positive facial cues -- especially smiles -- that are typically associated with higher demand, leading them to rely more heavily on skin color. In response, Black hosts updated their profile pictures to make their faces more visible and added basic amenities to their listings.
Generative modelling with Transformer architectures can simulate complex sequential systems across various %domains, including physics, chemistry, and health. applications. We extend this line of work to the social sciences by introducing a Transformer-based generative model tailored to longitudinal administrative data. Our contributions are: (i) we design a novel encoding method that represents socio-economic life histories as sequences, including overlapping events across life domains; and (ii) we adapt generative modelling techniques to simulate plausible alternative life trajectories conditioned on past histories or researcher-imposed variations. Using large-scale data from the Italian social security administration (INPS), we show that the model can be trained at scale, reproduces realistic labour market patterns consistent with known causal relationships, and generates coherent hypothetical life paths. %While our framework does not replace classical counterfactual identification strategies, it offers a complementary and scalable tool for individual-level simulation. This work demonstrates the feasibility of generative modelling for socio-economic trajectories and opens new opportunities for policy-oriented research, with counterfactual analysis as a particularly promising application.
Depression in old age has negative individual and societal consequences. With ageing populations, understanding life course factors that raise the risk of clinical depression in old age may reduce healthcare costs and guide resources allocation. In this paper, we estimate the risk of self-reported depression by combining adult life course trajectories and childhood conditions in supervised machine learning algorithms. Our contribution is threefold. Using data from the Survey of Health, Ageing and Retirement in Europe (SHARE), we first implement and compare the performance of six alternative machine learning algorithms. Second, we analyse the performance of the algorithms using different life-course data configurations. While we obtain similar predictive abilities between algorithms, we achieve the highest models' performance when employing high-dimensional and less structured data. Finally, we use the SHAP (SHapley Additive exPlanations) method to extract the most decisive depressive patterns by gender. Age, health, childhood conditions, and low education predict most depression risk later in life. In addition, we identify new predictive patterns in high-frequency emotion-enhancing life events and low utilization of dental care services.
This paper proposes spatial comprehensive composite indicators to evaluate the wellbeing levels and ranking of Italian provinces with data from the Equitable and Sustainable Well-Being (BES) dashboard. We use a method based on Bayesian latent factor models, which allow us to include spatial dependence across Italian provinces, quantify uncertainty in the resulting estimates, and estimate data-driven weights for elementary indicators. The results reveal that the inclusion of spatial information changes the resulting composite indicator rankings compared to those produced by traditional composite indicators’ approaches. Estimated social and economic well-being is unequally distributed among southern and northern Italian provinces. In contrast, the environmental dimension appears less spatially clustered, and its composite indicators also reach above average levels in the southern provinces. The time series of well-being composite indicators of Italian macro-areas shows clustering and macro-areas discrimination on larger territorial units.
This study investigates the relationship between quality of government and environmental wellbeing in 233 European regions at the NUTS-2 level. We find that subnational environmental data is spatially interdependent and construct a set of composite indicators of environmental wellbeing through Bayesian spatial factor analysis. By using these composite indicators in linear regressions, we demonstrate that institutional quality is a key determinant of environmental wellbeing. We also find that the institutions-environment nexus varies across dimensions of environmental wellbeing – institutions matter especially for air and soil quality. Policymakers should be aware that environmental destruction can be tackled by building more effective regional institutions.
The article provides a first quantification of the redistributive effects of automatic stabilizers and discretional policies imposed by the Italian government to limit the diffusion of COVID-19 in March 2020 and to compensate for income losses of individuals affected by the shutdown. In particular, we analyse the short term impact on family incomes, using the Italian module of EUROMOD which allow us to simulate the effects on incomes, poverty risks and inequality based on IT-SILC data combined with relevant information needed to identify the workers affected by the shutdown. The article provides timely evidence of the resilience of the Italian welfare state in the different geographical areas of the country facing an asymmetric shock, particularly strong from an economic perspective for some families and less for others even in the presence of compensative policies introduced by the government.
