Paolo Li Donni

How does exposure to automation at the beginning of workers’ careers affect employ- ment resilience in the Italian labor market? Using matched worker-firm data covering the full non-agricultural firm-size distribution, we study how exposure to multiple au- tomation technologies shapes young workers’ employment trajectories. Automation exposure is associated with substantially greater employment stability, with estimated effects of 8-10 percentage points. A complementary institutional exercise around the removal of Italy’s COVID-19 firing freeze yields a consistent pattern, with differences in employment trajectories becoming more pronounced once firms regain the separation margin. In a labor market characterized by strong dismissal regulation and contrac- tual rigidities, these gains are far from uniformly distributed within young cohorts: they are systematically larger among permanent workers, workers in firms above the employment-protection threshold, tertiary-educated workers, and workers in technolog- ically advanced sectors. The resulting disparities in retention probabilities point to an important distributional dimension of technological change, with potential implications for within-cohort inequality.

We examine the substitution between primary and emergency secondary care in patients with chronic conditions by using a dynamic structural model for longitudinal count data. The hidden Markov bivariate Poisson model allows for disentangling unobservable time-varying heterogeneity from the dynamic effect of utilization of primary and secondary care. Also, unobserved heterogeneity can be modeled as a function of observable individual characteristics, including age, income, and education. We find that an additional standard visit in primary care reduces utilization of emergency care by 9%, while the substitution effect of preventive assessment visits is two times larger. The substitution effect is larger in individuals with a high utilization profile, it decreases with individual age, while it shows little difference by socioeconomic status. Finally, the probability of switching utilization profile decreases with age. Our analysis contributes to the existing evidence on the substitution effect from policy evaluation studies.

Health systems increasingly face the dual challenge of rising demand and persistent inequalities in access to care. Understanding how patients use different services—particularly the balance between primary and secondary care—is essential for informing policy. This study develops a Bayesian hierarchical bivariate Poisson model to jointly analyze general practitioner consultations and emergency department visits among older adults newly diagnosed with hypertension, using data from the UK Clinical Practice Research Datalink. The framework accounts for both individual-level heterogeneity and practice-level fixed effects, capturing substitution or complementarity between services. Results show that multimorbidity strongly drives both utilization, while socioeconomic deprivation is mainly associated with higher emergency care use, pointing to persistent inequalities. We also find substantial heterogeneity across GP practices, suggesting that organizational and contextual factors significantly shape care pathways.

We study the dynamic relationship between outpatient secondary care and emergency secondary care using linked Danish register data on approximately 30,000 individuals with type-2 diabetes followed for 5 years after an emergency hospital admission. The data present 3 challenges: utilization is observed as 2 interrelated count outcomes, attrition is informative because it is largely driven by mortality, and the timing of death is recorded only at the quarterly level. We address these features using a hidden Markov dropout bivariate Poisson model that combines time-varying latent health states with a discrete-time specification for dropout. This framework allows us to separately identify dynamic state dependence and unobserved heterogeneity in both care utilization and survival. Accounting for latent health dynamics substantially alters the estimated relationship between the 2 services: standard random-effects models suggest mild complementarity, whereas the hidden Markov specification indicates substitution. Modelling dropout as informative further refines these estimates and clarifies the role of unobserved heterogeneity, disentangling persistent frailty from time-varying health dynamics. The results reveal substantial heterogeneity in patient trajectories and show that more recent cohorts experience improved survival together with a shift from emergency towards outpatient utilization. The framework is applicable to settings with interdependent recurrent events and endogenous attrition.

This study develops a composite index of perceptions of inequality (PI) in two domains: inequality of opportunity and inequality of outcome. Using International Social Survey Programme (ISSP) data for 24 OECD countries, we recover patterns in public views on inequality and derive an optimally weighted PI score within a latent-variable framework. The index aggregates multiple indicators of societal and economic inequality, explicitly addressing multidimensionality and the unobservability of PI. It enables consistent comparisons across groups and over time, revealing context-invariant perceptions in the outcome domain but multidimensional perceptions in the opportunity domain. We validate the score against objective indicators such as the Gini coefficient and intergenerational mobility indices and document systematic differences in PI across demographic and socioeconomic groups. The index provides a compact tool for analysing public perceptions of inequality and their drivers, with applications to policy evaluation and social research.

An extended stochastic block model for clustering dynamic weighted network data is introduced, where the block memberships are represented by a sequence of latent variables following a Markov chain. Motivated by the availability of original data on patient transfers within a network of Italian hospitals, we propose to rely on a bivariate zero-inflated Poisson distribution to characterize the sparse and correlated interactions of hospitals. In particular, we explicitly model the distribution of the dyad referred to each pair of nodes, conditional on the states occupied by both nodes at a given time occasion. This allows us to discover unobserved blocks of hospitals sharing the same propensity to send or receive patients and to set up reciprocated relations. Model estimation is based on an algorithm that maximizes a tempered version of a variational approximation of the likelihood to address the problem of multimodality of the target function. Selection of the number of blocks is based on the integrated classification likelihood criterion. In terms of clustering, it is possible to assign units to blocks on the basis of a local decoding procedure that exploits the variational posterior probabilities.

This study develops a composite index to measure perception of inequality (PI), focusing on two domains: inequality of opportunity and inequality of outcome. Using survey responses from 24 OECD countries within the International Social Survey Programme (ISSP), we identify latent patterns in public perceptions of inequality and construct a PI score using a weighting scheme optimized through latent variable models. By aggregating various indicators related to societal and economic inequalities, we address multidimensionality and the inherent unobservability of PI, ensuring our index provides a nuanced and comprehensive representation. The optimal weighting methodology allows for clear comparisons across groups and over time, reflecting context-invariance in the outcome domain but multidimensional perceptions in the opportunity domain. Validations with objective inequality indicators, such as the Gini index and intergenerational mobility indices, confirm the robustness of our score. Our findings reveal the sensitivity of the perception of inequality to demographic and socioeconomic factors, aligning with previous research. This index offers a valuable tool for understanding public perceptions of inequality and its drivers, with implications for policy evaluation and social research. JEL codes: D63, D31, D83