ABSTRACT. We study the impact of a public microcredit programme in a developed credit market. Using administrative records from Lazio’s Fondo Piccolo Credito merged to the Italian Credit Register and firm balance sheets, we compare beneficiary firms to applicants rejected for administrative reasons. Using staggered difference-in-differences, we show that programme access increases firms' investment and the stock of bank credit, with a corresponding increase in leverage, indicating substantial crowding-in rather than substitution across lenders. Real outcomes respond more slowly: employment and wages display limited adjustment, while sales rise modestly as capital deepens. These results show that even in a developed financial market, financial frictions can generate economically meaningful credit constraints for small businesses, and that targeted public interventions can relax these frictions. While our four-year observation window does not allow us to determine whether the short-run decline in profitability gives way to sustained long-run productivity gains, the evidence points to benefits that materialize primarily through capital accumulation.

ABSTRACT. Several countries use preferential tax schemes to attract high-skilled workers from abroad. While these schemes are motivated by the positive spillover effects of tax- induced immigration on receiving countries, there is limited empirical evidence documenting these human capital externalities in the context of tax-induced immigration. In this project, we investigate the firm-level effects of tax-induced immigration in Italy, which offer preferential tax schemes for high-skilled immigrants and returnees since 2010. Using social security data and leveraging pre-existing variation in firms’ exposure to the policy, we estimate the effects of tax-induced immigration on productivity, wages, employment and other outcomes of firms and co-workers.

ABSTRACT. This paper examines the distributional effects of tax incentives aimed at attracting high-skilled foreigners and returning expatriates to Italy. While a large literature shows that high-skilled migration is responsive to taxation, much less is known about how preferential tax schemes affect income inequality, tax progressivity, and horizontal equity. We address this gap by evaluating the Italian preferential tax regime over the period 2016–2022. Using EU-SILC microdata, administrative tabulations from the Italian Ministry of Economy and Finance, and official migration statistics, we build a static tax-benefit microsimulation framework to estimate counterfactual income distributions in the absence of the policy. We assess the impact of the scheme on a broad set of pre- and post-tax inequality measures, including the Gini coefficient, Generalized Entropy indices, top income shares, and percentile ratios. Our analysis provides new evidence on the trade-off between efficiency and equity in tax policies designed to attract mobile talent.

ABSTRACT. This study analyzes the staggered implementation of a state-level policy that mandates single-room nursing home quotas. Our difference-in-differences analyses draw on data from the German Care Statistics from 2007 to 2019. They offer detailed insights into the universe of care recipients and all nursing homes. The policy significantly decreases the likelihood of individuals in severe need of care securing a nursing home bed and significantly increases residents’ out-of-pocket payments. Instead of a shift to professional home health care, we observe a significant increase in the utilization of informal care following the policy implementation. The policy generates substantial direct net fiscal gains for long-term care insurance and local communities, which likely outweigh the potential indirect fiscal costs, such as lower income tax revenues from reduced labor supply among informal caregivers.

ABSTRACT. **Abstract**

Long waiting lists for elective care remain a major policy challenge in publicly funded healthcare systems. While a large literature has examined the determinants of waiting times, much less attention has been devoted to waiting lists, which capture the stock of patients awaiting treatment and therefore reflect the accumulation of unmet demand. Importantly, waiting times and waiting lists need not evolve in parallel, as they represent distinct dimensions of healthcare system performance. This paper investigates both theoretically and empirically the impact of hospital mergers on waiting lists in healthcare markets with regulated prices.

We develop a model of hospital competition in which patients choose providers based on travel distance and waiting lists, while hospitals maximize a weighted combination of profits and patient utility. Within this framework, a merger internalizes two opposing competitive externalities: an altruistic incentive to attract patients and a profit-oriented incentive to avoid treating financially unprofitable patients. The former tends to increase waiting lists after a merger, whereas the latter tends to reduce them. In addition, mergers may generate cost synergies that improve productive efficiency and further reduce waiting lists. The overall effect is therefore theoretically ambiguous. We derive the conditions under which mergers increase or decrease waiting lists and show that, in the presence of cost synergies, waiting-list reductions are more likely when hospitals are more profit-oriented.

To test these predictions, we exploit 19 years of panel data on English NHS hospital trusts and estimate the causal effect of hospital mergers using a staggered Difference-in-Differences design that accounts for variation in treatment timing. We find that hospital mergers lead, on average, to a significant increase in waiting lists. However, this aggregate effect masks substantial heterogeneity. Using Foundation Trust status as a proxy for hospitals’ degree of profit orientation, we show that mergers involving Foundation Trusts also generate significant increases in waiting lists. These findings suggest that, despite potential efficiency gains, mergers can exacerbate congestion in access to elective care and that organizational incentives play a crucial role in shaping post-merger outcomes.

By shifting attention from waiting times to waiting lists, the paper highlights a neglected dimension of healthcare rationing and provides new theoretical and empirical evidence on how hospital consolidation affects access to care in regulated healthcare markets. The results have important implications for competition policy, hospital restructuring, and the design of healthcare systems seeking to balance efficiency and equitable access.

ABSTRACT. Objectives: In healthcare systems with high provider fragmentation — arguably the case for Switzerland — integrated care has long been proposed as a way to increase efficiency and quality of care. However, the evidence in support of this policy lever remains mixed. This paper investigates whether the connectedness of primary care provider communities affects patient-level healthcare spending in Switzerland, and whether the connectedness of an individual patient's main primary care provider plays an independent role.

Methods: I use Swiss health insurance claims from 2015 to 2024, organised into three three-year periods (triennia): two prior to the Covid-19 pandemic (2015-2017 and 2017-2019) and one after (2022--2024). For each triennium, I use the first two years of claims data to construct patient-sharing networks among primary healthcare providers, applying a spatially-weighted Leiden community detection algorithm with the Constant Potts Model objective. I characterise each community by its internal density, the share of realised patient-sharing links over all possible links within the community, as a measure of care connectedness. Each patient is assigned to the community generating the largest share of their outpatient expenditure in year two. Patients who remain in the same dominant community between year two and year three serve as the comparison group (stayers), whilst those who switch are classified as movers. Moves are further characterised by direction towards communities with higher or lower density, as defined by national quintiles. I estimate the effect of moving on patient spending across six categories (outpatient, inpatient, drugs, laboratory, physiotherapy, and emergency care) using two complementary designs: a patient fixed-effects difference-in-differences approach Hull (2018) and a within-movers estimator Finkelstein et al (2021) that compares movers by direction conditional on sharing the same origin community. I complement this analysis with a descriptive cross-sectional specification linking patient spending to the degree centrality of their main primary care provider.

Results: On the extensive margin, moves towards less connected communities are associated with increased likelihood of spending on drugs, laboratory tests, physiotherapy, and emergency care relative to stayers. Moves towards more connected communities are associated with reductions in physiotherapy and emergency care spending. The within-movers specification confirms an asymmetric pattern: moves to less connected communities increase the likelihood of laboratory and emergency care spending, whilst moves to more connected communities reduce it. Results for the intensive margin are broadly consistent. Descriptive evidence further suggests that patients whose main primary care provider has a higher degree centrality spend less on drugs and physiotherapy, on both margins.

Discussion: These findings suggest that the connectedness of primary care provider communities shapes downstream healthcare utilisation. Moves toward more integrated communities are associated with lower spending on discretionary and potentially avoidable care categories, consistent with tighter collaboration improving information flows and reducing unnecessary duplication. The complementary provider-level evidence points in the same direction. These results have implications for policies aimed at promoting integrated care in fragmented primary care systems such as Switzerland's.

ABSTRACT. We study fiscal spillovers across Italian municipalities, asking whether a municipality's current public spending responds to that of its geographic neighbours. This relationship is difficult to estimate because neighbouring budgets are jointly determined and may also react to common spatial shocks. We revisit this classic issue by developing a design based on plausibly exogenous variation from the post-2014 expansion of the Italian refugee reception network, which will be compared in the next version with more traditional fiscal-spillover specifications. Using a panel of Italian municipalities over 2008--2021, we instrument neighbours' current expenditure with their exposure to SAI and CAS reception capacity and estimate spatial-lag models with municipality and year fixed effects. We focus on current expenditure because it captures recurrent service provision and short-run budget responses, whereas capital expenditure is typically lumpy, project-based and harder to interpret in a contemporaneous annual framework. The results show positive but function-specific spillovers, concentrated in environment and education, while total current spending does not display a significant contemporaneous spillover. The findings suggest that local fiscal interactions operate mainly through service-related interdependence rather than broad expenditure mimicking or electoral yardstick competition.

ABSTRACT. This paper examines how large-scale public investment shapes democratic accountability in local government. We study the rollout of the EU-funded Recovery and Resilience Facility (PNRR) across Italian municipalities in the aftermath of the COVID-19 pandemic. Recovery spending generated salient but noisy signals about mayoral performance. Using data on the per-capita recovery funds spent in each municipality, we show that incumbent mayors whose municipalities spent more recovery money are more likely to be re-elected and to win by larger margins. This reward, however, is concentrated in specific kinds of spending: among the five PNRR missions, only education-related investment (Mission~4) improves incumbents' electoral prospects, while digital, green, cohesion, and health projects show no measurable effect. Disaggregating Mission~4 into its sub-components, the returns accrue to tangible, family-salient items (nursery schools, school buildings, and educational-enhancement programmes) rather than to less visible school digitalisation. To probe the mechanism, we use a large language model to score each spending item along three dimensions (visibility, daily impact, and credit-attribution to the mayor ) aggregate these scores to the municipality level weighting items by their salience, and find that all three dimensions independently predict re-election. Together, the results indicate that electoral accountability for recovery spending operates through what voters can see and attribute to their mayor, rather than through the underlying substance or long-run value of the investment.

ABSTRACT. The Italian National Recovery and Resilience Plan (NRRP) is expected to generate both long-term structural effects and short-term demand stimuli through public investment. However, the location of NRRP interventions—and therefore the territories expected to benefit from their long-term structural effects—does not necessarily correspond to the location of the firms involved in their implementation and, consequently, to the territories where their short-term economic effects arise. This paper investigates the magnitude and determinants of the cross-territorial effects generated by NRRP-funded public works. Using data from the REGIS and INPS databases, projects are mapped according to both their location and the location of contractors’ operational units at the level of Local Labour Market Areas (LLM). Of the €101.8 billion in infrastructure projects awarded up to October 2025, 72% were assigned to firms located outside the LLM in which the project is implemented, highlighting substantial cross-territorial effects. Multivariate regression models, including aggregate and gravity specifications, are then used to identify the main determinants of these effects. The results indicate that territories with stronger construction-sector capacity retain a larger share of procurement activity and compete more successfully outside their borders, while large tenders and local market stress increase reliance on non-local firms. Gravity-model estimates further show that procurement interactions decline significantly with geographical distance and are stronger between territories with similar socioeconomic and construction-sector characteristics.

ABSTRACT. The coupling of environmental planning with local culture is pivotal to enhancing urban attractiveness. As the first and most significant megacity established since the implementation of China’s Reform and Opening-up policy 45 years ago, Shenzhen maintains a population where internal migrants account for at least 70%. The city has consistently prioritised the enhancement of migrant economic wellbeing (MEW) as a primary strategy to attract talent. This study utilises Kernel Density Estimation to analyse pollutant transitions from 2014 to 2025, identifying a shift from PM2.5 to O3, with NO2 as a primary precursor. By employing a semiparametric spatial autoregressive (PS-SAR) model via the pspatreg R package, the research examines the nexus between NO2, the job market, rental housing culture, and wellbeing. Empirical evidence reveals that NO2 has a significant negative impact, with the model showing an average loss of 0.404 units of wellbeing for every unit increase in NO2, while expanded job positions significantly enhance wellbeing. Regarding housing culture, a threshold effect exists where the negative influence of total floors on wellbeing diminishes below 12 floors and becomes positive above this level. Spatial analysis indicates that controlling for variables reverses wellbeing patterns, suggesting that policy-makers should target high-low value clusters. These findings also provide critical insights for the polycentric development of Shenzhen and other global megacities.

ABSTRACT. Climate-related mortality is increasingly shaped by spatially uneven patterns of exposure, vulnerability, and adaptive capacity. Yet most evidence remains spatially aggregated, limiting understanding of how local demographic, environmental, and socio-spatial conditions structure climate-health risks within regions. This study investigates the territorial organisation of temperature-attributable mortality across 326 municipalities in Emilia-Romagna, Italy, during 2011–2024. A two-stage distributed lag non-linear modelling framework was combined with interpretable machine learning and clustering techniques to examine how place-based characteristics shape local vulnerability to non-optimal temperatures. Overall, 9.35% of total mortality, corresponding to 67,000 deaths, was attributable to non-optimal temperatures. The burden was overwhelmingly dominated by cold exposure, with 63,573 attributable deaths, particularly due to moderate cold. Heat-related mortality accounted for a smaller absolute burden, with 3,701 deaths, but displayed stronger spatial concentration across municipalities. Temperature-attributable mortality was strongly age-structured, with individuals aged 85 years and older accounting for the largest share of attributable deaths. Substantial intra-regional heterogeneity emerged across municipalities. Heat-related mortality was concentrated in densely urbanised, low-altitude, and environmentally pressured municipalities, whereas lower burdens were generally observed in sparsely populated and mountainous areas. Machine-learning interpretation identified urbanisation, population density, social vulnerability, long-term PM₂.₅ exposure, and territorial morphology as major predictors of local heat-related vulnerability. Clustering analysis further revealed distinct municipal vulnerability profiles characterised by different combinations of environmental pressure, settlement intensity, demographic structure, and socio-economic conditions. These findings suggest that temperature-attributable mortality should be interpreted not only as a climate-related public-health outcome, but also as a spatially structured form of territorial inequality.

ABSTRACT. In this paper, we study the effect of fine particulate matter (PM2.5) on respiratory morbidity and in-hospital mortality in Italy, focusing on municipalities hosting industrial plants regulated under the European Pollutant Release and Transfer Register. We combine administrative hospital discharge records with high-resolution pollution data and exploit exogenous variation in wind as an instrument for local PM2.5 exposure. Wind patterns shift the origin of incoming air masses, generating plausibly exogenous changes in pollution that are unrelated to contemporaneous health shocks. Our results show that higher PM2.5 concentrations lead to significant increases in hospitalisations for COPD, pneumonia, and respiratory failure, as well as associated in-hospital mortality. We find that a 1 μg/m3 increase in PM2.5 significantly increases COPD admissions by approximately 5 cases per municipality-month. A decomposition by clinical severity reveals that pollution effects are strongest among moderate cases, where COPD hospitalisations increase by about 4 cases, consistent with a mechanism in which particulate exposure triggers acute exacerbations of pre-existing conditions. Heterogeneity analysis shows that the health burden falls disproportionately on older individuals, those with lower educational attainment, and residents of Northern industrial regions. These findings underscore the substantial costs of industrial air pollution and point to the importance of monitoring and regulating emissions in areas where industrial sources are spatially concentrated.

ABSTRACT. Using an online experiment, we study how awareness of the short- and long-term costs of part-time work shifts public attitudes towards mothers working part-time. By randomly providing information about the short-term (earnings) and long-term (pension) costs of part-time work in a two arms experiment, we find that individuals change their the attitudes towards mothers' reduced working hours. Namely, treated respondents receiving information about the pension costs of part-time work are 15 percentage points more likely to suggest longer working hours for a hypothetical female job seeker, a 25% increase as compared to the control group. Moreover, the effect of the treatment is stronger for individuals with stronger preferences for and experiences with part-time work (i.e., parents and individuals working part-time), and in regions with more conservative gender norms. Given that mothers’ working hours are shaped by broader societal expectations, our findings suggest that low-cost information provision targeted at the general public can meaningfully shift attitudes in the broader social environment where these working decisions are taken.

ABSTRACT. We estimate the labor market effects of Italy’s statutory priority to work from home (WFH) granted to parents with children below age 12. Using population-wide matched employer–employee administrative data and a sharp age-based regression discontinuity design, we show that, once remote work becomes operational during the pandemic, eligibility increases mothers’ wages in high-WFH-intensity firms by 1.6–2.3 percent and improves their work continuity, raising paid weeks and paid days. No effects emerge before the pandemic, among fathers, in firms that never adopted WFH, or in organizational environments with low remote-work intensity. Effects are concentrated in non-manual occupations and in technologically and managerially prepared environments, smaller and younger firms, municipalities with stronger broadband infrastructure, and Central–Northern Italy. The results show that flexibility rights improve mothers’ labor market attachment only where firms and local conditions possess the capacity to implement remote work effectively.

ABSTRACT. In this paper we examine whether local availability of immigrants working in the care sector increased the employment probability of Italian women when an unexpected pension reform, in 2011, delayed retirement for a group of them. The availability of immigrants providing care services might allow women to absorb the increased mandatory retirement age, while remaining employed, rather than being forced to move to unemployment or inactivity to care for their parents. We use a survey-administrative linked dataset on individual labor market histories over the period 2007–2015, and adopt an identification strategy based on a triple-difference approach comparing before and after 2011, women in the cohort affected by the reform, with those in the immediately earlier cohorts and not exposed to it, interacted with the local supply of immigrant workers in the domestic sector. We find that the delayed retirement from the reform was accompanied by higher unemployment and inactivity for women, except for those living in areas with a higher presence of immigrants who were much more likely to remain employed. These effects were stronger for women with lower levels of education, lower income, and older parents.

ABSTRACT. This paper examines how social norms shape anti-coordination behavior under asymmetric income distributions. We study two three-player games: the classic Volunteer’s Dilemma, where multiple players can benefit, and a novel Deserver’s Dilemma, where rewards are scarce and exclusive. Using a controlled experiment, we elicit normative expectations conditional on endowment differences and assess their influence on strategic choices. In the Volunteer’s Dilemma, less-endowed individuals are more likely to pursue the high payoff, while richer individuals tend to sacrifice when group composition makes them focal. In contrast, endowment heterogeneity does not affect behavior in the Deserver’s Dilemma, where scarcity shifts attention from shared responsibility to individual competition. Norm compliance strengthens when conformity entails low individual risk but weakens under concentrated rewards. These findings highlight how structural features of interaction—such as payoff asymmetries and reward scarcity—shape the effectiveness of social norms as coordination devices.

ABSTRACT. In this paper we discuss some limitations of using survey data to measure inequality of opportunity. First, we link two fundamental principles of the theory of equal opportunities – compensation and reward – with the concepts of power and confidence levels in hypothesis testing. This connection can be used to address, for example, whether a sample has sufficient observations to appropriately measure inequality of opportunity. Second, we propose a set of tools to normatively assess inequality of opportunity estimates in any type partition. We apply our proposal to Conditional Inference Trees, a machine learning technique that has received growing attention in the literature. We propose a method to arrange standard tree-based partitions to reduce the risk of violating the compensation and the reward principles. Finally, we apply these methodological contributions to a quasi-administrative sample of Italian PhD graduates. We find substantial levels of labor income inequality among two cohorts of PhD graduates (2012 and 2014), with a significant portion explained by circumstances beyond their control.

ABSTRACT. This paper explores critical sources of bias affecting standard non-parametric measures of inequality of opportunity (IOp) within the framework that distinguishes between circumstances and effort. Specifically, we assess the impact of (i) inherent randomness in outcomes and (ii) unobserved variables. We examine these sources of distortion in turn. First, we consider randomness in outcomes, modelled as measurement error in income, and distinguish between uncorrelated multiplicative error and mean-reverting error. Second, we turn to unobserved variables, distinguishing between omitted circumstances and omitted effort variables. Our theoretical analysis is supported by an empirical application to EU-SILC data for four countries—Belgium, Italy, Norway, and Spain. The results suggest that the divergence in IOp estimates documented in the literature may reflect not only differences in conceptual frameworks and data choices, but also the fact that alternative IOp measures react differently to specific sources of bias, such as measurement error and unobserved variables. In particular, direct ex-ante measures appear generally more robust to measurement error. By contrast, the effects of unobserved variables are more nuanced: omitted circumstances tend to produce downward bias, whereas omitted effort variables tend to produce upward bias, and the magnitude of these distortions depends on both the IOp metric and the inequality index considered (MLD versus Gini). In the final part of the paper, we further extend the analysis to examine whether the error component can be attributed to measurement error or to unobserved variables. Overall, the paper highlights the importance of aligning methodological choices in the measurement of IOp with the structure and quality of the available data.

ABSTRACT. Voter participation in Italian abrogative referenda has been declining over the past three decades, with most referenda failing to meet the quorum threshold. Understanding what drives citizens to vote or abstain is key to assessing the future of this democratic event. This study investigates the determinants of participation in the 2025 Italian abrogative referendum, focusing on perceived complexity of the ballot questions and thematic heterogeneity. Using an original survey of 2.076 Italian citizens, this paper explores both the drivers of rejecting the referendum proposals and the determinants of turnout. Results show that, despite the issue salience, voters who perceive the questions as more complex are more likely to favor the status quo (i.e., status quo bias), rejecting the propositions. In turn, a stronger preference for rejecting proposals is significantly associated with abstention. Moreover, turnout is lower among citizens who perceive the referendum themes as too diverse, who trust the current government, or identify with the political right, consistent with official abstention cues from government leaders. These findings provide evidence on the behavioral mechanisms behind status quo bias and propensity to abstention in referenda, based on voters’ stated preferences.

ABSTRACT. Youth political disengagement is widely attributed to apathy. This paper argues it is, in part, a misperception. We conduct a pre-registered information provision experiment with approximately 3,000 nationally representative Italian citizens aged 18-34, exploiting the timing of a 2026 constitutional referendum on judicial reform. Respondents overestimate peer political participation by nearly 17 percentage points relative to the ISTAT benchmark. Two treatment arms - a short video citing either low or high statistics on youth engagement, each paired with an identical injunctive statement on the importance of voting - successfully shift posterior beliefs and increase stated political engagement by 0.13-0.16 standard deviations, regardless of the direction of the descriptive information. Instrumental variable estimation reveals two distinct mechanisms: the high-participation treatment operates through descriptive norm conformity, while the low-participation treatment triggers a backlash response. These findings suggest that mobilisation campaigns should anchor any descriptive message with a normative appeal.

ABSTRACT. This paper studies the causal effect of early electoral enfranchisement on students' educational decisions using comprehensive administrative data from the Netherlands. We exploit institutional features of the Dutch electoral and educational systems to identify quasi-random variation in eligibility to vote during the final year of secondary school. Our difference-in-differences design compares students who are above versus below the voting age cutoff in cohorts that are or are not exposed to elections in their graduation year. We find that eligibility to vote modestly worsens academic performance on final secondary exams but significantly shifts students' preferences in tertiary education toward social sciences—particularly economics—at the expense of STEM fields. We interpret this as evidence that exposure to democratic participation during formative years alters civic identity, information acquisition, and preferences over fields of study. Our findings highlight a novel channel through which democratic institutions shape human capital formation and contribute to literatures on political socialization, education economics, and the impressionable years hypothesis.

ABSTRACT. We study whether exposure to far-right political success affects teachers’ behavior and the academic performance of immigrant students. We combine Italian administrative data on standardized test scores (INVALSI), covering 10 cohorts of students in 2nd, 5th, 8th, and 10th grade, with detailed information on electoral outcomes. Exploiting variation in local anti-immigrant political sentiment and the timing of election campaigns, we proxy local attitudes using the vote share of far-right parties in recent elections, and capture short-run fluctuations in political salience through indicators for pre-election periods. By comparing immigrant and native students within the same municipalities over time, we implement a difference-in-differences framework that isolates the impact of political exposure. We find that in municipalities with stronger support for far-right parties, electoral campaigns are associated with a widening gap between teacher-assigned grades and standardized test performance for immigrant students. This pattern suggests that electoral campaigns amplify the effects of anti-immigrant sentiment on educational inequality. Our results provide causal evidence that political environments shape behavior within public institutions and contribute to disparities in educational outcomes.

ABSTRACT. This paper applies a network-based approach to the educational mobility indicators of the Global Database on Intergenerational Mobility (GDIM) in order to study their structure. Using the Tri- angulated Maximally Filtered Graph (TMFG) methodology, the analysis examines how different indicators are connected through conditional associations across cohorts, parental channels, and, for the 1980 cohort, child gender. Rather than treating mobility measures independently, the paper reconstructs the multidimensional measurement space of intergenerational educational mobility. The results show a stable empirical separation between absolute mobility indicators and rank- based measures, while traditional persistence measures such as intergenerational correlation and regression coefficients occupy more peripheral positions. Among all indicators, MIX consistently emerges as the main bridge between different measurement families. The analysis also reveals that the organisation of the indicator system varies across maternal and paternal networks and changes further when gender-specific structures are considered. Overall, the findings suggest that GDIM indicators capture distinct dimensions of educational mobility and are not fully interchangeable, highlighting the importance of indicator choice in comparative mobility research.

ABSTRACT. Mothers' life-cycle events play a crucial role in shaping their children’s outcomes. This paper examines whether adverse economic conditions experienced by mothers during adolescence have intergenerational consequences for newborn outcomes and early childhood development. Using rich longitudinal data from the UK, we follow first-born children from birth through age three and link their outcomes to local labour market conditions faced by their mothers at ages 16 and 18. Exploiting cohort variation in adolescent unemployment exposure, we show that children born to mothers exposed to worse labour market conditions—particularly at age 16—experience significantly poorer birth outcomes, including a higher probability of pre-term birth, and lower birthweight. By age three, the effects are more muted: maternal unemployment exposure is associated with a higher probability of low adaptive behaviour and with disadvantages in specific socio-emotional dimensions. Explanatory mechanism analyses suggest that adolescent economic shocks affect maternal labour market trajectories, earnings, health, and the timing of fertility before childbirth. Overall, the findings indicate that macroeconomic conditions experienced during mothers’ formative years can contribute to the intergenerational transmission of inequality.

ABSTRACT. This study investigates whether Italy’s National Recovery and Resilience Plan (NRRP) effectively addresses territorial fragility, focusing on municipal participation and implementation across missions. Using the Municipal Fragility Index (IFC) to capture structural vulnerabilities, we assess whether more fragile municipalities participate more intensively and receive greater resources. We also examine the role of local government capacity—both bureaucratic and political—and selected fiscal conditions in shaping participation and implementation performance. The empirical analysis relies on a municipal-level dataset and applies a Heckman sample selection model to account for non-random participation. Results indicate that territorial fragility is positively associated with project intensity in several mission–components, though less consistently with funding intensity. Political capacity plays a central role in explaining participation, while fiscal conditions related to municipalities’ COVID-induced budget dynamics significantly affect programme entry. A persistent North–South divide emerges: Southern municipalities show higher predicted participation, whereas the alignment between fragility and implementation is often stronger in the North. The results contribute to the debate on the territorial effectiveness of recovery policies and offer insights for improving place-sensitive investment strategies.

ABSTRACT. This paper investigates the local implementation of Italy’s National Recovery and Resilience Plan (NRRP) Mission 5, the component explicitly devoted to social inclusion and territorial cohesion. The analysis is conducted at the municipal level and is structured in three steps. First, the paper examines whether the allocation of Mission 5 resources reflects territorial fragility and pre-existing gaps in local social-service provision. Second, it evaluates municipal implementation performance through a two-stage Network Data Envelopment Analysis (NDEA), distinguishing between administrative efficiency in transforming allocated funding into actual payments and operational effectiveness in transforming payments into realized outputs. Third, it relates the estimated stage-specific scores to municipal institutional, social, and territorial characteristics. The evidence suggests that the allocation of Mission 5 funding is not neutral across municipalities. More fragile municipalities and municipalities with weaker pre-existing social and socio-educational service provision tend to receive more resources per capita, pointing to a redistributive component in the territorial distribution of funds. The NDEA results show that administrative absorption and operational effectiveness are distinct dimensions of municipal performance and should not be conflated. They also show that the estimated stage-2 scores are sensitive to the breadth of the output vector, confirming that Mission 5 is too heterogeneous to be represented adequately through a single narrow output definition. At the same time, the comparison between pooled Mission 5 results and the component-specific evidence for M5C1 indicates that a more homogeneous policy domain yields more stable and interpretable implementation patterns. Overall, the paper argues that evaluating Mission 5 requires separating allocation equity, financial absorption, and realized outputs rather than collapsing them into a single notion of performance.

ABSTRACT. The National Recovery and Resilience Plan (NRRP, Mission 4) allocates substantial resources to school infrastructure investments, including the renovation of existing school buildings, the construction of school canteens to support the expansion of full-time schooling, and the development or upgrading of school sports facilities. These investments aim to reduce territorial disparities and improve access to educational services. The main implementing bodies are municipalities, as well as provinces and metropolitan cities, which own school buildings for primary and lower-secondary education and upper-secondary education, respectively. However, since local authorities’ participation has largely relied on demand-driven application procedures, it is important to assess whether resources have been effectively allocated to areas with the greatest infrastructure deficits and the highest levels of socioeconomic vulnerability. This paper provides an empirical assessment of the territorial distribution of NRRP investments in school infrastructure, examining whether the targeting of resources is consistent with pre-existing infrastructure needs. The analysis builds on the infrastructure endowment indicators developed by Bucci et al. (2024), integrating them with data from the National School Building Registry and information on NRRP-funded projects at the municipal and provincial levels. The main objective is to assess whether investments have been directed towards territories characterized by lower levels of school infrastructure, greater building obsolescence, more limited availability of school canteens, sports facilities and full-time educational services, and higher levels of socioeconomic fragility. Particular attention is devoted to territorial differences in local administrative capacity, investigating whether resource allocation reflects actual needs or, alternatively, the varying planning and administrative capabilities of local governments. The paper also addresses the issue of the financial sustainability of these investments. While renovated and modernized school facilities may generate efficiency gains and expenditure savings, they also entail additional costs related to service provision, operation, and maintenance of new or expanded facilities. These costs will fall on the ordinary budgets of municipalities, provinces, and metropolitan cities. The analysis therefore explores whether beneficiary territories have adequate fiscal and financial resources to sustain the long-term management of NRRP-funded infrastructure and whether pre-existing financial constraints may have discouraged local authorities from participating in NRRP investment opportunities. The study contributes to the broader debate on place-based policies and the redistributive effectiveness of the NRRP in reducing territorial inequalities. It aims to provide evidence relevant both for the evaluation of school infrastructure investment policies and for understanding the financial and administrative implications that such investments generate for local governments responsible for the day-to-day provision of educational services.

ABSTRACT. Executives may implement legally contestable policies aggressively before courts reach a final legality determination, creating reliance and other sunk effects that make reversal costly. We study a complete-information sequential game in which an executive chooses policy aggressiveness and a court then decides whether to uphold the policy or strike it. If reversal costs increase sufficiently convexly in aggressiveness, the court strikes mild policies but upholds sufficiently aggressive ones to avoid disruption. Anticipating this, the executive overreaches-choosing a policy more extreme than its ideal point-to deter full reversal, yielding inefficient excess implementation relative to a commitment benchmark. Institutions that limit pre-review sunk effects (stays, phased implementation, expedited review) mitigate this distortion.

ABSTRACT. Judicial inefficiency is widely regarded as a structural constraint on economic performance, particularly in civil law systems characterized by procedural rigidity and prolonged litigation. This paper assesses the causal impact of Italy’s 2016 bankruptcy reform (Legislative Decree No. 59) on dispute duration and investigates its broader economic spillovers. The reform introduced digital communication tools, more stringent procedural deadlines, and streamlined mechanisms for creditor coordination to expedite insolvency proceedings. Utilizing an original micro-level dataset comprising over 40,000 civil cases from the Court of Catania, along with a Difference-in-Differences (DID) design incorporating extensive fixed effects and robustness checks, the study demonstrates that partial implementation in 2016 reduced dispute duration by approximately 12%. In contrast, full operational consolidation by 2018 yielded a substantially larger effect, decreasing proceedings by nearly 50%. These efficiency improvements remain evident even in complex, multi-stage cases. Beyond judicial performance, the analysis provides evidence of genuine economic spillovers: reform implementation correlates with reductions in provincial unemployment rates and a medium-term revival in firm investment. The findings imply that judicial reform transcends mere legal reform and constitutes a strategic institutional investment. By bolstering enforcement credibility and facilitating the reallocation of resources, procedural modernization can enhance creditor protection, sustain employment stability, and cultivate a more resilient economic environment.

ABSTRACT. Building on a novel dataset covering the universe of Italian magistrates from 2004 to 2025, we study how expected workload shapes the staffing of courts and how new entrants affect court productivity. We exploit the serial-dictatorship allocation mechanism through which prospective magistrates are matched to vacancies and outline a discrete-choice model of preferences over court attributes. We document that magistrates discount destinations with higher expected workload, and find evidence of assortative matching: higher-score magistrates are more likely to select more efficient courts. These patterns raise important concerns for the delivery of a key public good such as justice, especially because higher-score judges are themselves more productive. We show this using event-study evidence on the effects of incoming prosecutors on office-level productivity. New hires are associated with persistent productivity gains, with stronger effects for higher-ranked incoming prosecutors and for offices facing greater pre-event stress.

ABSTRACT. Abstract The digital economy and Artificial Intelligence (AI) are reshaping the economy in several ways: while consumers are often not mobile, the same is not true for production, which is dominated by a restricted number of multinational firms which operate in a sort of monopolistic market which makes it difficult to tax such activities so that the pressure to corporate tax competition is somehow reduced. Although these technologies might significant tools for carbon mitigation and resource optimization, they also carry a substantial physical footprint through energy consumption, resource extraction, and e-waste generation. One of the most effective instruments for Governments to internalise these externalities is the use of Pigouvian taxes. . In this paper we exploit an important equivalence between production and consumption Pigouvian taxes to show that the use of a combine production and consumption tax may in fact be a superior instrument from a policy point of view. We show that both a combined tax on production and consumption and a tax on consumption allows to reach the optimal autarky level, albeit with a different distribution of welfare. Taxing both production and consumption allows to exploit the gains from international trade since in this case the welfare of both countries is higher than in autarchy. However, a tax on production usually allows the country where goods are shipped to other countries to raise a Pigouvian tax that may be more in line with the actual environmental damage produced.

ABSTRACT. This paper estimates the size and drivers of the informal economy in 13 Central and Eastern European (CEE) countries over 2000–2022 by integrating the Currency Demand Approach (CDA) within a Pooled Mean Group – Autoregressive Distributed Lag (PMG-ARDL) framework. The long-run equation includes the tax burden, real income, the change in interest rates, an enforcement–income interaction, and internet penetration, with a COVID-19 dummy capturing pandemic shocks. Taxation is the dominant long-run driver of cash demand, while stronger enforcement, higher interest rates and broader digital-payment access act as significant restraints; the error-correction term confirms a stable equilibrium. Money velocity is estimated through a separate econometric specification that explicitly accounts for the feedback from shadow size, avoiding the bias of fixed-velocity CDA. Estimated shadow economy spans a wide range across the 13 CEE economies: about 10% of GDP in Czechia, Poland, about 30% in Romania, Bulgaria, Albania, Ukraine, Serbia, and above 30% in the easternmost economies (Moldova, Russian Federation, Bosnia-Herzegovina, North Macedonia).

ABSTRACT. We propose a new model of revenue-maximising lottery pricing under income-related heterogeneity and apply it to 13 years of UK National Lottery data. Our pricing rule shows how a state’s lottery power is shaped by heterogeneities in lotto demand and the income distribution and we study its efficiency and distributional implications. We find that UK lotto was regressive, underpriced by up to 20% for Saturday games and overpriced by up to 20% for Wednesday games. But annual revenue losses reached only 3% (£57m) and legislated pricing was less regressive than optimal on Saturdays and more regressive on Wednesdays. We contrast results from our model with those of two extensions: 1. a dynamic, rollover-induced, model of revenue-maximising pricing and 2. a model of social welfare maximisation with a lotto revenue target.

ABSTRACT. Objective: Personal Health Budget (PHB) has been recently provided to patients with severe mental illness within health policies supporting person-tailored mental health treatments based on individual unmet needs. However, evidence on the beneficial effects of PHB approach is still limited, even in Italy. The main aim of this research was to discuss preliminary data on the social and clinical benefits of adding PHB to standard pharmacotherapy in SMI patients during 2 years of follow-up. Methods: Patients (n=63) were individuals with SMI, aged 18-60 years, recruited from an adult mental healthcare service in Bologna. They completed the Health of the Nation Outcome Scale (HoNOS) at the beginning and end of our follow-up period. The Wilcoxon test for repeated measures was used to assess the longitudinal stability of outcomes. Multiple linear regression analysis was also performed to explore baseline predictors for HoNOS score improvement. Results: A significant longitudinal decrease in HoNOS total scores was observed. This was found to be associated with multiaxial intervention as the most robust predictor. Conclusions: Our results further support the useful implementation of the PHB approach for patients with SMI within the Italian network of public mental health services.

ABSTRACT. We examine how a change in the administrative governance shapes equity in access to Italy’s main universal non-means-tested cash long-term care benefit, the Indennità di Accompagnamento (IdA). We exploit a 2010 reform that strengthened central oversight of local disability-assessment commissions without altering formal eligibility rules. Using two decades of INPS data on a sample of Italian pensioners (around 159 millions of individual-years data), we combine a macro analysis of territorial dynamics with a micro analysis of post-reform socioeconomic selectivity. At the macro level, piecewise linear mixed-effects models show that the reform contributed to reducing coverage territorial heterogeneity. At the micro level, high-dimensional linear probability models indicate that tighter administrative oversight reshaped the socioeconomic composition of beneficiaries and the programme's compensatory role across different metrics of disadvantage.

ABSTRACT. This paper examines how generic entry affects prescribing choices across therapeutically close drugs. We study the Italian statin market in the wake of the loss of patent protection for atorvastatin, focusing on transitions between active substances within the same family. Using administrative data from Emilia-Romagna, we reconstruct individual prescription histories in the year before and after generic atorvastatin entered the market. We document strong persistence in prescription choices before the shock, followed by a marked increase in transitions towards atorvastatin. These patterns point to a reassessment of treatment choices when atorvastatin became cheaper for the National Health Service. When regulatory measures entail no financial consequences, as in the pre-break period, they risk being merely cheap talk, whereas, when patent expiration affects the cost-effectiveness of treatment, GPs seem to rebalance competing objectives and to become more sensitive to costs and more loyal to the NHS. Still, patients’ characteristics and health conditions, as well as financial incentives, play a pivotal role in shaping GPs’ prescribing decisions.

ABSTRACT. This paper documents the use of the Italy’s First-Home Guarantee Fund (Fondo di garanzia
per la prima casa) and estimates its effectiveness. The empirical analysis is based on an original
dataset based on data from the Italian Central Credit Register (Centrale dei Rischi) and covers
the period from 2011 to 2023. This time span includes the period following the introduction
of the guarantee covering 50 per cent of the mortgage principal, as well as the period during
which young borrowers under the age of 36 with a low ISEE (Italy’s means-tested indicator of
household economic conditions) were eligible for an enhanced guarantee covering up to 80 per
cent of the principal, together with tax incentives. The empirical design combines panel analysis,
Bartik-style specifications, and difference-in-differences estimates. The results suggest that the
strengthening of the scheme significantly increased mortgage originations to young borrowers,
with stronger effects in economically disadvantaged areas.

ABSTRACT. This paper shows that disposable income shocks affect local house prices through a mortgage credit-supply transmission channel. I provide evidence consistent with a mechanism in which negative income shocks weaken indebted households’ repayment capacity, increase mortgage delinquency risk, and are followed by a tightening of mortgage credit. Using U.S. state-level personal income tax (PIT) reforms approved between January 2003 and December 2019 as disposable income shocks, I compare adjacent counties on opposite sides of state borders. A 1 pp increase in the PIT rate lowers local house prices by 5.2\%. The same shock is followed by a 9.3\% increase in mortgage delinquency in reforming states, consistent with weaker repayment capacity among indebted households. Lenders respond by tightening mortgage credit: denial shares increase by 1.1 pp and the total amount of mortgage credit originated declines by 6.5\%. This tightening is concentrated among banks with greater exposure to the reforming state relative to more diversified banks, and exposed banks increase their cash holdings by 2 pp, consistent with defensive portfolio management. Finally, consistent with tighter underwriting constraints, the effects of credit tightening are strongest in the mortgage segment where borrowers are more likely to be close to those constraints. Although housing demand also weakens after negative disposable income shocks, exposure to borrowers close to underwriting constraints predicts where credit tightens and where house prices fall most, but does not predict demand-side responses. This pattern points to an amplification mechanism in which disposable-income shocks affect house prices through lenders’ credit-supply response.

ABSTRACT. This study examines how local demographic shocks affect residential mobility among households and individuals in an urban setting. Using longitudinal administrative population-register data for the period 2001–2022, we construct a panel at the census-section, household, and individual levels. Demographic shocks are identified as the unexpected component of local population growth, obtained from residuals of regressions that control for demographic characteristics, census-section and year fixed effects, and spatial dependence across contiguous areas. We estimate the dynamic response of residential mobility to these shocks using Local Projections. The outcomes include the probability of moving out of the neighborhood, moves toward more peripheral or more central areas, changes in travel time to the city center, and changes in urban zone. The analysis also assesses the persistence and predictability of shocks, tests for pre-trends, and explores spatial heterogeneity in the estimated effects. The paper contributes to the empirical literature by showing how unexpected demographic changes at a highly local scale may shape residential choices and the spatial structure of urban mobility.

ABSTRACT. The debate on environmental policy increasingly focuses on aligning private incentives with social objectives in imperfectly competitive markets. While traditional literature has centred on public-based mechanisms like taxes and subsidies, a growing strand emphasizes private-based mechanisms, particularly green consumerism, where consumer preferences can drive firms' adoption of clean technologies. Recent game-theoretic analysis shows that consumers' willingness-to-pay can lead to various market equilibria, from all-green to all-brown outcomes. This paper complements this analytical approach by developing an agent-based model (ABM) to study the dynamic evolution of a spatial market where firms, based on relative performance, decide whether to supply brown or green products to heterogeneous consumers. Our computational simulations confirm that all three market structures—all-brown, all-green, and mixed—can endogenously emerge depending on average green consumer preferences. Furthermore, we evaluate the effectiveness of three policy instruments: an environmental tax, a subsidy to green firms, and a subsidy to green consumers. We find that supply-side policies are more effective than demand-side subsidies. Specifically, an environmental tax ensures the fastest convergence to an all-green market, while a production subsidy is most effective at reducing the share of brown firms and consumers in mixed-market scenarios. By bridging game-theoretic insights with agent-based computational analysis, this paper provides a dynamic and policy-relevant perspective on the transition to sustainable markets.

ABSTRACT. This paper proposes a simple empirical framework to measure installation-level compliance costs under the EU Emissions Trading System (EU ETS). Using installation-level emissions data and institutional features of allowance allocation, we derive a tractable measure of compliance costs per unit of output. Applying the method to Phase III of the EU ETS (2013–2020), we document substantial heterogeneity in compliance costs across installations. A variance decomposition shows that most of the total variation arises from firm-level differences rather than sector or country effects. Linking installations to firm-level balance sheet data reveals that firm age is a key determinant: an additional ten years of age is associated with about 20 EUR per ton higher compliance costs on average, consistent with technological lock-in. We further provide a back-of-the-envelope calculation of the gains from permit trading, showing that access to the permit market generates economically meaningful cost savings at the installation level, averaging about \euro 76,000 per installation per year. These findings highlight the efficiency advantages of emissions trading systems and suggest that policies facilitating technological upgrading of older firms and firm turnover may complement carbon pricing.

ABSTRACT. As climate-change uncertainty rises, societies increasingly rely on new technologies to adapt. While such measures can improve resilience, they may also foster maladaptation—private responses that reduce damages locally while imposing external costs on society. We model a private agent that can invest irreversibly in a clean adaptation technology under stochastic climate risk. In a business-as-usual scenario, the agent instead relies on a lowercost practice that partially mitigates climate impacts but generates a negative externality (e.g., intensive soil fertilizer use). Adopting the clean technology achieves comparable adaptation performance without these social damages. We characterize the privately optimal investment trigger and robustness choice, and then study policy design when the government can subsidize adoption. The optimal subsidy balances fiscal costs against avoided external damages and accounts for how subsidies affect both the timing and design of private investment, reducing the likelihood of widespread maladaptation.

ABSTRACT. Does retirement change the marginal propensity to consume? This paper shows that it does.

Using quasi-experimental variation induced by statutory pension eligibility thresholds, the analysis implements a fuzzy regression discontinuity design to identify the causal effect of retirement on the marginal propensity to consume (MPC). Retirement increases the MPC by approximately 7.5 percentage points, corresponding to a rise of about 19 percent relative to individuals just below the eligibility threshold, and raises the probability of reporting high consumption responses.

The evidence points to leisure availability as a central mechanism. The retirement increase in MPCs is strongest among individuals facing greater work-related time scarcity before retirement and becomes more pronounced once working retirees are excluded from the analysis. By contrast, alternative explanations based on scale effects, liquidity constraints, balance-sheet conditions, or precautionary-saving motives receive limited support.

A simple life-cycle framework rationalizes these findings by linking consumption responses to leisure availability. The results imply that demographic structure shapes the aggregate transmission of fiscal policy through household consumption responses.

ABSTRACT. This paper examines how policy uncertainty influences retirement decisions. We develop a simple model in which individuals face a one-time choice between immediate retirement and continued employment until the statutory retirement age. In the absence of policy uncertainty, re- tirement decisions depend solely on the standard income–leisure trade-off. When future pension reforms are uncertain, however, individuals also take into account the perceived risk of increases in the retirement age or reductions in benefit generosity, and may choose to accept the offer in order to lock in current pension rules. Using administrative data from a large Italian bank that offered a one-time early-retirement scheme in 2017, we find that acceptance rates decline with the expected income loss but rise with the number of years to retirement. These patterns are consistent with workers using early retirement as an insurance against potential policy changes, underscor- ing the importance of incorporating behavioural responses to policy uncertainty in the design of pension systems. Our findings suggest that individuals with an average annual income of €35,000 are willing to pay an annual premium of €481 to insure against the probability that the pension system is reformed.

ABSTRACT. This paper studies how information on demographic trends affects youths’ perceptions of the labor market and migration intentions. Using a survey experiment with over 900 high school students in Rome, we randomly provide information on population aging in Italy to the treatment group, while the control group receives unrelated information. We find that exposure to aging information worsens perceptions of job opportunities for young people, with limited effects on inequality perceptions. Heterogeneity analysis, motivated by Eurobarometer evidence, reveals that the treatment negatively affects female students’ perceptions, while improving males’ views on job opportunities and reducing their migration intentions.

ABSTRACT. Between-country GDP inequality in Europe declined markedly from the early 1990s to the eve of the global financial crisis, but this convergence has since stalled and partly reversed, pointing to a structural change in the macroeconomic forces shaping cross-country disparities. This paper investigates the contribution of macroeconomic factors to that evolution by decomposing between-country GDP inequality into expenditure-side components, income-side components, and demographic and productivity-related channels. Using annual OECD National Accounts data for 17 European countries over 1990–2024, we implement a non-parametric two-stage decomposition strategy that combines a counterfactual Shapley procedure with a Shorrocks factor decomposition. The estimation strategy we propose provides an exact accounting of inequality in terms of interpretable factor contributions, separates demographic effects from per-capita GDP dynamics, and is well suited to settings in which macro aggregates are jointly determined. Our findings show that the decline in inequality up to the late 2000s was largely driven by wage- and consumption-led convergence, underpinned by narrowing productivity gaps. Since the 2000s, however, growing dispersion in exports has increasingly acted as the main source of divergence, while profits have progressively replaced wages as the main income-side channel through which productivity differences translate into inequality. These results suggest that, in a changed macroeconomic environment, reducing between-country inequality requires not only policies that reconnect wage growth to productivity and sustain domestic demand, but also interventions on the profit side and greater coordination of external balances.

ABSTRACT. This paper studies gender inequality in intra-household resource allocation in Italy over the period 1998--2024. We estimate women’s relative resource shares—the fraction of total adult household expenditure allocated to women—using the framework of Blundell et al.\ (2025), which identifies resource shares from Engel curves for assignable goods.

Our empirical analysis uses 28 annual cross-sections from the Italian Household Budget Survey (ISTAT) and exploits gender- and age-specific clothing expenditures as assignable goods. We document the evolution of within-household gender inequality over time and examine heterogeneity across household types, geographic areas, and income groups. We further decompose changes in resource shares into components driven by shifts in household characteristics and changes in the resource-share function.

Women's relative resource shares are estimated to have increased, on average, from about 45\% to about 55\% over our investigation period, rising from disparity to above parity with men. We find that changes in relative resource shares of childless women are mainly driven by changes in women's characteristics over time. Instead, rising relative resource shares of mothers are mainly driven by changes in women's bargaining power within the household.

ABSTRACT. Population ageing is reshaping European societies and raising important questions about the distribution of well-being among older adults. This paper analyses multidimensional inequality among individuals aged 50 and over using data from the Survey of Health, Ageing and Retirement in Europe (SHARE). Moving beyond traditional income-based approaches, we construct a multidimensional inequality index combining dimensions related to economic resources, education, health and social participation. Multidimensional inequality is measured using inequalities indices that allow decomposition across dimensions. Preliminary results suggest that disparities in social participation account for a substantial share of overall inequality among older Europeans. These findings highlight that inequality in later life increasingly reflects differences in technological and social resources in addition to traditional economic disparities.

ABSTRACT. In the wake of anti-immigration backlash, countries are increasingly imposing restrictions on high-status migrants. This study investigates how such policies influence the choices of domestic migrant-background populations, focusing on the 2014 Swiss popular initiative ``Against Mass Immigration.'' Approved by 50.3\% of voters, the initiative imposed new employment and residence restrictions on EU migrants. Using the full Swiss registry data, we analyze migrant responses in terms of residential patterns, marital decisions, fertility, and assimilation choices. Difference-in-differences estimates comparing temporary EU migrants to other immigrants unaffected by the policy allow us to isolate the economic effects from those of increased perceived discrimination. Results suggest that the initiative generated conservative life choices: it reduced internal migrant mobility, fostered desegregation, and reduced fertility. Yet we find no evidence of either reduced or increased cultural assimilation efforts. Our findings contribute to debates on restrictive immigration policy, with insights on how negative economic incentives affect migrant integration.

ABSTRACT. This paper investigates the causal effect of language assimilation on the cultural identity and educational aspirations of teenagers using a well-established instrumental-variable strategy based on age at migration and linguistic proximity between origin and host languages. While the instrument significantly reduces language fluency scores for both boys and girls, I find that language skills do not affect cultural identity nor educational aspirations outcomes for girls. In contrast, boys exhibit a lower probability of feeling like foreigners and a higher likelihood of wanting to live in the host country in adulthood, as well as an increased desire to pursue further education.

ABSTRACT. This paper studies borrower discouragement among immigrant households in Euro area credit markets. Using harmonised microdata from the Household Finance and Consumption Survey (HFCS), we show that immigrant households are more likely than comparable native households to refrain from applying for credit because they expect rejection. The gap is concentrated among migrants from outside the Euro area and declines with years spent in the host country, consistent with the role of financial-history portability and familiarity with host-country credit institutions. As complementary evidence, we show that immigrant applicants are also more likely to be denied credit or to receive less credit than requested. The results highlight borrower discouragement as an important pre-application margin of financial inclusion and show that applicant-based analyses miss a form of credit-market exclusion that occurs before any lender decision is observed.

ABSTRACT. This paper proposes a novel methodology to identify the geographic market of local newspapers when information on their diffusion is unavailable or insufficiently granular. We illustrate the methodology using historical data from 154 newly digitized newspapers published in Italy between 1919 and 1922. Combining machine learning-augmented optical character recognition techniques, multi-way fixed-effect regressions, and GIS tools, our approach allows us to estimate markets from news content. Text-based location of newspaper markets significantly improves over assuming that market boundaries coincide with administrative aggregations. We discuss how our technique strengthens the use of newspapers as a detailed source of historical information.

ABSTRACT. Are political careers facilitated by family networks? We address this question using a novel dataset assembled from various sources and combining information on the political careers of British MPs with information about the family trees of the British aristocracy. Our data span almost two centuries (1831-2023) and document the evolution over time of the background of British MPs, their social origin (in particular whether aristocratic or not), education (university and school attended), profession and political experience. Using information on family trees we measure MPs' distances from relatives with nobility titles and their centrality in the aristocratic network. We use these indicators to assess the role of aristocratic ancestry and connections in determining election and career progression. These indicators, calculated at birth, explain a large part of an early start advantage, i.e. the lower age of entry into the House of Commons. We document how connected MPs put less effort in legislative activities, by participating less in voting and speaking less frequently. Progression into the Cabinet is facilitated by aristocratic connections but this is mostly an indirect effect via longer experience (in years) accumulated in parliament due to the early start advantage. We also show how the decline of the aristocracy in British politics is mostly due to its decline as a source of political personnel for the Conservative party.

ABSTRACT. This paper examines how within-city government amalgamations affect democratic representation and electoral behavior by analyzing the 2013 reform of sub-municipal units in Rome, which reduced their number from nineteen to fifteen. The empirical analysis exploits electoral precinct-level data covering over 2,000 precincts and all sub-municipal and municipal elections between 2001 and 2021. Using a boundary discontinuity design, we find no effect of sub-municipal consolidation on voter turnout. In contrast, preference-based outcomes respond markedly, with a temporary increase in proactive political engagement, a persistent concentration of preference votes, and a decline in support for local lists.

ABSTRACT. This paper investigates the correlates of defense burden sharing in the European Union over the period 1980–2023. We argue that institutional constraints and strategic endowments systematically shape how member states distribute the costs of collective defense. Using a panel of 28 EU countries, we combine machine-learning–based variable screening with fixed-effects panel estimation and explicitly distinguish between relative (between-ally) and domestic (within-ally) measures of defense effort. We introduce four underexplored correlates — constitutional complexity, distance from the colonial past, nuclear latency, and arms technology level. Findings show that countries possessing nuclear latency consistently devote lower relative and domestic resources to defense, indicating a plausible substitution effect. Constitutional complexity reduces relative burden contributions, possibly reflecting higher transaction costs. Colonial legacy shapes domestic defense prioritization, while arms technology exhibits weaker effects. The findings remain robust across robustness checks. Notably, excluding the leading states yields sharper estimates, suggesting that their structural commitments may partially overshadow baseline relationships. Replication on the NATO sample further supports the external validity of the findings.

ABSTRACT. We study how local exposure to Ukrainian refugees shapes the political attitudes and immigration preferences of British citizens, exploiting the design of the Homes for Ukraine sponsorship programme. Launched in March 2022 until 2025, the scheme relied on voluntary private hosting where British citizens hosted refugees in their homes, making pre-existing local spare bedrooms availability a credible source of exogenous variation in the intensity of refugee exposure. We construct a Bartik-style shift-share instrument based on the share of under-occupied houses in 2021 and merge it with data from waves 14--29 of the British Election Study Internet Panel (2014-2024). Our results reveal a pronounced temporal dynamic: in 2022, residents of more exposed local authorities displayed significantly more favourable attitudes toward immigration, consistent with an initial solidarity response. However, this positive effect erodes substantially as exposure grows. Investigating potential mechanisms, we document that local authorities receiving larger numbers of refugees experienced significant increases in rental prices, consistent with refugees gradually transitioning from sponsored accommodation into the private rental market. These findings suggest that backlash toward immigration is more likely to emerge when host communities perceive concrete pressures on housing markets and local public services, and that accompanying policies mitigating such pressures may be essential to sustain long-run popular support for refugee integration.

ABSTRACT. This paper investigates intra-industry trade in major conventional weapons (MCW) among European Union countries over the period 1958–2022. We compute the Grubel–Lloyd index at the country level to measure the extent of intra-industry trade and analyse its distribution across EU members. The empirical analysis relies on correlated random effects fractional response models to examine the economic, political, and security-related determinants of intra-industry trade. The results suggest that higher market concentration is associated with lower intra-industry trade, while labour productivity, capital formation, and technological development tend to increase it. Political orientation and external security threats also appear to play a significant role.

ABSTRACT. Economic liberalization is widely expected to generate demand for political voice—a conjecture associated with Friedman (1962), who argued that competitive capitalism is a necessary, though not sufficient, condition for political freedom. Yet authoritarian regimes routinely liberalize markets without destabilizing. This paper asks how. Two channels can neutralize the political content of economically induced contention: coercion (repressing emergent dissent) and absorption (resolving grievances through state-controlled institutions). I distinguish them using China’s staggered rollout of Pilot Free Trade Zones (FTZs) across 64 prefectures (2013–2023), which introduced commercial arbitration, enforceable contracts, and regulatory streamlining without any expansion of political rights. Using Callaway–Sant’Anna estimators, I find that FTZ activation raises total contention by roughly 32% (p < 0.05), yet this contention is neither politicized nor repressed. The share of rights-oriented protest does not increase, and state repression shows no detectable change (narrow and broad measures, null under OLS, Poisson, and Callaway–Sant’Anna). Instead, the grievances are absorbed: China Labour Bulletin data show that FTZ arbitration reduces institutional labor demands, i.e. disputes over social insurance, pensions, and contract enforcement, by 19% (p < 0.05), while non-resolvable wage-arrears protests persist. The reform channels contention into state-controlled institutional arenas rather than into political demands or visible repression. Consistent with an institutional-novelty mechanism, the effect is concentrated where formal commercial norms are genuinely new (non-treaty-port cities). The findings point to institutional absorption—rather than a visible coercive response—as the most consistent explanation for how commercial reform is rendered politically inert in this authoritarian setting.

ABSTRACT. We study the environmental impact of artificial intelligence (AI) using a novel dataset that links measures of AI penetration, the location of data centers and power plants, and CO2 emissions across US commuting zones between 2002 and 2022. Our analysis yields four main findings. First, exploiting a shift–share identification strategy, we show that localities more exposed to AI experience relatively faster emissions growth. Second, decomposition results indicate that scale effects dominate, while changes in industrial composition exert at most a weak mitigating effect; at the same time, electricity generation becomes more carbon intensive. Third, AI penetration raises dependence on non-renewable electricity. Fourth, proximity to data centers is a key driver of this effect, as nearby power plants shift toward greater fossil fuel use. These findings suggest that, absent a rapid decarbonization of power generation, the diffusion of AI is likely to exacerbate environmental externalities through the energy demand of data centers.

ABSTRACT. Whether the informal wage gap reflects labor market segmentation or competitive sorting on productivity remains a central question in development economics, with direct implications for policy design. Existing evidence is largely based on average wage comparisons that may conceal economically meaningful heterogeneity. We revisit this debate by estimating the full distribution of individual-level informal wage penalties in urban South Africa. Using nationally representative data from the National Income Dynamics Study and a causal machine learning framework---double machine learning with doubly robust scores---we flexibly control for a high-dimensional set of worker characteristics and recover individualized treatment effects without pre-specifying subgroups. We find an average informal wage penalty of approximately 21\%, consistent with the existing literature on developing economies. However, this mean masks substantial heterogeneity: individual penalties range from near zero to almost 50\% at the 1st percentile, while a small fraction of workers faces no penalty or even a mild premium. Crucially, the dimensions along which heterogeneity is most pronounced are identified endogenously by the machine learning algorithm rather than imposed ex ante by the researcher. A classification analysis reveals that the largest penalties concentrate among workers belonging to the African racial group and speakers of minority languages, while White and Coloured workers, Afrikaans speakers, and Western Cape residents experience comparatively smaller gaps. These patterns are difficult to reconcile with a frictionless competitive model and point toward search frictions, unequal outside options, and employer market power as plausible---though not uniquely identified---mechanisms shaping wage inequality across both formal and informal sectors.

ABSTRACT. Social security systems feature complex tax and benefit rules and offer limited individualized counseling. This complexity restricts the ability of social policies to reach their intended populations. We distinguish between awareness of rules (general, static information) and computational complexity (personalized, interactive information) and test their respective roles in transmitting financial work incentives. We develop an online benefit simulator and evaluate its impact in a randomized field experiment. Providing general information does not affect behavior, whereas reducing computational complexity through the simulator increases re-employment among social security claimants by 3.75 % after one year. Simulator data reveal that financial incentives to work are non-monotonic in hours, difficult to compute, and highly heterogeneous across claimants. Personalization thus appears more effective than salience in complex policy environments, highlighting the high potential returns to digital counseling tools in social insurance.

ABSTRACT. This study investigates the electoral consequences of environmental taxation in the European Union (EU-27). Using panel data on parliamentary elections from 2000 to 2022, we construct a measure of incumbent electoral performance based on changes in vote shares of governing parties across elections. We examine the relationship between environmental taxation and electoral outcomes and explore whether this relationship is conditioned by income inequality and social expenditure. The results show that increases in environmental taxes are associated with a reduction in incumbent electoral support. However, this effect is not uniform across contexts. While we find limited evidence that inequality amplifies electoral backlash, the results provide suggestive evidence that higher levels of social expenditure may mitigate the political cost of green taxation. These findings suggest that the political feasibility of environmental taxation depends crucially on redistributive policies, highlighting the importance of compensation mechanisms in sustaining public support for the green transition.

ABSTRACT. This paper investigates whether fiscal rules designed to contain public debt may unintentionally weaken innovation and long-run economic growth. We develop a quality-ladder Schumpeterian growth model in which government-financed R&D subsidies promote technological progress, while public debt evolves under a fiscal rule linking expenditure to the debt-to-output ratio. The model predicts that fiscal rules can have opposite effects on growth depending on their design. Rules that reduce public spending when debt rises may depress innovation through a negative market-size effect, weakening growth and generating macroeconomic instability. In contrast, fiscal frameworks that preserve or expand spending in high-debt periods can support innovation, strengthen growth, and improve debt sustainability. Under plausible conditions, such a response is shown to be necessary for saddle-path stability. To assess these predictions, we propose an empirical strategy based on heterogeneous-coefficient panel models that examines the relationship between fiscal rules, government spending, and R&D activity across countries. The analysis aims to provide new evidence on how fiscal institutions shape innovation-driven growth in high-debt economies.

ABSTRACT. This paper explores the hypothesis that tax evasion is a critical determinant of income inequality in Italy. While the relationship between fiscal systems and inequality is widely studied, the specific contribution of tax evasion to the unequal distribution of wealth and income remains underexplored. Drawing from recent empirical studies and theoretical models, we argue that tax evasion disproportionately benefits high-income individuals and undermines the redistributive capacity of progressive tax systems. We estimate an OLS and GLS panel model using annual data from the Italian provinces (NUTS-3 level) for the 2004–2012 period. Although tax evasion is often treated as a symptom of weak governance or complex tax codes, our analysis suggests it should be viewed as an active driver of inequality. The paper develops a conceptual framework integrating fiscal evasion into broader inequality models and outlines a methodology for future empirical testing. We argue that greater attention should be paid to the interaction between wealth concealment and redistributive inefficacy, particularly in countries, like Italy, where taxation plays a major role in income equalization. Preliminary data collection and methodological design are presented, with early results to follow. Ultimately, this study contributes to a more nuanced understanding of the mechanisms that perpetuate inequality and calls for policy interventions targeting the opacity of wealth and the effectiveness of tax enforcement mechanisms across diverse institutional settings.

ABSTRACT. VAT gaps remain a major fiscal problem for many European Union countries. They reduce public revenues, weaken fiscal capacity and make budget adjustment more difficult, especially in countries already facing high deficits. This paper studies why VAT gaps remain persistently high in some European countries and whether EU countries are gradually moving toward similar levels of VAT compliance. The analysis covers 25 European countries over the period 2000–2023. It combines convergence analysis, club convergence methods, variable selection, ordered probit models, Blinder–Oaxaca decomposition and exploratory panel regressions. The purpose is not to estimate one single causal effect, but to identify whether VAT gaps are mainly temporary administrative problems or whether they reflect deeper structural differences across countries. The results show that VAT gaps have become less dispersed over time, but this does not mean that all countries are converging toward the same compliance standard. The cross-country standard deviation of the VAT gap fell from 10.06 percentage points in 2000 to 6.70 in 2023. However, the Phillips–Sul convergence test rejects a single common convergence path. Instead, countries are grouped into four convergence clubs, while Romania remains outside the clubs as a divergent case. This means that VAT compliance in Europe has improved unevenly: some countries move together within groups, but high-gap countries do not fully catch up with better-performing countries. To interpret these differences, the paper classifies countries into three compliance regimes: high compliance, medium compliance and a low-compliance trap. These groups are not only statistical categories. They have clear economic and institutional profiles. The high-compliance group has an average VAT gap of 6.9%, while the trap group has an average gap of 23.8%. Countries in the trap group also have lower GDP per capita, weaker governance indicators and larger shadow economies. The estimated shadow economy rises from 16.9% of GDP in the high-compliance group to 27.9% in the trap group. The results suggest that VAT gaps are closely linked to broader development and institutional conditions. Variable-selection results point to a governance-income cluster as the most stable set of factors associated with VAT gaps. In the ordered probit model, GDP per capita is the strongest predictor of regime membership. Institutional quality moves in the expected direction, but its separate effect is harder to estimate precisely because governance and income are closely related. The evidence therefore points to a broader development-institutional gradient: richer countries with stronger institutions are much more likely to belong to low-gap compliance regimes. The Blinder–Oaxaca decomposition reinforces this interpretation. The average VAT gap difference between the high-compliance and trap regimes is 16.86 percentage points. Only 3% of this difference is explained by observable differences in income and institutional quality evaluated at common returns. The remaining 97% lies in the structural or unexplained component. This suggests that high-gap countries are not separated from low-gap countries only because they have different observed characteristics, but also because those characteristics translate differently into compliance outcomes. In policy terms, this points to persistent compliance environments rather than short-term revenue administration problems alone. Exploratory panel regressions further show that the relationship between VAT gaps, informality and governance differs across compliance regimes. These estimates are not interpreted as causal proof, but they support the idea that the same policy lever may not work in the same way across all countries. Overall, the paper shows that lower dispersion in VAT gaps should not be confused with full convergence. European countries have not moved toward a single VAT compliance path. Instead, they remain grouped into different compliance regimes, shaped by income, governance quality and informality. The policy implication is clear: countries with persistently high VAT gaps may need more than technical fixes, digital reporting tools or short-term administrative reforms. They require broader fiscal-capacity strategies that address institutional quality, informality and the structural conditions that sustain low tax compliance.

ABSTRACT. This paper studies deterrence when individuals are uncertain about enforcement and learn only through their own offending experience. We develop a dynamic model in which potential offenders repeatedly decide whether to violate the law while updating beliefs about the probability of detection through Bayesian learning. Since informative signals arise only after violations, learning is endogenous and asymmetric: those who stop offending cease to acquire information, while those who continue gradually learn the true enforcement environment. Enforcement uncertainty has two contrasting effects. In the short run, it may reduce deterrence because offending has informational value and forward-looking individuals optimally experiment. In the long run, however, asymmetric learning generates persistent heterogeneity in beliefs and behavior. Some individuals stop offending after unfavorable outcomes and retain pessimistic beliefs, while others continue offending and converge toward accurate beliefs. This endogenous stopping mechanism generates over-deterrence even though Bayesian beliefs remain unbiased on average. Under continuous stochastic learning, the long-run distribution of beliefs coincides with the outcome of an optimal Bayesian persuasion problem.

ABSTRACT. Background. The growing mental health gap between the most and least economically fortunate people globally requires urgent and innovative financial solutions, even in industrialized countries. Microcredit was designed to alleviate poverty by promoting access to loans for people excluded from the financial system. Although studies have demonstrated the link between debt and mental health, the mechanisms underlying the impact of microcredit on psychological well-being have been little explored, and the few published findings published are contradictory. The aim of this narrative review was to fill this gap by synthesizing current evidence on microcredit and its effects on mental health. Methods. A systematic search was conducted on Web of Science, PubMed, PsycINFO, and EMBASE from their inception up to March 2026, using the keywords “microcredit” AND “mental health”. Only articles that qualitatively or quantitatively assessed parameters related to microcredit and mental health were included in the review. Results. Nine of the initial seventy-five articles met eligibility criteria. Our findings first suggest that microcredit programs are not immune to the negative effects of debt on mental health (i.e., depression, anxiety, and stress). However, microcredit also induced positive changes in several mental health indicators (e.g., self-esteem, life satisfaction, and social support). The crucial limitations of these studies were related to their almost exclusively cross-sectional design and the lack of surveys in industrialized countries, which did not allow for the examination of bidirectional causality and the generalization of findings. Conclusions. Microcredit is an approach to financial inclusion that can improve mental health and enable participants to maintain social connections, although receiving it has proven stressful for some individuals. Microcredit can potentially play a role as a public health initiative to support alternative economic spaces serving the “financially excluded”, including those with mental disorders, even in industrialized countries.

ABSTRACT. Introduction

Digital health technologies are rapidly transforming healthcare delivery, nonetheless their large-scale implementation remains fragmented. Adoption is constrained by system-level barriers such as infrastructural limitation, financial constraints and organisational complexity but also by patient-level factors related to lack of digital skills, privacy concerns, trust in new technologies and perceived quality of care. Understanding the drivers of these barriers is essential for designing health policies that ensure an equitable and efficient integration of digital health technologies into novel healthcare practices. Therefore, this study examines individual-level predictors of barriers to patient facing digital health technologies adoption. While previous research has documented key barriers to digital health access, fewer studies have examined which individual-level characteristics drive heterogeneity in these barriers, and even fewer have considered income and contextual factors, limiting our understanding of what shapes access and utilisation of these technologies.

Methods

Leveraging novel survey data from 2,769 participants in the Emilia-Romagna region in Italy, we estimated a series of probit, negative binomial, hurdle and generalised ordered logit models to examine: i) the likelihood of reporting any barrier, and ii) a summative index which counted the total number of barriers experienced by each participant iii) the probability of transitioning across ordered categories of increasing barrier severity. Models were adjusted for socio-demographic and socio-economic characteristics, health status and contextual factors such as province of residence and degree of urbanisation. The modelling approach allowed us to address issues of overdispersion and excess zeros in the count data, while relaxing the assumption of homogeneous effects of covariates across outcome thresholds, providing a robust framework to capture the predictors of both the occurrence, intensity, and typology of barriers.

Results

At the extensive margin, being aged 55 or more, being in the lowest income quintile, and having below-average health increase the probability of reporting any barrier by 5.6 pp (p<0.001), 12.2 pp (p<0.001), and 3.5 pp (p<0.05), respectively. These patterns were confirmed in the count and ordered models examining the number, intensity, and severity of barriers. No significant effects were detected for degree of urbanisation or province-level indicators.

Discussion

Findings established a clear socioeconomic gradient which confirmed the less educated individuals, those from the lowest income quintiles and those living alone more likely to experience barriers to accessing patient-facing digital health services. Gender differences were only significant at the intensive margin. We found limited evidence of a geographical digital health divide, as factors such as degree of urbanisation and province of residence showed no significant effect. These results can inform the design of digital health policies aimed at improving the uptake and utilisation of patients facing digital health technologies by targeting support to those groups identified more at risk of experiencing barriers to access.

ABSTRACT. We present the first systematic survey of assistive and digital technology (AT) adoption in residential care facilities for dependent older people (Case Residenza per Anziani non autosufficienti, CRA) in the Italian region of Emilia-Romagna. The region hosts 370 CRA with roughly 21,650 beds, a stock larger than the regional hospital-bed capacity, which makes residential long-term care a structural component of the regional health system. Using a REDCap questionnaire administered in October–November 2025 with the support of the Regional Government, we collected responses from 160 of 370 facilities (response rate 43.2%). The picture that emerges is a two-speed sector: information-management and basic-safety technologies are pervasive (electronic health records and communication tools in 90–95% of facilities, safety systems in 82%), second-generation technologies (sensors and monitoring, telemedicine, smart-home) reach 25–52%, while the most advanced technologies (virtual reality, care robots, socially assistive robots) are adopted by fewer than 6% of facilities. Once basic technologies are accounted for, Poisson regressions show that territorial differences—rather than facility size or ownership—are the main correlate of intermediate-technology adoption, with a robust urban–inland gradient. Adoption is mostly motivated by quality-of-care improvement (81%) rather than cost reduction (8%). The results provide a structured, replicable empirical baseline for regional long-term-care policy and for subsequent inferential and record-linkage analyses.

ABSTRACT. Emergency department crowding raises concerns about treatment quality, but field evidence cannot cleanly determine whether observed deterioration reflects a capacity effect or a failure to prioritise patients by clinical needs. We address this issue using a controlled laboratory experiment in which subjects allocate treatment resources across heterogeneous patients under exogenously varied capacity constraints. Subjects act as physicians deciding resource intensity at the bedside, after triage has already been completed. Laboratory sessions are ongoing and will be completed by July 2026. Preliminary results suggest that treatment intensity falls significantly once capacity crosses a binding threshold, with heterogeneous effects across patient urgency types and a non-trivial incidence of intensity-ordering violations under binding constraints. A between-subjects information treatment identifies whether decision-support benchmarks improve prioritisation under capacity pressure.

ABSTRACT. Institutional quality is increasingly recognized as key determinant of individual mobility decisions and regional development dynamics. This paper examines how this dimension shape the medium- and long-term migration trajectories of young individuals from Southern Italy, a context marked by persistent socioeconomic disparities and uneven institutional performance. Using rich longitudinal microdata and discrete-time competing risks models, we track the post-migration pathways of youths who moved from the South to the Centre-North in the early 2010s. Over the subsequent decade, they faced multiple potential destinations: remaining in the North, returning to the South, or moving abroad. We show that institutional quality — measured in the municipalities of origin and in the initial destination — exerts a significant and heterogeneous influence on these transitions. Stronger institutions reduce the likelihood of return migration and increase the probability of onward migration to international destinations. More generally, the findings underscore the importance of local institutional environments for retaining and attracting human capital.

ABSTRACT. Despite growing literature on the impact of migration shocks on labour market outcomes, the effect on perceived quality of urban and cultural amenities remains underexplored. This is the first paper to examine the causal effect of migration shocks on place perception using big social media data from TripAdvisor ratings of cultural attractions at the municipality level. We exploit the opening of refugee centres as an exogenous shock to migration inflows in Italy. We apply a staggered difference-in-differences (DiD) estimator using a panel of 3,929 Italian municipalities from 2012 to 2022. Our findings reveal heterogeneity by reviewer origin, with a positive effect among foreign reviewers. This suggests that the negative political narratives around migration shocks may not necessarily result into a negative perception of place.

ABSTRACT. Heritage sites, often considered as public goods, generate a range of collective benefits arising from identity, amenity and cultural values. Therefore, their preservation mostly relies on government intervention. Economic development, however, could have both positive and negative effects on heritage site conservation, suggesting a nonlinear relationship similar to the Environmental Kuznets Curve. On one side, these forms of public goods provision are generally stronger in more affluent economies with greater resources and institutional capacity. At the same time, economic growth might exert several pressures that directly affect the integrity and vulnerability of heritage resources. We address this question by investigating whether the relationship between economic development and threats to World Heritage Sites displays a Heritage Kuznets Curve. Drawing on threats recorded in UNESCO State of Conservation reports for the period 2000-2021, we construct a country-level panel capturing the incidence of human-induced and management-related threats across nine categories. We test whether the relationship between per capita income and an index of threat to cultural heritage exhibits systematic non-linearities. To account for cross-country heterogeneity in both baseline threat levels and income effects, we adopt a mixed-effects modelling framework that allows for heterogeneous intercepts and slopes across countries. The findings provide evidence of a Heritage Kuznets Curve for four categories of threats: management, building development, pollution and resource extraction.

ABSTRACT. This paper investigates the impact of PRIN competitive research funding on the spatial distribution of academic performance across Italian universities over the period 2007–2020, with a specific focus on regional inequalities and territorial dynamics. While international literature has extensively examined the relationship between research funding and scientific productivity, providing mixed and often inconclusive evidence, considerably less attention has been devoted to the spatial and distributional implications of competitive funding schemes. This gap is particularly relevant in countries characterized by persistent regional divides, where place-based structural differences may shape both access to funding and the ability to transform resources into scientific output. Italy provides a particularly suitable case study. Its higher education system operates within a centralized institutional framework, yet it is embedded in a territorially heterogeneous context marked by pronounced North–South disparities in economic development, institutional quality, research infrastructure, and innovation capacity. These structural differences may influence both the allocation of competitive research funding and its effectiveness, potentially reinforcing cumulative advantage mechanisms. Descriptive evidence shows that PRIN funding is highly concentrated in Central and Northern regions, while Southern regions—despite comparable levels of participation in research collaborations—receive a significantly smaller share of financial resources. This pattern suggests that inclusion in research networks does not necessarily translate into equal access to funding intensity. To empirically investigate these dynamics, the paper combines administrative data on PRIN-funded projects with bibliometric indicators of academic performance derived from Scopus. The analysis is conducted at the university–field–year level, allowing for a detailed examination of both institutional and disciplinary heterogeneity. The empirical strategy adopts a continuous treatment framework, where the intensity of PRIN funding—normalized by faculty size—serves as the main explanatory variable. Identification relies on a generalized propensity score approach combined with fixed effects models, enabling the estimation of causal effects under selection on observables. In addition, spatial analysis techniques are employed to capture potential spillovers, clustering effects, and spatial dependence in research performance. The main objective is to assess whether competitive research funding generates spatially heterogeneous effects and whether it contributes to reinforcing or mitigating existing territorial inequalities. In particular, the paper tests whether PRIN funding strengthens cumulative advantage dynamics, favoring already well-performing universities and regions, or whether it promotes convergence in research outcomes across institutions and territories. The analysis also considers the potential trade-offs between efficiency and equity embedded in merit-based allocation mechanisms. From a policy perspective, the findings contribute to ongoing debates on the design of research funding instruments. While competitive schemes are often justified on efficiency and excellence grounds, their territorial implications remain less explored. The results of this study provide evidence on whether such instruments inadvertently amplify regional disparities in contexts characterized by structural heterogeneity. More broadly, the paper highlights the importance of integrating place-sensitive considerations into higher education and innovation policies, suggesting that purely meritocratic funding mechanisms may need to be complemented by targeted interventions to ensure a more balanced and inclusive development of national Research systems.

ABSTRACT. This paper provides causal evidence on the long-term welfare consequences of marital dissolution for health, economic security, and lifetime labour supply in later life. We use data from the Survey of Health, Ageing and Retirement in Europe and exploit the staggered introduction of unilateral divorce laws (UDLs) as an instrumental variable for divorce. We find that divorce significantly improves long-term self-reported health and life satisfaction, while effects on specific clinical indicators (depressive symptoms, BMI, and frailty) remain statistically imprecise. The most substantial adjustment is observed in the labour market: conditional on having ever worked, divorce increases total career length by 5.7 to 8.6 years, representing an 11 to 15 percentage point increase in the fraction of adult life spent working. These findings suggest that legal frameworks governing family dissolution permanently reshape lifetime economic and health trajectories well into old age.

ABSTRACT. Does public expenditure reduce the burden of unpaid work? Is there a connection between the provision of public infrastructure and individual time allocation? We identify positive and negative spikes in public expenditure within the United States care regime between 2004 and 2021, defined as episodes of extraordinary expansions or contractions in public investment in education, healthcare, and public welfare. Using a dynamic difference-in-differences design that exploits these spending shocks, we provide new evidence that public investment significantly affects individual time allocation in an asymmetric way. Positive spending spikes increase women’s paid working time and reduce the burden of domestic and care responsibilities. Negative spikes lead to a rise in female unpaid working time. These effects highlight the importance of welfare-state institutions in shaping gender inequalities in time use. Gender budgeting, specifically accounting for diverse transmission channels during periods of expansion and recession, is advocated.

ABSTRACT. Population ageing has increased reliance on informal caregiving, rais- ing concerns about its implications for caregivers' mental health and qual- ity of life. This paper examines the relationship between informal care- giving and multiple dimensions of psychological distress, with particular attention to territorial context. Using data from a health survey conduc- ted in the metropolitan area of Genoa, Italy, we analyse seven ordinal mental health outcomes through ordered logit models recalling interac- tions between caregiving status and municipality type (urban, coastal, and rural). Results show that individuals living in rural municipalities exhibit lower baseline odds of psychological distress than urban resid- ents. However, this advantage is substantially reduced among caregivers in rural areas, who experience signicantly worse mental health outcomes than non-caregivers. No comparable dierences emerge for coastal mu- nicipalities. Marginal eects indicate that caregiving is associated with pronounced increases in the probability of distress in rural settings, espe- cially for sleep problems, overeating, and low self-esteem. These ndings underscore the importance of spatial context in shaping the mental health consequences of informal caregiving.

ABSTRACT. Population aging is transforming electoral coalitions, fiscal priorities, and public policy across advanced and middle-income economies. This paper develops a political-economy framework linking demographic aging to political support for military spending. Integrating cross-national macro evidence, micro-level opinion research, and institutional mediation mechanisms, we show that aging does not produce a universal decline in defense spending. Instead, demographic change operates through the median voter, whose relative valuation of social protection versus security determines fiscal trade-offs. The effect of aging on military spending is conditional on income level, threat environment, alliance commitments, regime type, and elite securitization strategies. To test our theoretical framework, we draw on individual-level data from five waves of the Eurobarometer (2021–2024), pooling approximately 185,000 respondents across all EU member states. The empirical results consistently indicate a non-linear, U-shaped life-cycle pattern: support for European defence policies declines from early adulthood through middle age, reaching a minimum around age 45, and increases again among older respondents. These results are robust across four distinct measures of defence support and across survey waves spanning different geopolitical contexts.

ABSTRACT. This paper presents the first estimate of an actuarial balance sheet (ABS) for Italy’s mandatory public pension system, covering old-age, survivor, and disability contingencies. Using administrative data from the Italian National Institute of Social Security (INPS), we derive the main components of the balance sheet by combining a standard actuarial approach for the estimation of accrued-to-date liabilities (ADL) with the framework proposed by Settergren and Mikula (2001, 2005) to compute the contribution asset (CA). Our results indicate an average annual solvency gap of approximately 12% over the period 2019–2021. This figure increases to 32% when transfers from general taxation are excluded from the system’s assets. Counterfactual exercises show that these estimates are sensitive to the choice of the discount rate and highlight the role of legacy defined benefit (DB) rules in shaping current liabilities. Extending the analysis through a long-term simulation of contributors and pensioners up to 2070, we find a gradual improvement in the system’s financial sustainability, driven by the increase in the effective retirement age and by the stronger alignment between liabilities and the contribution base associated with the progressive phase-in of the notional defined contribution (NDC) system.

ABSTRACT. This paper examines whether physician gender and age influence prescribing behaviour in primary care, with a particular focus on mental health treatments. Using a panel dataset of 7,431 general practitioner (GP) practices in England from 2017 to 2024, we analyse how variation in the gender composition of the GP workforce affects prescription patterns. Our empirical strategy exploits within-practice variation over time, controlling for patient demographics, workforce characteristics, and local socioeconomic conditions through fixed effects. We find that a higher share of female GPs is associated with a modest increase in mental health prescribing, but this effect is entirely driven by younger female physicians (under 40). In contrast, older female GPs exhibit no statistically significant association with prescribing behaviour. Importantly, these patterns are specific to mental health medications and do not extend to other drug categories such as antibiotics, statins, or analgesics. Further analysis shows that the effect is stronger in more deprived areas and in practices serving younger populations, where diagnostic discretion is likely to be greater. These findings highlight the role of physician cohort and professional socialisation in shaping clinical decision-making, suggesting that workforce composition can influence treatment patterns in domains characterised by uncertainty and subjective assessment

ABSTRACT. Retirement reduces personal income and may shift the balance of bargaining power between partners in elderly couples. Using data from the English Longitudinal Study of Ageing, I show that male partners are significantly less likely to be the household financial respondent once they retire, while no significant change for “who has the final say” over major financial decisions. Overall, results suggest a decline of men control in everyday financial matters, but no significant reallocation of bargaining power for formal financial choices.

ABSTRACT. This paper investigates how the physical environment of law enforcement facilities affects victims' willingness to report sexual violence. We evaluate a nationwide Italian intervention: the creation of "listening rooms" (stanze di ascolto)—dedicated, victim-centered spaces designed to provide privacy and psychological safety within police stations. Using a staggered Difference-in-Differences framework across 106 provinces (2016–2023), we find that the introduction of these rooms leads to a statistically significant increase in reported sexual crimes, with an average effect of 0.17 standard deviations. The impact is primarily driven by the initial introduction of the service, suggesting an extensive-margin mechanism where the availability of a protected environment facilitates the decision to report. Falsification tests on alternative crime categories and stable trends in Anti-Violence Center admissions confirm that the findings reflect the emergence of previously latent victimization rather than an increase in the underlying prevalence of violence. Our results demonstrate that specialized, victim-centered infrastructure can significantly bridge the reporting gap by fostering a secure and supportive environment for victims to come forward.

ABSTRACT. Administrative records of harm in closed institutions combine the true incidence of events with the probability that those events are observed and recorded. Separating the two is difficult because most interventions move both margins at once. I exploit the staggered timing of monitoring visits by Antigone, an Italian civil-society organization that observes and documents prison conditions but holds no managerial or enforcement authority. Recorded critical events rise sharply after the first visit, and the increase grows over time across staggered difference-in-differences estimators. The response is concentrated in reporting-sensitive outcomes such as self-harm and attempted suicides. Completed suicides and homicides-events that are difficult to conceal-do not move, and neither do staffing, overcrowding, or detainee composition. The pattern is what a reporting channel predicts and a deterioration channel does not: monitoring makes latent distress visible rather than worse. The findings show that the welfare content of administrative harm data depends on the form of oversight that produces it.

ABSTRACT. This paper investigates the determinants of legal gambling across Italian provinces (NUTS-3) over the period 2015–2023, with a particular focus on the interaction between legal and illegal gambling activity. Using a set of explanatory variables, including enforcement-related measures, illegal gambling, crime rates, income, education, unemployment, and other socioeconomic and demographic characteristics, we address two main research questions: (i) what are the main drivers of the legal gambling sector in Italy over the considered period, and (ii) whether illegal gambling acts as a substitute for legal gambling or instead reinforces legal gambling participation. While prior research has examined gambling behavior and market dynamics, evidence at the provincial level on the joint role of enforcement, local socioeconomic conditions, and illegal gambling in shaping legal gambling volume remains limited. The findings might offer policy-relevant insights for the design of gambling regulation and enforcement strategies.

ABSTRACT. Understanding the dynamics of relational networks is crucial in many social contexts, where interactions are often observed as count data. This paper proposes a dynamic latent factor model for count networks that captures both the formation of connections and the intensity of interactions over time. The framework separately models the probability of observing a link through a Bernoulli equation and the magnitude of positive interactions through a zero-truncated Poisson equation, allowing the model to account for the large proportion of zeros typically observed in sparse networks. The model incorporates time-varying sender and receiver effects to account for node-level heterogeneity, while latent dyadic effects capture unobserved pairwise relationships between nodes. Temporal dynamics are introduced through autoregressive processes AR(1) for both nodal and dyadic components, while the dyadic latent effects are allowed to be correlated across the two equations governing connection and intensity. To illustrate the model’s applicability, we analyse patient transfers within the Sicilian regional healthcare system. The empirical application reveals evolving interaction patterns and latent relational structures in hospital transfer networks, highlighting how the organization of the regional healthcare system has changed over time in response to unobserved factors that may reflect healthcare policy interventions.

ABSTRACT. This paper proposes a positional poverty gap measure of multidimensional poverty within the Alkire-Foster counting framework. The measure captures the depth of deprivations even when indicators are ordinal, unlike the standard poverty gap, which requires cardinal variables. The proposed method draws on the fuzzy set literature and introduces a distribution-based measure of deprivation depth using the empirical cumulative distribution of each indicator, with the most deprived group as the benchmark. For each deprived individual, the method assigns a score based on the individual's relative position in the distribution. Depth is thus expressed as a difference in distributional positions, motivating the label positional poverty gap. The paper demonstrates that this measure preserves the identification and aggregation structure of the counting approach and satisfies its axiomatic properties when the reference distribution remains fixed over time. The framework remains flexible because it accommodates different identification rules, deprivation cutoffs, and variable types. Overall, it offers a simple, meaningful, and theoretically grounded way to incorporate depth into multidimensional poverty measurement with ordinal data.

ABSTRACT. Binary indicators remain dominant in energy poverty (EP) measurement, yet their reliance on sharp thresholds obscures the continuous and heterogeneous nature of household vulnerability. This paper proposes a fuzzy logic–based framework that preserves comparability with official incidence metrics while explicitly modelling deprivation intensity. Building on the Faiella–Lavecchia (2015) expenditure rule, we reinterpret energy affordability as a continuous membership condition, where each household is assigned a degree of belonging to the EP set reflecting how strongly the affordability predicate is satisfied. The key methodological contribution is a calibration strategy in which the severity parameter of the fuzzy membership function is estimated so that the weighted mean membership reproduces the observed headcount ratio, ensuring empirical coherence between binary and fuzzy representations. The framework uses Verma-type membership functions, incorporates design weights, and is estimated on ISTAT Household Budget Survey microdata. Bootstrap-based variance estimation supports statistically valid regional and climatic comparisons. Results show high stability across alternative functional forms and near-perfect concordance in regional rankings, while revealing substantial heterogeneity in deprivation intensity that binary measures conceal. The proposed methodology bridges headcount indicators and multidimensional welfare analysis, offering a replicable template for national and comparative EP measurement.

ABSTRACT. We propose Bayesian networks as an innovative empirical approach to measure inequality of opportunity. We present this method as a viable alternative to the more advanced machine learning methods discussed in the literature, outlining its features and advantages. This methodology employs structural learning algorithms to generate an index of inequality of opportunity that is comparable with existing alternatives. In addition, it offers insights into the network of relationships between variables. We illustrate how this approach relates to established measurement methods using simulated data, and provide an application for five European countries.

ABSTRACT. This study examines how electoral competitiveness, voter heterogene- ity, and informational conditions shape politicians’ incentives to engage in corruption. Using a laboratory experiment in which participants act as in- cumbents allocating resources to bribe voters (targeting hesitant voters), we test predictions from a vote-maximisation framework enriched with behavioural considerations. Electoral competitiveness is varied through initial vote shares (35%, 40%, 45%), while voter heterogeneity is captured by differing proportions of voters who are highly versus weakly sensitive to corruption. The results reveal three key patterns. First, bribery ex- hibits a non-linear relationship with electoral competitiveness: investment in corruption is lowest under moderate competition (40%) and higher when incumbents are either far from or close to victory, consistent with reference-dependent risk behaviour predicted by prospect theory. Sec- ond, bribery increases when voters are more responsive to inducements and declines when moral or reputational costs dominate. Third, under uncertainty about voter types, participants behave as if adopting neu- tral priors, treating the electorate as evenly mixed. Overall, the findings provide micro-level behavioural evidence for strategic models of electoral corruption and help explain the persistence of vote buying in transitional democracies.

ABSTRACT. This paper studies repeated elections in which citizen-candidates differ along two dimensions: ideology and competence (valence). The model extends the dynamic citizen-candidate framework of Bernhardt et al. (2011) by allowing voters and incumbents to observe a costless public signal about an untried challenger's expected valence before the election. We find that in a symmetric threshold equilibrium, moderate incumbents choose policies that guarantee re-election, intermediate incumbents choose policies that win only against low-signal challengers, and sufficiently extreme incumbents implement their ideal policies and leave office. We characterize the continuation values generated by informative signals, derive the threshold equations that determine equilibrium policy compromise, and obtain comparative statics for signal informativeness, incumbent valence, office rents, patience, and exogenous turnover.

ABSTRACT. I study the relationship between technological progress and the extractive nature of political power. A citizen and a ruler interact over time. The citizen owns a stock of capital which -- in every period -- yields some output according to a time-varying technology. The ruler can choose how much output to expropriate for her private consumption, subject to the citizen's participation constraint. The speed of technological progress is endogenous to the citizen's choice of effort. The citizen optimally exerts more effort when the ruler reduces rent extraction. As technology improves, the ruler faces a dynamic trade-off. Higher rent extraction increases her payoff today, but discourages the citizen’s effort, thereby slowing the pace of technological progress and ultimately lowering her payoff tomorrow. I show that the equilibrium of the model features an endogenous transition from a regime with high rent extraction and slow technological progress to a regime with low rent extraction and fast technological progress.