Research

Using a unique dataset of administrative data from municipalities in the Veneto region of Italy for 2010–2019, we develop a spatial econometric model to study the effects of two waste management policies: Door-to-Door collection and Pay-As-You-Throw tariff. We focus on the impact of these policies on waste sorting and accumulation, with particular attention to spatial spillovers. Both policies show similar effects on the outcome variables, leading to an increase in waste sorting and a reduction in waste accumulation; interestingly, we also find evidence of spatial spillovers. However, we identify unintended negative spillovers, where waste is diverted to neighboring municipalities with less stringent regimes, potentially undermining policy effectiveness. This study is the first to use a spatial econometric model to investigate how the adoption of a policy in a municipality affects waste production in surrounding areas, highlighting the need for coordinated decisions in the implementation of waste management policies.

Using real platform data, this paper assesses donors’ reactions to a rare event; specifically, we demonstrate how donors were able to compensate for (or stabilize) healthcare deficits and policy restrictions used to mitigate the spread of COVID-19. Based on a sample of 20,784 COVID-related donation campaigns promoted on the GoFundMe platform from January 30, 2020 to November 23, 2021, the paper shows that donors decided to donate based on campaign purpose and reliability: specifically, campaigns with the purpose to support hospitals or to buy healthcare materials in countries more affected by the pandemic received more funds. Moreover, the paper provides evidence of the impact of different government policy restrictions on campaign outcome in terms of the amount raised and the number of campaigns promoted in six European countries. As donations during COVID-19 were valuable (in particular, for public institutions) in terms of amount and timeliness, our findings offer valuable insights into how governments encouraged donations via fiscal incentives and policy restrictions.

Adaptation to climate-related hydrological risks—such as floods and landslides—is increasingly central to disaster management and resilience policy. Yet measuring adaptation efforts at scale remains challenging. This paper introduces HYDROADAPT, a novel dataset of adaptation policies related to hydrological risk, constructed using a custom classification pipeline powered by large language models (LLMs). Drawing from the Italian registry of public investments (OpenCUP), I focus on the Emilia-Romagna region and identify over 24,000 projects semantically linked to hydrological instability. I classify these projects as either Ex-Ante (preparedness) or Ex-Post (remedial) interventions based on their textual descriptions. Each policy is geolocated, time-stamped, and enriched with metadata on funding volume, policy instrument, and implementing body. I present descriptive patterns in adaptation activity across time, space, and intervention types. To validate the classification, I use the timing of the May 2023 floods that hit the region. HYDROADAPT provides a scalable, transparent, and replicable framework for measuring climate adaptation policy — and lays the foundation for future empirical evaluation of its effectiveness.