Abstracts (alphabetical order by speaker's family name)

Speaker: Nicola Campigotto

Title: Inequality Aversion in AI-Assisted Routine and Creative Work

Abstract: This project examines whether public acceptance of AI-assisted work differs between creative and routine tasks and whether these differences shape fairness judgments about performance-based pay. The study uses an incentivized online experiment with three phases. In the first phase, workers complete either a creative task (writing a short story for children) or a routine real-effort task. Some workers have access to a GPT chatbot while others do not. In the second phase, workers’ performance is evaluated either by objective output in the routine task or by ratings of stories in the creative task. Workers are then paired within treatment conditions and receive unequal earnings based on performance. In the final phase, spectators observe a worker pair, learn how performance was assessed, who performed better, the initial earnings allocation and, when relevant, which worker relied more on AI. They then decide whether to redistribute earnings. The experiment varies three dimensions: the nature of the task, whether AI assistance is available, and whether creative work is evaluated by humans or AI. The main outcome is the Gini coefficient implemented by spectators, used as a measure of inequality aversion. The central hypothesis is that AI assistance affects fairness perceptions differently across tasks. In routine work, AI may be seen as a legitimate productivity-enhancing tool, so workers who use it more effectively may be viewed as deserving higher earnings. In creative work, however, AI may be perceived as undermining originality or as a form of cheating, leading spectators to be less accepting of income differences when performance depends on AI use.

Speaker: Roberto Caputo

Title: Anticipated Discrimination, Gender Stereotypes, and Competitive Choices in Job Applications: A Large-Scale Online Experiment

Abstract: We study to what extent including evaluator subjectivity in tournament decisions affects the documented gender gap in competition. We conduct a large-scale pre-registered online experiment (n = 7,310) in which workers choose between a non-competitive piece rate and a competitive payment scheme whose payout depends on being selected by an incentivised human evaluator. By randomly varying whether the evaluator can observe the worker’s gender across a male-stereotyped and a non-stereotyped task, we exogenously induce the scope for anticipated discrimination. In the male-stereotyped task, gender revelation breaks the link between women’s objective performance beliefs and their beliefs about being selected conditional on the subjective probability of having the best signal: moderately confident women report significantly lower selection expectations when their gender is known. This distortion reduces female entry rates, with the strongest effects among non-risk-averse participants, those who made up the majority of the competitor pool when their gender was concealed. The revelation of gender in the male-stereotyped domain particularly affects low-ability women, destabilising the negative association between ability and competition among females. We validate that this effect is driven by the anticipation of gender discrimination and not by a preference for opting out of settings in which gender is made salient: in a non-stereotyped task, we find no change in beliefs nor in competition behaviour when their gender is revealed to the evaluator. Our findings suggest that the extent to which the gender competition gap distorts female labour supply may be underestimated in the face of the potential for evaluator bias, even before it is enacted.

Speaker: Isil Celik

Title: Do social norms of a country impact personal tax morale in the context of tax compliance?

Abstract: Method: To understand whether social norms impact tax morale, we employ data from the World Value Survey (WVS) 7th Wave, which was collected among 66 countries; participants were surveyed once per wave between 2017 and 2022. WVS 7th Wave provides information on financial satisfaction and confidence in government institutions, perceived corruption, perceived corruption of ordinary people, perceived risk of corruption, tax morale, perceived democracy and political satisfaction, nationality pride and feeling close to different social groups in a country, sex, age, level of education, and employment status. We will employ a linear probability regression in which tax morale is a binary dependent variable: if participants believe tax evasion is never justifiable = 1, 0 otherwise; perceived corruption and feeling close to the country as main variables of interest, and remaining variables as controls. We will also check the interaction between perceived corruption and feeling close to the country. Expected results: We expect perceived corruption hinders tax morale, whereas feeling close to the country would positively impact tax morale. Regarding the interaction term, feeling close to a country will empower the negative impact of perceived corruption, as the closer you feel to the country and the more corrupt you see your country will overall hinder your own tax morale. Motivation: Developed and developing countries built Behavioural Insights Units to empower the impact of public policies with behavioural interventions. They run certain randomized controlled trials, convert their findings into behavioural public policies. These behavioural interventions include improving tax morale and compliance, since even a 1 percent significant result may result in millions of pounds of collected tax. If social norms significantly impact personal tax morale, a country could save a million pounds in a budget year.

Speaker: Linda Dezsö

Title: Partial spectators: Identity-biased third-party allocations

Abstract: Third parties are assumed to allocate impartially because they have no monetary stake in the outcome. We ask whether sharing a political identity with an affected party compromises this impartiality. Using US partisan identities, we test whether redistribution favors co-partisans or disadvantages opposing partisans when inequality is transparently due to luck. We conducted an incentivized online experiment with US participants recruited through Prolific. Workers completed the same fixed-effort task and were randomly assigned 1 or 3, creating luck-based inequality within pairs whose 4€ pool spectators divided. The final spectator sample comprised 2,616 self-identified Democrats and Republicans. Across eight treatments, we varied whether the low earner was a co-partisan, the high earner an opposing partisan, or either worker apolitical. One payoff-relevant allocation per spectator identified identity bias, ingroup love, and outgroup hate relative to an apolitical baseline. Critically, unlike in previous experiments, spectators did not know that their identities were observed when they made their choices, so that identity cues were implicit and grouping was not mentioned. Partisan identity shifted redistribution toward the low earner among both Republican and Democratic spectators. Among Republicans, a co-partisan low earner increased redistribution by 2.14 percentage points, an opposing-partisan high earner by 2.16, and both cues by 3.19. Among Democrats, the corresponding effects were 3.34, 1.70, and 2.95 percentage points. All six contrasts were significant in the primary specification and substantively unchanged with demographic and heterogeneity controls. Overall, identity bias did not differ significantly between Republican and Democratic spectators. Thus, removing monetary self-interest does not guarantee impartiality: politically irrelevant identity cues can alter third-party fairness judgments. We discuss policy implications of these results.

Speaker: Alexandra Geis

Title: Menstrual Norms: Experimental Evidence from Nepal

Abstract: Cultural taboos and stigma severely limit menstrual health management in many low- and middle-income countries. Policies often target visible consequences but fail to address underlying cultural stigma and social norms perpetuating these harmful practices. We study underlying harmful norms that nurture cultural stigma and shed light on how these norms are perpetuated using a lab-in-the-field experiment with over 1,800 school children in Nepal. We elicit social norms and measure support for abolishing menstrual stigma using a Donation Game where participants can show support for abolishing menstrual stigma by donating to a menstrual charity fighting menstrual restrictions. We vary the visibility of the decision among peers and the availability of third-party punishment. Sixty percent of the participants fully support ending menstrual restrictions. Social image concerns among peers and peer punishment crowd in support for removing menstrual restrictions, while social sanctions by the elderly, particularly elderly women, reduce this support. We conclude that the social expectations of elder generations maintain harmful practices and that any efforts to demolish menstrual restrictions and stigma needs to include multiple generations, in particular the elderly.

Speaker: Rui Guan

Title: Fair Cooperation

Abstract: How fair is inequality, and when does it help or hinder cooperation? Prior work shows that fairness perceptions depend on the source of inequality, while cooperativeness differs across societies. We build on these insights with an online experiment that varies redistributive regimes and ability requirements to study how inequality and fairness perceptions shape cooperative behaviour. We show that perceived fairness varies systematically with taxation levels and is positively associated with cooperation. Strikingly, regimes with stronger ability requirements and lower redistribution foster higher cooperation. This pattern is broadly consistent with inequality aversion giving way to reciprocity in regimes with weaker ability requirements or with higher redistribution, alongside systematic heterogeneity across beneficiary status.

Speaker: Pietro Guarnieri

Title: Unequal Endowments in a Multilevel Public Goods Game

Abstract: This study examines the effect of competition and horizontal inequality on contribution decisions in a repeated multilevel public goods game (MLPGG). On the one hand, competition is manipulated by ranking subjects’ performances in a real effort task and attributing a high endowment to the high-performers (the first half of the ranking) and a low endowment to the low-performers (the second half of the ranking). On the other hand, horizontal inequality is manipulated by grouping in the same local group of the MLPGG only subjects with the same level of endowment, i.e. only subjects with high endowment (H) in one local group and only subjects with low endowment (L) in the other local group. Based on these manipulations, we set up a 2X2 between-treatments design, giving rise to four conditions: C1 = no competition and no horizontal inequality; C2 = competition and no horizontal inequality; C3 = no competition and horizontal inequality; C4 = competition and horizontal inequality. When subjects are attributed to conditions with no competition, high and low endowments are attributed randomly. When subjects are attributed to conditions with no horizontal inequality, each of the local groups is composed of two H and two L subjects. Our results show that treatment conditions primarily affect the decision of whether to contribute, whereas endowment mainly determines the amount contributed conditional on contributing. Although the magnitude of the endowment effect varies across treatments, these differences are not statistically robust, suggesting a broadly homogeneous effect across experimental conditions.

Speaker: Mimmi Gustafsson

Title: Information intervention to increase gender equality in expectant mothers’ parental leave plans

Abstract: Parenthood has been identified as a primary driver of gender disparities in labor market outcomes. Women remain the main childcare providers and take the majority of parental leave. Beyond its negative consequences for mothers’ labor market trajectories, this unequal distribution limits fathers’ relationships with their children. Identifying effective measures to encourage a more balanced sharing of leave is therefore central to achieving more equitable economic and social outcomes. This ongoing study investigates whether an information intervention can shift expectant mothers’ plans for parental leave toward a more equal division. The intervention consists of brief information about positive long-term consequences of a more equal division of leave, including a more equal sharing of household work, fewer conflicts, fathers feeling more confident in their parental role, and greater relationship satisfaction. The study is conducted as a pre-registered survey experiment with a between-subjects design. Respondents are expectant mothers recruited from maternal health care centers across several Swedish regions. They are randomly assigned to a treatment group receiving the intervention or an active control group receiving only neutral information about parental leave regulations. A key strength of the study is its high external validity: participants are recruited from the exact population that would be targeted by a real-world policy implementation, enabling follow-up and making results directly relevant for policymakers. Data collection is scheduled for August. At the conference, preliminary findings regarding the intervention’s effect on planned leave division will be presented and discussed.

Speaker: Amir Jafarzadeh

Title: Is Inequality Acceptable? An Experiment on Procedural Fairness

Abstract: Do concerns about the fairness of procedures translate into stronger redistributive preferences? In this paper, we examine the relationship between procedural fairness and preferences for redistribution by designing a novel lab experiment within a new game-theoretic framework based on Sugden and Wang (2020). First, participants play a card game with either equal or unequal opportunity sets, i.e. fair or unfair rules. We use the minimal group paradigm to implement group-based discrimination against disadvantaged players in unfair rules. Second, participants make a dictator-style transfer decision over the resulting unequal payoffs. We find no evidence that redistribution differs between the fair- and unfair-rule treatments, even though participants clearly recognise the group-based unfair rules as unfair. Our findings suggest a disconnect between perceptions of procedural fairness and preferences for redistribution. Although individuals recognise and evaluate procedural unfairness, this recognition does not translate into different redistribution decisions. The paper’s main contribution is the design of a controlled experiment that isolates and identifies this disconnection.

Speaker: Martin Ljunge

Title: Patience Has No Natural Political Home: Inherited Beliefs about Time, Uncertainty, and the Zero-Sum Pie in the Demand for Redistribution

Abstract: Attitudes toward inequality are shaped not only by where people stand in the income distribution, but by what they believe about the future and about how wealth is produced. We study three culturally transmitted orientations—long-term orientation, uncertainty avoidance, and zero-sum beliefs—and show theoretically and empirically that each governs the demand for redistribution through a distinct channel. We formalize a tax-choice model in which long-term orientation sets the weight on the future, uncertainty avoidance the value of insurance against ambiguous income shocks, and zero-sum thinking the perceived cost of taxation via the perceived elasticity of the tax base: a pure fixed-pie believer sees output as appropriated rather than produced, so taxation looks cheap. Zero-sum beliefs and uncertainty avoidance raise desired redistribution, while the effect of patience is conditional, negative precisely when perceived future growth costs outweigh the insurance motive, so the same forward-looking orientation can fuel either opposition to or support for redistribution. We test these predictions with an epidemiological design comparing second-generation immigrants who live in the same Western European country and survey year but differ in ancestral-country beliefs, using European Social Survey matched to World Values Survey and Hofstede measures. A standard deviation of ancestral zero-sum beliefs raises support for redistribution by 4.0 percent of an outcome standard deviation, uncertainty avoidance by 3.6 percent, and long-term orientation lowers it by 2.4 percent, each between a quarter and a half of the education gradient. Semi-structural estimation places the average patience effect on the negative side of the model’s threshold. A companion information-provision experiment manipulates zero-sum framing to test the perceived-tax-cost mechanism directly.

Speaker: Regine Oexl

Title: Perceived discrimination

Abstract: TBA

Speaker: Damiano Paoli

Title: Human Oversight of AI Redistributive Decisions

Abstract: Artificial intelligence is increasingly integrated into high-stakes decisions, from benefits allocation to hiring and healthcare, promising substantial efficiency gains. Human oversight of consequential AI decisions is already required by law, but its effectiveness is not guaranteed. If individuals harbor reservations about AI judgment, they may scrutinize AI decisions excessively, offsetting the efficiency gains that motivated adoption; if they defer automatically, the oversight mechanism fails. This paper investigates whether individuals exhibit AI aversion when overseeing other-regarding decisions made by an artificial agent. I seek to disentangle two potential mechanisms: the black-box effect, arising from uncertainty about the AI’s decision process, and intrinsic AI aversion, a fundamental reluctance to defer to algorithmic judgment. I develop a theoretical framework for the oversight of redistributive choices of humans and artificial agents and test its predictions in a two-session online experiment. The main finding is that participants do not exhibit AI aversion: they do not intervene more when overseeing an AI rather than a human, and their decisions are driven by the expected fairness cost of non-intervention rather than by the identity of the agent. These results suggest that human oversight of AI redistributive decisions is unlikely to generate excessive scrutiny, though this may also limit its effectiveness when AI decisions are biased or mistaken.

Speaker: Andrea Pogliano

Title: Facing Unequal Opportunities: Does Experience Shape Redistributive Preferences?

Abstract: This paper studies whether experiencing unequal opportunities changes redistributive preferences. We design a spectator experiment in which third-party spectators first complete one of two versions of a real-effort task that differ in difficulty and then decide how to redistribute a bonus between two workers, one of whom completed the easy version of the task and one of whom completed the hard version. This design allows us to test whether own experience changes how spectators interpret merit, fairness, and redistribution. We find that spectators who experience the disadvantaged task revise their beliefs about the role of opportunities, view the initial allocation as less fair, and redistribute more toward the disadvantaged worker. This increase persists even when redistribution requires spectators to pay a personal cost, suggesting that experience affects not only fairness views but also the willingness to implement them. In an extension, we show that directly experiencing both the advantaged and disadvantaged tasks generates even stronger redistributive responses, while receiving comparable information through reports or statistics produces weaker effects. The findings suggest that redistributive preferences are shaped by how unequal opportunities are encountered.

Speaker: Maria Luigia Signore

Title: Equity or Efficiency: An Experimental Investigation

Abstract: Policymakers face the challenge of designing redistributive policies that promote equity without undermining economic efficiency. This paper contributes to the debate on the equity-efficiency tradeoff by investigating how individuals allocate resources between themselves and an anonymous counterpart under different levels of efficiency loss and responder veto power. Specifically, we examine whether fairness concerns persist when redistribution becomes increasingly costly. In our experimental design, we distinguish three conditions reflecting the theoretical effects of taxation on output: a Baseline condition with no efficiency loss (lump-sum taxation) and two distortionary taxation conditions generating either Low or High Efficiency Loss. Responder veto power is manipulated through the Ultimatum, Impunity, and Dictator Games. Using the strategy vector method, participants make decisions in both proposer and responder roles. Each participant plays all three games under either the Baseline-Low or the Baseline-High Efficiency Loss treatment. Thus, efficiency loss varies between subjects, while game type and baseline comparison vary within subjects. Our findings reveal that (i) efficiency-seeking behavior increases as the equity-efficiency tradeoff worsens, with the extent of this shift varying across game types; (ii) equity-seeking behavior increases with the responder’s veto power; and (iii) despite changes in efficiency loss and veto power, individuals’ equity preferences remain remarkably stable, suggesting that fairness is rooted in persistent behavioral traits rather than being entirely driven by contextual factors.

Speaker: Magdalena Smyk-Szymańska

Title: Heuristics and Signals: Experimental Evidence on Information and Wage Discrimination

Abstract: Statistical discrimination occurs when employers rely on group-level averages instead of individual productivity under incomplete information. While previous studies have mainly examined hiring decisions, less is known about wage-setting and how individuals update their decisions when accurate performance information becomes available. We report evidence from a laboratory experiment with 223 university students acting as managers allocating wages between a female and a male worker. Managers first observed only statistical information about average gender differences in productivity in either a math or an emotion-recognition task. They then received workers’ actual performance and revised their allocations. Participants were randomly assigned to free or paid information treatments. Managers initially relied on statistical information, demonstrating the use of heuristics under uncertainty. After observing individual productivity, most shifted toward performance-based allocations, reducing reliance on statistical stereotypes. Rather than maximizing their own earnings, managers balanced fairness and self-interest. We also find task-specific patterns: performance was rewarded more strongly in the math task, whereas allocations remained closer to equality in the emotion-recognition task. Charging for information had no significant effect. Our findings contribute to behavioral economics by demonstrating how information quality, heuristics, and fairness jointly shape allocation decisions under uncertainty.

Speaker: Hector Solaz

Title: The Behavioral Limits of Inclusion Policies

Abstract: This paper examines how citizens perceive fairness in social inclusion policies under immigration, demographic pressure, institutional mistrust, and misinformation. We study the Valencian Community in Spain, where immigration is central to welfare-state sustainability but politically contested. Using a representative survey experiment with 1,700 adults and qualitative focus groups, the study tests whether perceived discrimination reflects informed expectations or distorted beliefs. The design uses two randomized vignettes. The first varies only the name and implied origin of a female applicant for the Renta Valenciana de Inclusión (Spanish, Colombian, or Moroccan). The second varies the socioeconomic status of a male applicant seeking dependency benefits while holding need constant. Respondents evaluate expected civil-servant assistance, benefit access, and reasons for failure. Results show robust perceptions of unequal treatment. Respondents expect the Spanish-born applicant to receive worse assistance and have a lower probability of obtaining the benefit than migrant applicants. They also expect the low-status applicant to fare worse than the high-status applicant. These perceptions coexist with deep institutional mistrust and a severe knowledge gap: most respondents claim familiarity with social services, but only a small minority pass an objective knowledge test. Heterogeneity analyses show that perceived discrimination is concentrated among respondents without university education and among those who overestimate their knowledge of social programs. The evidence supports a mechanism of false discrimination: citizens perceive bias against native-born and low-status applicants not because of accurate knowledge of favoritism but because misinformation, low trust, and metacognitive error shape expectations about public administration. The study shows how welfare-state legitimacy can be undermined even when formal policy rules are inclusive.

Speaker: Henrike Sternberg

Title: Non-universal Universalists: In-group Favoritism in a Redistribution Context

Abstract: This paper investigates how individuals decide in the presence of two conflicting moral principles: perceived obligations towards in-group members and distributive fairness. In a representative two-wave online survey of 4,060 Germans, I randomly assigned respondents within- and between-subject to two versions of a spectator game that varied with respect to the salience of these conflicting moral principles but were both designed to elicit the level of universalism, the degree of identical distributive decisions across varying group identities. In the Allocation SG, individuals decided how much of a budget to allocate to in- versus out-group members, while in the Redistribution SG they reallocated existing and fairly obtained monetary amounts. I found that respondents were on average less in-group biased and more universalist in the redistribution context than in the simple allocation context (43% versus 15% full universalism). Mechanism analyses suggest that meritocratic fairness views were an important driver behind an individual’s cross-game difference in universalism. I further used this within-subject heterogeneity to construct different types of universalists—strict particularists, strict universalists, and malleable particularists—defined by the consistency of expressed universalism across games, or put differently by the relative importance of group membership compared to conflicting moral principles.

Speaker: Sofie R. Waltl

Title: What's in a Name? Perceived Ethnic Discrimination in Tokyo's Rental Housing Market

Abstract: This paper studies perceived ethnic discrimination as a market-access friction in Tokyo’s rental housing market. We first show that Japan’s private rental market offers limited statutory protection against nationality-based exclusion, leaving discriminatory screening weakly constrained. We then use a name-based survey experiment in which native Japanese and foreign-born residents evaluate whether otherwise identical applicants would receive a viewing invitation. Both groups expect non-Japanese applicants to face lower access probabilities than Japanese applicants. Foreign-born respondents’ evaluations capture group-level auto-stereotypes of minority disadvantage; native Japanese respondents’ evaluations capture hetero-stereotypes. Native Japanese respondents perceive larger penalties, indicating a Group-Group Discrimination Discrepancy. Among foreign-born respondents, prior everyday discrimination predicts more pessimistic access expectations, yet respondents rate their own prospects more favourably than those of their broader nationality group, consistent with the Personal-Group Discrimination Discrepancy. Platform evidence showing visible nationality-related screening links these perceptions to an institutional setting in which exclusion remains publicly observable.

Speaker: Doris Weichselbaumer

Title: Parental Leave and Discrimination in the Labor Market

Abstract: Encouraging fathers to take parental leave is considered a key policy tool for promoting gender equality, yet little is known about its impact on paternal labor market outcomes. We conduct a large-scale correspondence study in Germany to examine whether leave-taking fathers face hiring discrimination in three occupations that differ in their gender composition. Results show no penalty for fathers who took parental leave in female-dominated or gender-neutral occupations. In contrast, fathers who took long leave are significantly less likely to be invited to job interviews in a male-dominated occupation. Regardless of leave-taking, fathers are treated less favorably than mothers in female-dominated and gender-neutral occupations, but more favorably in a male-dominated occupation. These findings suggest the presence of strong gender norms concerning the perception of ideal employees in different occupations.

Speaker: Yung-Shiang Jasmine Yang

Title: Discriminated but not Discouraged: Investment Responses to Algorithmic Discrimination

Abstract: We study how individuals respond to algorithmic evaluation with group-based priors and whether disadvantage discourages or instead induces compensatory investment. In a pre-registered online experiment (N = 553), participants are randomly assigned to either an advantaged or disadvantaged identity group and choose how much to invest to improve their hiring prospects under an algorithm that applies different prior beliefs across groups. Participants in the disadvantaged group invest 17.7% more than advantaged participants, consistent with compensatory investment rather than disengagement. We then test two interventions designed to encourage investment: information-based Role Model treatments and a No Risk treatment that conditions investment repayment on hiring. While the role model treatments have limited effects, the No Risk treatment increases investment among disadvantaged participants by 24%. Our findings suggest that reducing downside exposure associated with investing may be more effective than social-comparison interventions in encouraging self-investment under algorithmic statistical discrimination.