Filippo Pavesi

We investigate how generative AI shapes creative performance and human-AI interaction in an open-ended writing task that employs a laboratory experiment in which participants are randomly assigned to either receive access to a large language model (ChatGPT-4.2) or not. Creative performance is measured by the average score assigned by independent evaluators recruited through the Prolific platform, and detailed logs of human-AI interaction are analyzed to measure AI use, prompting intensity, ideation requests, and the textual overlap between AI outputs and participants' final writings. Three main results emerge. First, AI access increases performance, but the gain is entirely driven by active use: participants with access who do not submit queries perform no better than those without AI. Second, the relationship between interaction intensity and performance is concave, peaking at roughly eight queries, consistent with iterative exploration rather than mechanical copying. Third, structural mediation analyses show that ideation requests affect performance primarily indirectly, by increasing downstream incorporation of AI-generated language; the direct effect of requesting an idea from the AI is negligible once execution-stage reliance is accounted for. We further document heterogeneity in AI reliance: cultural capital (proxied by books owned) predicts lower AI use, while prior AI exposure predicts higher use. By contrast, incentive schemes have limited effects on both outcomes and AI-related behaviors.

In this study, we use an administrative dataset of 406 students from a compulsory economics course for students enrolled in a MSc in Business Administration at a private Italian university, to examine how allowing students the option of attending either in person or online affects both their choice of attendance mode and academic performance. Our analysis exploits the rules implemented in Italy during the 2020-21 fall semester, when students were initially allowed to choose between in-person and online attendance, but then forced to attend online due to the pandemic in the second half of the semester. A key feature of the study is that the two attendance modes differ merely in terms of whether the student is physically present in the classroom or connected via video conferencing, whereas other learning features, such as the possibilities of interacting with professors, are identical. According to our results, female students are more likely to attend online than male students. Despite this, the online technology under evaluation has a neutral effect on students' academic performance. These findings suggest that an online teaching technology that allows students to interact with the class as if they were in the classroom can be welfare improving, because it allows for greater flexibility in allocating time beyond the classroom.

In this paper, we test the hypothesis that participation in cultural associations improves social capital, and more specifically generalized trust, which in turn fosters environmental concern. Using a dataset combining two large cross-national socioeconomic surveys and an instrumental variable mediation approach, we find that a measure of generalized trust fully mediates the relationship between people’s participation in cultural organizations and their environmental concerns. This result suggests that governments can increase citizens' environmental concern, a prerequisite for pro-environmental behaviors, by encouraging their participation in cultural associations.

Decision-making crucially relies on information. There are many cases in which those in charge of making decisions face a relative scarcity of relevant information with respect to other agents who instead possess it (i.e., experts). With information being available but asymmetrically distributed, a problem arises about how effectively it can be transmitted and used. In particular, when an expert is known (or simply believed) to face conflicts of interest and the decision maker cannot use contracts to align incentives, a problem of credibility arises with respect to the information conveyed by the expert. This chapter offers a selected review of the theoretical literature on strategic information transmission which studies the issue of how much information can be credibly conveyed by an informed but biased agent in the absence of explicit incentives. Given the vast size of this literature, the contribution of this review is to identify four broad institutional aspects that are deemed to be key: (i) the extent to which delegation can be superior to communication; (ii) the role that experts’ competition plays in helping the decision maker elicit information; (iii) the effectiveness of reputation as an implicit incentive that can align incentives; (iv) the role played by networks. We outline the role of these key institutional aspects in influencing information revelation with the scope of providing a platform for the design of expert markets and institutions.

Il cibo è una necessità assoluta per il sostentamento fisico della vita umana, ma è anche un’espressione culturale irrinunciabile per ciascuno individuo. Il testo si addentra in un campo difficoltoso e complesso quello del cibo e della cucina, che richiede una spiccata sensibilità, permeata di radicati e irrinunciabili principi etici. Per farlo si avvale, oltreché della competenza degli autori, anche delle esperienze e tecniche professionali maturate in anni di appassionato lavoro e approfondite ricerche sul campo di numerosi esperti e studiosi. L’opera prende le mosse da alcune notazioni sulla "Filosofia del cibo", argomento già affrontato dagli Autori nel testo «Metafisica, Filosofia e Scienza del Cibo» pubblicato nel 2022 dalla nostra casa editrice, grazie al contributo di Grazia Bottiglieri. Vengono quindi forniti alcuni pochi storiografici sul cibo, in modo da renderli immediati ed agevoli per richiamarci alla mente l’articolata problematica del cibo in tavola, senza debordare in ipotesi, magari molto suggestive, ma prive di quel filo logico che rende l’esposizione di questo delicato argomento solida e certa, quale si è mantenuta da oltre 3.500 anni. La prima parte del testo è dedicata alle diete alimentari e le loro caratterizzazioni, introducendo alcune considerazioni propedeutiche e articolate sulla specie umana, spingendosi poi sulla suddivisione e classificazione delle scienze alimentari, sulle relazioni tra l’organismo umano e il suo ambiente e sul carattere delle popolazioni che vi risiedono. L’introduzione affronta poi la relazione fra uomo e cibo, inteso come questione tecnica, mediatica e morale, senza tralasciare osservazioni sul binomio cibo – medicina. Segue un articolato capitolo dedicato all’arte cibaria, che parte dalla visione del cibo come medicina naturale per passare ad analizzarne gli aspetti culturali ed artistici e persino spirituali. Si analizza poi la catena alimentare umana nella società contemporanea e si definiscono gli elementi nutrizionali, per passare all’’evoluzione della cucina nel tempo, dalle tecniche fermentative ai metodi e sistemi di cottura dei cibi, fino a considerare le diete contemporanee, anche in relazione al cancro. La seconda parte si sofferma sull’universalizzazione della cucina e lo specialissimo "caso italiano", percorrendone i tratti salienti – compresi valori simbolici e valori mitici – , fornendo precisazioni di carattere generale e tecnologico sugli alimenti maggiormente diffusi nel nostro paese e alcuni cenni storici sul pane italiano, passando poi a pasta, riso, carne e pesce. Chiude il testo una postfazione che propone qualche consiglio su come condurre una "vita cibaria igienica" per mantenerci sani il più a lungo possibile.

We explore the impact of narratives on beliefs and policy opinions through a survey experiment that exposes US subjects to two media-based explanations of the causes of COVID-19. The Lab Narrative ascribes the pandemic to human error and scientific misconduct in a Chinese lab, and the Nature Narrative describes the natural causes of the virus. First, we find that both narratives influence individual beliefs about COVID-19 origins. More precisely, individual beliefs tend to be swayed in the direction of the version of the facts to which one is more exposed generating a potential source of polarization by exposure. Second, only the Nature Narrative unidirectionally affects policy opinions by increasing people’s preferences toward climate protection and trust in science, therefore representing a channel for one-sided polarization by exposure. Finally, we also explore the existence of heterogeneous effects of our narratives, finding that the Lab Narrative leads to opinion polarization between Republican- and Democratic-leaning states on climate change and foreign trade. This indicates the existence of an additional channel that can lead policy opinions to diverge, which we denote polarization by social context.

By promoting economic growth, human capital may contribute to the rise in CO2 emissions, but it may also stimulate emission-reducing technologies. Starting from a Green Solow model augmented with human capital, we show that the former effect dominates the latter when human capital is below a critical value, while the opposite is true when human capital becomes sufficiently high. We also find that this result may delay the observability of an EKC and that human capital is more important than savings and depreciation rates in predicting CO2 growth. This evidence has relevant policy implications regarding which factors should be considered to mitigate carbon emissions.

We investigate how the opportunity cost of delegation influences the impact of communication on trust, trustworthiness and communication style. To do this, we adopt a modified trust game in which the trustee may send a message to the trustor, before the latter chooses whether to delegate decision rights to the former. Our analysis shows that, only when the opportunity cost of delegation is high does communication have a positive effect on the beliefs (first and second order) relative to the amount the trustor will receive if he chooses to delegate, as well on the trustee's beliefs on the probability of delegation occurring. This implies that the incentives to use communication to increase trust are greater when the opportunity cost of delegation is higher, even if it does not affect the decision to delegate and is associated with deception. Regarding communication style, our analysis shows that non-precise promises are used more frequently when the opportunity cost of delegation is low, and the trustor has less to lose from delegating. We also document an “illusion effect” whereby trustees erroneously expect non-precise promises to have an impact on the beliefs of trustors.

We present evidence that international trade may exacerbate the initial unequal distribution of hydric resources. This result is driven by the fact that countries exporting agricultural goods are relatively abundant (with respect to capital) in the combined availability of water and arable land but, in absolute terms, scarce in capital and not richer in water in comparison to more developed ones. Due to both the scarcity of capital and the lower relative price of natural resources with respect to capital, the total value of production in these developing countries is modest, implying that international trade can lead to a less even distribution of the water content of consumption. Policies sustaining water prices and, more generally, those of natural resources (or lower capital costs) may contribute to offsetting this effect and allow for trade to play a positive role in reducing the uneven distribution of water endowments.

Decision makers often face uncertainty about the ability and the integrity of their advisors. If an expert is sufficiently concerned about establishing a reputation for being skilled and unbiased, she may truthfully report her private information about the decision-relevant state. However, while in a truthtelling equilibrium the decision maker learns only about the ability of the expert, in an equilibrium with some misreporting she also learns about the expert's bias. Although truthtelling allows for better current decisions, it may lead to worse sorting outcomes. This occurs if misreporting is principally caused by biased experts driven by their conflict of interest rather than by unbiased experts attempting to signal their type. Whenever lying has these features, it increases the decision maker's expected utility with respect to truthtelling if she is sufficiently concerned about future choices. In these cases, it is optimal to adopt policies aimed at reducing advisors' career concerns and we suggest real world examples in which these could be implemented.