We use experimental data from the Head Start Impact Study to examine the effect of sequential participation in childcare programs on cognitive outcomes. Using a sequential threshold model, which accounts for selection beyond initial randomization, we estimate causal returns to program sequences, including joint and cross returns across skill investment programs. We then estimate a dynamic structural model as a juxtaposition and evaluate a counterfactual policy reform which limits individuals to one year of Head Start. Our results support Head Start implementation earlier in life, and support engaging low-ability children with center-based care and high-ability children with some home care.

Standard intergenerational measures have been shown to understate the long-run persistence of socioeconomic advantages in developed countries. We study theoretically and empirically whether this pattern extends to less developed settings, using Indonesia as a case study. Using the Indonesian Family Life Survey (IFLS) and Census data, we study multigenerational correlations in education across three generations. Contrary to previous findings, we observe greater multigenerational mobility than parent-child correlations alone would suggest. We develop a theoretical framework to highlight two key factors influencing multigenerational dynamics in developing countries: (1) financial and credit constraints, and (2) cultural norms related to marital sorting. To confirm their relevance, we exploit regional variations in exposure to the 1997 Asian financial crisis and in marital customs.

This paper shows how the Marginal Value of Public Funds (MVPF) can guide treatment allocation to improve social welfare. Under budget constraints, the optimal treatment targets individuals with MVPFs above a threshold that minimizes the opportunity cost of treatment. Using experimental data, we show that prioritizing high-MVPF groups under tight budgets can double Head Start’s social benefits compared to random assignment. Analyzing joint allocation across early (Head Start) and late (Job Corps) skill investment programs, we find that exclusive investment in early interventions is not optimal unless substantially higher welfare weights are placed on young children.

This paper presents a novel methodology to estimating the prevalence of informal (dependent) employment by leveraging publicly available and timely data that are maintained in most countries. The method employs partial identification techniques to derive bounds on the proportion of informal workers by comparing aggregate data from the Labor Force Survey with public-use Social Security records. Applied to Italian data under minimal assumptions, this approach yields credible estimates with narrow bounds: informal employment remained stable at 10–12% until 2020 but declined to 8–10% in the post-pandemic period. As official estimates indicate, informal employment is more prevalent in southern regions, though it has followed a downward trend. Notably, before a recent methodological revision, official estimates may have overstated the true rate in several instances.

This paper studies the consequences of the buildup of a new economic sector—the Norwegian petroleum industry—on investment in human capital. We assess both short-term and long-term effects for a broad set of educational margins, by comparing individuals in regions exposed to the new sector with individuals in unexposed regions. Importantly, we analyze how the effects and the mechanisms change as the sector develops. Our results indicate that an initial increase in the high school dropout rate is short-lived both because dropouts get their degrees later as adults, and because later-born cohorts adapt to the new needs of the industry by enrolling more in vocational secondary education. We also observe a decrease in academic high school and college enrollment except for engineering degrees. Financial incentives to both completing high school and field of study, are the most likely channels driving these effects.

We study the effects of changing the rule that defines how students are selected into high schools in a context where school capacity is fixed. Schools for which demand exceeds supply must necessarily exclude some students from enrollment. We provide a theoretical framework to analyze the overall effect of policy changes, taking into account the crowding-out effect. Exploiting a reform that implemented merit-based allocation in Norway, we show that we can identify the relevant parameters. The reform had an overall negative effect because of the negative impact on crowded-out students. Different allocation rules would result in higher average outcomes.

Do large economic shocks increase intergenerational earnings mobility through creating new economic opportunities? Alternatively, do they reduce mobility by reinforcing the links between generations? In this paper, we estimate how the Norwegian oil boom starting in the 1970s affected intergenerational mobility. We find that this resource shock increased intergenerational mobility for cohorts entering the labor market at the beginning of the oil boom in those labor markets most affected by the growing oil industry. In particular, we show that individuals born to poor families in oil-affected regions were more likely to move to the top of their cohort's earnings distribution. Importantly, we reveal that preexisting local differences in intergenerational mobility did not drive these findings. Instead, we show that changes in the returns to education offer the best explanation for geographic differences in intergenerational mobility following the oil boom. In addition, we find that intergenerational mobility was significantly higher in oil-affected labor markets across three generations and that the oil boom broke the earnings link between grandfathers and their grandsons.

This paper explores the household adoption of Generative Artificial Intelligence (GenAI) in the United States and Italy, leveraging survey data to compare usage patterns, demographic influences, and employment sectoral composition effects. Our findings reveal higher adoption rates in the US, driven by socio-demographic differences between the two countries. Despite their lower usage of GenAI, Italians are more confident in its potential to improve their well-being and financial situation. Both Italian and US users tend to trust GenAI tools less than human operated services, but Italians report greater relative trust in government and institutions when handling personal data with GenAI tools.

We quantify how the employment expansion accompanying Italy's post-pandemic recovery mitigated the distributional consequences of the contemporaneous surge in prices, which disproportionately affected households at the bottom of the expenditure distribution. Using linked administrative employment records and household survey and expenditure data, we examine labour income dynamics, employment transitions, differential inflation exposure, and the redistributive role of the tax–benefit system for Italian households without pension or self-employment income over 2018–2023.  Despite elevated inflation, households in the lowest expenditure quintile experienced gains in real labour income, whereas those higher in the distribution did not. The decline in inequality is driven primarily by employment entry among previously non-employed household members, while adjustments among continuously employed workers played a limited role. Extensive-margin gains reflect stronger demand for low-skilled labour rather than differential labour-supply responses to inflation. Microsimulations indicate that fiscal measures cushioned disposable incomes at the bottom but did not alter the central role of employment growth in shaping distributional outcomes.

Artificial intelligence (AI) is a general-purpose technology with broad applicability across domains and economic sectors, which is expected to have a significant impact on the labour market in the coming years. We review some of the most recent measurements of labour market exposure to AI in advanced economies and then assess the implications for the Italian labour market. We find that occupations that are more exposed to AI, i.e. more at risk of being complemented or substituted by it, are in the top two quintiles of the income distribution, mostly in the service sector, and employ a large share of women and of highly-skilled workers. Substitutable workers are more protected from the risk of job displacement as they are less likely to be self-employed or on fixed-term contracts. Current patterns of job-to-job mobility show high degrees of persistence within occupation types. We provide indicative evidence that moving out of the most exposed and substitutable occupations might be difficult and costly in terms of wage, especially when movements are towards less exposed occupations. 

Il presente lavoro analizza il sistema di ammortizzatori sociali (il sussidio di disoccupazione e le integrazioni salariali in costanza di rapporto di lavoro) e di politiche attive vigente in Italia e lo confronta con i sistemi prevalenti nei principali paesi europei. L’analisi include gli schemi di reddito minimo garantito in quanto strumenti di ultima istanza per quelle situazioni in cui gli ammortizzatori sociali non sono accessibili o sufficienti. Nel confronto con Germania, Francia e Spagna, il modello italiano appare caratterizzato da limiti nel disegno e nell’implementazione dei singoli istituti più che nel suo impianto complessivo. Gli aspetti più critici riguardano la scarsa implementazione delle politiche attive; l’accesso alle integrazioni salariali da parte di imprese in procinto di cessare l’attività, in contrasto con l’obiettivo della misura di salvaguardare posti di lavoro; i disincentivi all’offerta di lavoro e le iniquità insite nel disegno del Reddito di cittadinanza. 

In this paper we estimate the causal effect of a training program for disadvantaged youths on their long-run labor market outcomes. Individuals receive lottery offers to participate in the program, but are allowed to choose when to leave the program and to participate alternative programs. We consider a multi-stage decision setting, where individuals sequentially select which program to participate at every stage. The standard IV estimator using initial random assignment as instrumental variable identifies a weighted average of the effects of the treatment for subgroups of individuals differing in terms of potential duration of participation and choice of the alternative programs. We estimate a sequential choice model which allows us to separately estimate the effect of the treatment for these different subgroups. We use the estimated model to investigate the dynamic complementarity between different training programs and explore program targeting to improve the cost-effectiveness relative to the existing program.

We examine the effects of school closures on parental labor participation. We investigate how mothers and fathers respond differently to these closures and find that mothers are more inclined to take time off from work. However, in a country with a rigid labor market like Italy, this primarily manifests as an increase in both paid and unpaid leave (intensive margin) rather than a decrease in employment (extensive margin). We thus assess the consequences of a policy reform aimed at reducing the costs associated with taking leave, considering both short-term and long-term maternal outcomes. Additionally, we perform a welfare analysis to determine the overall benefits of implementing such a policy.

Informal networks are an important job search channel. While a growing literature focuses on referred workers within the firm, little is known about the long-term worker-level benefits from job searches conducted with the help of social ties. In this paper, I use Norwegian registry data and focus on displaced workers. I follow these workers with respect to the jobs they find after displacement and show that those who are newly employed where a member of their network is also employed, on average earned higher salaries before displacement, suggesting positive selection. In the years immediately following displacement, the earnings gap favoring workers employed with a member of their network increases, partly due to shorter unemployment spells. However, this gap vanishes completely seven years after displacement. I discuss possible explanations for the decrease in the earnings gap and I describe how this new empirical evidence relates to different theoretical models developed to explain the usage of networks in job search.