• Other research productsclose
• 2013-2022close
• Other ORP type close
• French National Research Agency (ANR) close
• FR close
• English close
• Hal-Diderot close
• HAL-Inserm close

This is a position paper presented at CHI2021 Workshop called "Towards Explainable and Trustworthy Autonomous Physical Systems".; The spread of AI-embedded systems involved in human decision making makes it critical to build these systems according to trustworthiness standards. To understand whether this goal was achieved, users' trust in these systems must be studied. However, empirically investigating trust is challenging. One reason is the lack of standard protocols to design trust experiments. To get an overview of the current practices in the experimental protocols for studying trust in the context of AI-assisted decision making, we conducted a systematic review of such papers. We annotated, categorized, and analyzed them along the constitutive elements of an experimental protocol (i.e., participants, task). Drawing from empirical practices in social and cognitive studies on human-human trust, we provide practical guidelines and research opportunities to improve the methodology of studying Human-AI trust in decision-making contexts. In this workshop, we would like to start the discussion about how these guidelines and research questions can be used in the laboratory and in the wild.

URL des Documents de travail : https://centredeconomiesorbonne.cnrs.fr/publications/; Documents de travail du Centre d'Economie de la Sorbonne 2021.04 - ISSN : 1955-611X; Immigration and crime are two first-order issues that are often considered jointly in people's minds. This paper analyzes how media reporting policies on crime impact natives' attitudes towards immigration. We depart from most studies by investigating the content of crime-related articles instead of their coverage. Specifically, we use a radical change in local media reporting on crime in Germany as a natural experiment. This unique framework allows us to estimate whether systematically disclosing the places of origin of criminals affects natives' attitudes towards immigration. We combine individual survey data collected between January 2014 and December 2018 from the German SocioEconomic Panel with data from more than 545,000 crime-related articles in German newspapers and data on their diffusion across the country. Our results indicate that systematically mentioning the origins of criminals, especially when offenders are natives, significantly reduces natives' concerns about immigration.

URL des Documents de travail : https://centredeconomiesorbonne.cnrs.fr/publications/; Documents de travail du Centre d'Economie de laSorbonne 2021.02 - ISSN : 1955-611X; Judgment-aggregation theory has always focused on the attainment of rational collective judgments. But so far, rationality has been understood in static terms: as "coherence" of judgments at a given time, understood as consistency, completeness, and/or deductive closure. By contrast, this paper discusses whether collective judgments can be dynamically rational, so that they change rationally in response to new information. Formally, a judgment aggregation rule is dynamically rational with respect to a given revision operator if, whenever all individuals revise their judgments in light of some information (a learnt proposition), then the new aggregate judgments are the old ones revised in light of this information, i.e., aggregation and revision commute. We prove a general impossibility theorem: if the propositions on the agenda are sufficiently interconnected, no judgment aggregation rule with standard properties is dynamically rational with respect to any revision operator satisfying some mild conditions (familiar from belief revision theory). Our theorem is the dynamic-rationality analogue of some well-known impossibility theorems for static rationality. We also explore how dynamic rationality might be achieved by relaxing some of the conditions on the aggregation rule and/or the revision operator.

URL des Documents de travail : https://centredeconomiesorbonne.cnrs.fr/publications/ Voir aussi l'article basé sur ce document de travail paru dans "Journal of Economic Theory", 194, 2021, 105255; Documents de travail du Centre d'Economie de la Sorbonne 2020.14R - ISSN : 1955-611XVersion originale, tittre "The Rational Group" Juin 2020, révisée en Janvier 2021, titre "Fully Bayesian Aggregation"; Can a group be an orthodox rational agent? This requires the group's aggregate preferences to follow expected utility (static rationality) and to evolve by Bayesian updating (dynamic rationality). Group rationality is possible, but the only prefefence aggregation rules which achieve it (and are minimally Paretian and continuous) are the linear-geometric rules, which combine individual values linearly and individual beliefs geometrically. Linear-geometric preference aggregation contrasts with classic linear-linear preference aggregation, which combines both values and beliefs linearly, and achieves only static rationality. Our characterisation of linear-geometric preference aggregation implies as corollaries a characterisation of linear value aggregation (Harsanyi's Theorem) and a characterisation of geometric belief aggregation.

URL des Documents de travail : https://centredeconomiesorbonne.cnrs.fr; Documents de travail du Centre d'Economie de la Sorbonne 2020.27 - ISSN : 1955-611X; It has been claimed that climate policies can be evaluated by the Pareto principle. However, climate policies lead to different identities and different numbers of future people. Even if one assumes that the number of future people is countably infinite independently of policy choice, the problem is that there exists no natural one-to-one correspondence between the components of the compared alternatives. This non-existence means that the components of streams are indexed by natural numbers that do not correspond to particular people, making a case for impartiakity in the sense of Strong anonymity. Strong anonymity is incompatible with Strong Pareto. The paper re-examines this incompatibility and investigates how far sensitivity for the well-being at any one component can be extended without contradicting Strong anonymity. We show that Strong anonymity combined with four rather innocent axioms has two consequences: (i) There can be sensitivity for the well-being at a particular component of the stream if and only if a finite set of people have higher well-beings, and (ii) adding people to the population cannot have positive social value.

URL des Documents de travail : https://centredeconomiesorbonne.cnrs.fr/publications/; Documents de travail du Centre d'Economie de la Sorbonne 2020.15R - ISSN : 1955-611XVersion originale Septembre 2019, révisée en Mars 2021; Maximising expected value is the classic doctrine in choice theory under empirical uncertainty, and a prominent proposal in the emerging philosophical literature on normative uncertainty, i.e., uncertainty about values. But how should Expectationalism be stated in general, when we can face both uncertainties simultaneously, as is common in life? Surprisingly, different possibilities arise, ranging from Ex-Ante to Ex-Post Expectationalism, with several hybrid versions. Expectationalism thus faces the classic dilemma between ex-ante and ex-post approaches. Different expectational theories reach diverging evaluations, use different modes of reasoning, and take different attitudes to risk; but they converge undre an interesting (necessary and sufficient) condition. We relate choice under normative uncertainty to choice by Harsany's impartial observer' and Rawls's person behind the 'veil of ignorance', who are uncertain about their own identity and values; and we relate Expectationalism under normative uncertainty to Harsanyi's linear approach to aggregating values of different individuals or identities.

Summary:The European economic union is incomplete, which makes it vulnerable to macroeconomic shocks. The opportunity to move forward in the integration process was highly debated even before the Covid-19 crisis.Yet the diverging views among countries and political groups are often considered as an obstacle on the path to required agreements for completing the Economic and Monetary Union (EMU). We present the results of a survey conducted in 2018 among members of national parliaments (MPs) in France, Germany and Italy on European integration in policy fields related to risk-sharing and budgetary institutions, asking for their opinion on proposals such as the creation of a European Unemployment Insurance (EUI), Eurobonds, or an EU tax. We find that nationality and political groups are key determinants of support for such proposals, the latter being the strongest. We describe how opinions are divided and try to identify policy proposals which could gather enough political support. The agreement reached on July 21st, 2020 at the last European summit includes financial transfers between States and the creation of Eurobonds, thus representing an important institutional move and an application of some of the reforms suggested by our survey. Yet what has been decided upon is only temporary and leaves open the question of the future of European integration.Key points: At first glance, the answers show diverging opinions on most questions between countries with Italy supporting more integration, and Germany opposing it for most proposals. France has an intermediate position, leaning towards Italy. A breakdown of the results by party affiliation shows a more nuanced picture. For cross-country comparisons, we build a party indicator using the affiliation of national parties to European political groups. National MPs associated with the group of Socialists and Democrats (S&D) at the European level show strong support for the creation of new fiscal institutions and a new EU tax, and for risk sharing institutions (European Unemployment Insurance, Eurobonds). On the contrary, MPs associated with the European People’s Party (EPP) are mildly positive or against risk-sharing and fiscal institutions. National MPs affiliated to Renew Europe hold similar views to S&D MPs, but are less supportive of risk-sharing mechanisms. There is a substantial diversity of positions between the German AfD, the Italian Lega and the 5-star movement: the three parties have diverging views on the future of integration. Our econometric analysis shows that party affiliations have more explanatory power than nationality for all questions. This clearly shows that outcomes of national parliamentary elections could change the overall support for any issue.; Summary:The European economic union is incomplete, which makes it vulnerable to macroeconomic shocks. The opportunity to move forward in the integration process was highly debated even before the Covid-19 crisis.Yet the diverging views among countries and political groups are often considered as an obstacle on the path to required agreements for completing the Economic and Monetary Union (EMU). We present the results of a survey conducted in 2018 among members of national parliaments (MPs) in France, Germany and Italy on European integration in policy fields related to risk-sharing and budgetary institutions, asking for their opinion on proposals such as the creation of a European Unemployment Insurance (EUI), Eurobonds, or an EU tax. We find that nationality and political groups are key determinants of support for such proposals, the latter being the strongest. We describe how opinions are divided and try to identify policy proposals which could gather enough political support. The agreement reached on July 21st, 2020 at the last European summit includes financial transfers between States and the creation of Eurobonds, thus representing an important institutional move and an application of some of the reforms suggested by our survey. Yet what has been decided upon is only temporary and leaves open the question of the future of European integration.Key points: At first glance, the answers show diverging opinions on most questions between countries with Italy supporting more integration, and Germany opposing it for most proposals. France has an intermediate position, leaning towards Italy. A breakdown of the results by party affiliation shows a more nuanced picture. For cross-country comparisons, we build a party indicator using the affiliation of national parties to European political groups. National MPs associated with the group of Socialists and Democrats (S&D) at the European level show strong support for the creation of new fiscal institutions and a new EU tax, and for risk sharing institutions (European Unemployment Insurance, Eurobonds). On the contrary, MPs associated with the European People’s Party (EPP) are mildly positive or against risk-sharing and fiscal institutions. National MPs affiliated to Renew Europe hold similar views to S&D MPs, but are less supportive of risk-sharing mechanisms. There is a substantial diversity of positions between the German AfD, the Italian Lega and the 5-star movement: the three parties have diverging views on the future of integration. Our econometric analysis shows that party affiliations have more explanatory power than nationality for all questions. This clearly shows that outcomes of national parliamentary elections could change the overall support for any issue.

URL des Documents de travail : https://centredeconomiesorbonne.univ-paris1.fr/documents-de-travail-du-ces/; Documents de travail du Centre d'Economie de la Sorbonne 2020.14 - ISSN : 1955-611X; Can a group be a standard rational agent? This would require the group to hold aggregate preferences which maximise expected utility and change only by Bayesian updating. Group rationality is possible, but the only preference aggregation rules which support it (and are minimally Paretian and continuous) are the linear-geometric rules, which combine individual tastes linearly and individual beliefs geometrically.

URL des Documents de travail : https://centredeconomiesorbonne.cnrs.fr/publications/ Voir aussi l'article basé sur ce document de travail paru dans "Journal of Economic Methodology, Taylor & Francis (Routledge), 2021, 28 (2), pp.143-164. Documents de travail du Centre d'Economie de la Sorbonne 2020.16 - ISSN : 1955-611X International audience Leonard Savage famously contravened his own theory when first confronting the Allais Paradox, but then convinced himself that the had made an error. We examine the formal structure of Savage's ‘error-correcting’ reasoning in the light of (i) behavioural economists' claims to identify the latent preferences of individuals who violate conventional rationality requirements and (ii) John Broome's critique of arguments which presuppose that rationality requirements can be achieved through reasoning. We argue that Savage's reasoning is not vulnerable to Broome's critique, but does not provide support for the view that behavioural scientists can identify and counteract errors in people's choices.

URL des Documents de travail : https://centredeconomiesorbonne.cnrs.fr; Document de travail du Centre d'Economie de la Sorbonne 2020.26 - ISSN : 1955-611X; Climate change-related mortality may strongly affect human well-being. By reducing life expectancy, it reduces the well-being of some infividuals. This may exacerbate existing inequalities: ex-ante inequality among people in different groups or regions of the world; ex-post inequality in experienced well-being by people in the same generation. But mortality may also reduce total population size by preventing some individuals from having children. This raises the population-ethical problem of how total population size should be valued. This paper proposes a methodology to measure te welfare effects of climate change through population and inequality change. We illustrate the methodology using a climate-economy integrated assessment model involving endogenous population change due to climate change-related mortality.