Environmental, Social, and Governance (ESG) scores are quantitative assessments of companies’ commitment to sustainability that have become extremely popular tools in the financial industry. However, transparency in the ESG assessment process is still far from being achieved. In fact there is no full disclosure on how the ratings are computed. As a matter of fact, rating agencies determine ESG ratings (as a function of the E, S and G scores) through proprietary models which public knowledge is limited to what the data provider effectively chooses to disclose, that, in many cases, is restricted only to the main ideas and essential principles of the procedure. The goal of this work is to exploit machine learning techniques to shed light on the ESG ratings issuance process. In particular, we focus on the Refinitiv data provider, widely used both from practitioners and from academics, and we consider white-box and black-box mathematical models to reconstruct the E, S, and G ratings’ assessment model. The results show that it is possible to replicate the underlying assessment process with a satisfying level of accuracy, shedding light on the proprietary models employed by the data provider. However, there is evidence of persisting unlearnable noise that even more complex models cannot eliminate. Finally, we consider some interpretability instruments to identify the most important factors explaining the ESG ratings.

Environmental, Social, and Governance (ESG) scores are crucial for evaluating corporate sustainability. However, the undisclosed and complex methodologies used by rating agencies have raised concerns about their reliability and consistency. This study replicates LSEG's ESG scoring methodology using machine learning to shed light on the key drivers behind ESG ratings, with a focus on the balance between forward‐looking promises (aspirational) and past achievements (performance). Our analysis finds that approximately 60% of ESG scores are based on aspirational promises, while only approximately 40% reflect actual performance. This imbalance suggests a potential over‐reliance on future commitments, which could inflate ESG scores and mislead investors about a company's true sustainability efforts, even accounting for LSEG's transparency stimulation mechanism, where non‐disclosure of material data is penalized. The findings emphasize the need for greater transparency and a clearer distinction between aspirational and performance metrics to ensure credible ESG assessments for informed investment decisions.

(1) Background: COVID-19 vaccination campaigns offer the best hope of controlling the pandemic. However, the fast production of COVID-19 vaccines has caused concern among the general public regarding their safety and efficacy. In particular, patients with chronic illnesses, such as celiac disease (CD), may be more fearful. Information on vaccine hesitancy plays a pivotal role in the development of an efficient vaccination campaign. In our study, we aimed to evaluate COVID-19 vaccine hesitancy among Italian CD patients. (2) Methods: an anonymous questionnaire was sent to CD patients followed at our tertiary referral center for CD in Milan, Italy. Patients were defined as willing, hesitant and refusing. We evaluated the reasons for hesitancy/refusal and the possible determinants, calculating crude and adjusted odds ratios [AdjORs] with 95% confidence intervals [CIs]. (3) Results: the questionnaire was sent to 346 patients with a response rate of 29.8%. Twenty-six (25.2%) of the 103 respondents were hesitant, with a total refusal rate of 4.8%. The main reason was fear of adverse events related to vaccination (68.2%). Among hesitant patients, 23% declared that their opinion was influenced by their CD. The determinants positively influencing willingness to be vaccinated against COVID-19 were adherence to a GFD, perception of good knowledge about COVID-19 and its vaccines, and a positive attitude to previous vaccines (AdjOR 12.71, 95% CI 1.82–88.58, AdjOR 6.50, 95% CI 1.44–29.22, AdjOR 0.70, 95% CI 0.11–4.34, respectively). (4) Conclusions: CD patients should be vaccinated against COVID-19 and a specific campaign to address the determinants of hesitancy should be developed.

As climate change accelerates, the frequency and severity of extreme weather events, such as wildfires, are increasing, with profound impacts on human health. While much attention has been paid to the direct health consequences of these events, such as chronic diseases from poor air quality, less is known about how behavioral shifts induced by such events can influence the transmission of infectious diseases. This study investigates how wildfire-induced changes in human behavior during the U.S. West Coast wildfires of 2020 may affect the spread of airborne diseases. Using a mobility data-driven indoor activity index, we find that the wildfire-induced deterioration of air quality led to a substantial increase in indoor activities, fostering conditions conducive to airborne disease transmission. Specifically, counties in Oregon and Washington experienced an average 10.8% and 14.3% increase in indoor activity, respectively, during the wildfire events, with major cities like Portland and Seattle experiencing increases of 11% and 16%, respectively. We quantify these behavioral changes and integrate them into an SIR epidemic model to characterize the increased indoor activity and disease dynamics. The model predicts the greatest impact on diseases with shorter generation times, such as RSV and influenza. Our results show that even a modest increase in indoor mask-wearing (as low as 10%) could significantly reduce the risk of disease spread in these settings, with higher compliance needed for more substantial reductions. As wildfires and other climate-related events become more frequent, integrating behavioral responses into public health policies will be crucial to mitigate the compounded risks of climate change and its secondary health impacts.