|
on Experimental Economics |
| By: | Gallo, Misty; Wolf, Patrick J.; Patrinos, Harry A. |
| Abstract: | Government programs often experience low take-up, limiting their potential positive impact. We examine whether reducing registration frictions can increase participation in a state-financed literacy intervention program. In a randomized field experiment, families received either encouragement and a simplified enrollment process, or these interventions plus text message nudges, relative to a business-as-usual control group. We find that encouragement and simplified registration increased take-up by more than 10 percentage points, while a fuller treatment of encouragement, simplified enrollment, and text nudges increased take-up by 6 percentage points. We also estimate the impact of these treatments to promote use of funding and amount of funding used to access tutoring. We find positive intent-to-treat effects with the fuller treatment showing a strong impact on use and amount of funding used. These low-cost interventions offer a promising strategy for increasing participation in government programs. |
| Keywords: | Literacy, Tutoring, Take-up, Nudge, Behavioral economics, Encouragement, Regulatory friction, Parental engagement, Choice, Implementation |
| JEL: | I21 D91 C93 |
| Date: | 2026 |
| URL: | https://d.repec.org/n?u=RePEc:zbw:glodps:1836 |
| By: | Rongrong Fu (HU Berlin); Fabian Gerstmeier (TU Berlin); Michel Tolksdorf (TU Berlin) |
| Abstract: | Understanding how decision makers incorporate external information is increasingly important as algorithmic predictions become a common source of advice. The literature on advice-taking suggests that people discount others' advice, especially when they feel competent. We experimentally study how self-confidence affects advice utilization and how algorithmic advice shapes confidence in final judgments. In our online experiment, participants make age predictions and assess the prediction accuracy before and after receiving algorithmic advice. We exogenously vary self-confidence through (perceived) task difficulty and vary the (observable) quality of the algorithmic advisor. While participants generally put more weight on high-quality ("good") advice, we find, contrary to the existing advice-taking literature, that they are less influenced by advice when they perceive the task as difficult. By contrasting age and accuracy predictions, we show that advice can influence beliefs without substantially changing point predictions: people who feel proficient at a task barely adjust their predictions toward high-quality advice but become more confident in them. When people feel less proficient, they are more reluctant to adjust their predictions toward low-quality ("bad") advice. Our design shows that prediction and confidence updating respond differently to perceived task difficulty and advice quality, implying that measures based solely on changes in point predictions can misstate the influence of advice. Comparing observed updating with a Bayesian benchmark further reveals systematic underutilization of advice, driven in particular by insufficient confidence updating following high-quality advice in tasks perceived as easy. |
| Keywords: | algorithmic advice; algorithm appreciation; confidence; online experiment; |
| JEL: | C91 D83 D91 |
| Date: | 2026–10–22 |
| URL: | https://d.repec.org/n?u=RePEc:rco:dpaper:589 |
| By: | Cruces, Guillermo; Fernández Meijide, Diego; Galiani, Sebastián; Gálvez, Ramiro H.; Lombardi, María |
| Abstract: | Does generative artificial intelligence (AI) reinforce or reduce productivity differences across workers? Existing evidence largely studies AI within firms and occupations, where organizational selection compresses educational heterogeneity, leaving unclear whether AI narrows productivity gaps across individuals with different levels of education. A related question is whether the productivity gains from AI reflect productive use of the tool or mere delegation, fading once AI is unavailable, and whether the answer differs across education groups. We address both questions using a randomized online experiment outside firms, in which 1, 174 adults aged 2545 complete an incentivized, workplacestyle business problem-solving task with or without a generative-AI assistant, followed by a non-AI-assisted follow-up module. AI increases performance for all participants, with substantially larger gains for lower-education individuals. In the control group without AI, higher-education participants outperform lower-education participants by 0.548 standard deviations; with AI, this gap falls to 0.139 standard deviations, closing about three-quarters of the initial gap. Chat-log measures explain both why this gap narrows and why it does not disappear: lower-education participants successfully obtain substantial assistance from AI, while higher-education participants use the tool somewhat more effectively across several margins. The follow-up results showthat these gains are not purely driven by delegation; treated participants do not perform worse than controls once AI is removed, and lower-education participants retain part of their gain, although a sizable education gap re-emerges. Consistent with this interpretation, intensive use of the assistant produces strong task performance regardless of how much time and effort participants devote themselves, but follow-up performance is substantially higher only when intensive AI use is combined with sustained effort on the task. We interpret these findings as evidence that generative AI narrows effective productivity differences in task execution while underlying human-capital differences continue to shape unassisted performance and effective use of the tool. |
| Keywords: | Productivity;artificial intelligence;Education;human capital;Inequality |
| JEL: | J24 O33 |
| Date: | 2026–10 |
| URL: | https://d.repec.org/n?u=RePEc:idb:brikps:14819 |
| By: | Rosino, Maria Victoria; Scartascini, Carlos |
| Abstract: | How citizens want artificial intelligence (AI) regulated will shape which AI governance instruments are politically feasible. Yet we know little about public preferences over AI policy in middle-income democracies, where state capacity is contested. This paper provides the first multi-country experimental evidence on preferences over AI labordisplacement policy instruments in middle-income democracies. Using a randomized framing experiment in a sur-vey of Argentina, Brazil, Colombia, and Mexico, we examine how citizens map beliefs about AI-driven job disruption into demand for three governance responses: worker-retention reg-ulation, tax-financed redistribution to displaced workers, or non-intervention. Nearly three-quarters of respondents prefer employment-preserving regulation, about one in twenty favors tax-financed redistribution, and one-fifth prefer nonintervention. A three-arm framing taxonomy spanning alarmist, delayed-threat, and task reallocation narratives produces no average shift in this distribution, and the heterogeneous effects we detect at the margin run from non-intervention toward regulation, not toward redistribution. The pattern is consistent with preferences anchored in priors about the state's redistributive capacity rather than in correctable beliefs about AI, with direct implications for AI governance design in low-credibility institutional environments. A within-survey wording experiment on employment protection regulation elicits a substantial shift in preferences among the same respondents, indicating that the null on the AI framings is not an artifact of respondent inattention. |
| Keywords: | regulation;employment protection;Trust;Survey experiment;AI governance;Labor Market Regulation;Public Opinion;Framing Experiment |
| JEL: | D72 D83 J38 O33 |
| Date: | 2026–10 |
| URL: | https://d.repec.org/n?u=RePEc:idb:brikps:14824 |
| By: | Jozwik, Jan (School of Economics, University of Nottingham Ningbo China); Lane, Tom (Newcastle University Business School); Song, Lina (Nottingham University Business School); Zhang, Fan (Independent researcher) |
| Abstract: | Building on evidence that group cohesion drives economic efficiency, we hypothesise income inequality influences efficiency by altering cohesion, with the direction depending on inequality type. Using a novel experimental design with extensive group interactions and a state-of-the-art group cohesion measure, we confirm stronger cohesion is associated with more efficient interactions. Group cohesion is strengthened by inequality between groups, a mechanism previously overlooked. Furthermore, unexpectedly, inequalities within groups do not hinder cohesion. These results suggest interpersonal relationships display resilience to economic disparities, challenging the prevailing argument that inequality must harm cohesion, and undermining a pathway through which it could reduce efficiency. |
| Keywords: | economic inequality, group cohesion, social identity, cooperation, coordination, public goods, weak-link game, economic experiment |
| JEL: | C99 D31 D64 D91 |
| Date: | 2026–10 |
| URL: | https://d.repec.org/n?u=RePEc:nwc:wpaper:2026-02 |
| By: | Giorgia Barboni; Lucas A. Mariani; Jose Renato Haas Ornelas; Bernardo Ricca |
| Abstract: | Does improving knowledge about open finance increase its adoption and lead to greater financial inclusion? We implement a randomized experiment that provides Brazilian microentrepreneurs with targeted information about open finance. The intervention increases open finance adoption by more than 7 percentage points, a 40% increase relative to a control group mean of 17%. While the intervention increases adoption broadly across the study population, the impacts are largest among firms with greater baseline digital payment (Pix) activity, consistent with lower adoption barriers for digitally engaged firms. Analysis of firm-level administrative data shows that the intervention increases the probability of having an available credit card limit by 3.6–5.1 percentage points, or approximately 5–7.5% relative to the control-group mean. Consistent with this effect, the intervention increases the probability of credit-card spending by 2.7–4.1 percentage points, equivalent to approximately 4–6% of the control-group mean. Over a six-month horizon, the intervention also increases access to credit through business bank accounts to a greater extent than through personal bank accounts, raises reliance on working capital loans by 4.8 percentage points, equivalent to 50% of the control-group mean, and the number of Pix transactions in business accounts by 15%. |
| Date: | 2026–10 |
| URL: | https://d.repec.org/n?u=RePEc:bcb:wpaper:653 |
| By: | Kemper, Jan; Rostam-Afschar, Davud |
| Abstract: | The advantages of adaptive experiments have led to their rapid adoption in economics, other fields, and among practitioners. However, adaptive experiments pose challenges for causal inference. This note suggests a BOLS (batched ordinary least squares) test statistic for inference on treatment effects in adaptive experiments. The statistic provides a precisionequalizing aggregation of per-period treatment-control differences under heteroskedasticity. The combined test statistic is a linear combination of heteroskedastic per-period z-statistics and can be used to construct asymptotically valid confidence intervals. We provide simulation results comparing rejection rates in the typical case with few treatment periods and few (or many) observations per batch. |
| Keywords: | Adaptive experiments, Heteroskedasticity, Causal inference, Randomized controlled trial |
| JEL: | C12 C13 C9 D83 |
| Date: | 2026 |
| URL: | https://d.repec.org/n?u=RePEc:zbw:glodps:1854 |
| By: | Deborah A. Cobb-Clark; Juliana Silva-Goncalves; Agnieszka Tymula; Xueting Wang |
| Abstract: | Self-control, attention, and impulsivity jointly support goal-directed behavior yet are often examined in isolation with heterogeneous measures. We integrate validated self-reported scales with a hybrid, incentivized behavioral task that synthesizes two canonical experimental designs and adds novel extensions to measure these constructs in a typical work context. Using data from a preregistered four-session online study (N = 443 adults), we characterize the cross-relationships and the relationships between different measures of self-control, attention, and impulsivity and evaluate their contributions to effort allocation over time. More broadly, the study advances an emerging research program that leverages laboratory settings with well-structured economic incentives to examine the role of personality traits and cognitive limitations in economic decision making. |
| Keywords: | self-control, impulsivity, attention, measurement |
| JEL: | D81 D83 D91 |
| Date: | 2026 |
| URL: | https://d.repec.org/n?u=RePEc:ces:ceswps:_13111 |
| By: | Bogler, Lisa; Neubauer, Florian; Rose, Julian; Ankel-Peters, Jörg |
| Abstract: | Mohanan et al. (2021a) test the effect of input and output incentive contracts for health providers on maternal and neonatal health outcomes using a randomized controlled trial in Karnataka, India. They find that both contract types led to a reduction in the probability of experiencing postpartum hemorrhage and that the input incentive contract did so more cost-effectively. The authors cautiously suggest heterogeneity by provider qualification for output contracts, claiming that providers with higher qualification performed better than those with lower qualification under the output incentive contract but not under the input incentive contract. We use the authors' reproduction package to replicate their results and to perform a series of robustness checks. The first claim regarding the effectiveness of both contract types holds in our robustness reproduction. The evidence for the paper's second-order heterogeneity claim is tentative and sensitive to how provider qualification is operationalized. The pre-analysis plan prespecifies heterogeneity by provider 'Qualifications' but does not define the measure; the published paper operationalizes skill/human capital more narrowly as advanced obstetric specialization (MBBS plus). Using an alternative operationalization of the broader preregistered qualification construct - MBBS-or-higher versus BAMS/other - yields negative interactions under both contract types. We interpret the difference as sensitivity to how the broader preregistered qualification construct is measured, not as evidence that MBBS status is the uniquely correct measure of clinical skill. |
| Date: | 2026 |
| URL: | https://d.repec.org/n?u=RePEc:zbw:i4rdps:345 |
| By: | Ankel-Peters, Jörg; Bensch, Gunther; Brodeur, Abel; Hauser, Christina Sarah; Neubauer, Florian; Rose, Julian |
| Abstract: | Using a randomized controlled trial, Islam, Lee and Nicholas (2021) study the impact of a short-term chess program for primary school children on over 15 cognitive and non-cognitive outcomes, with a particular focus on time and risk preferences. The only significant effect is a reduction of children's risk aversion. This main finding of the paper survives the robustness checks proposed in this reproduction analysis. The study is not pre-registered. A pre-analysis plan is not available but would have been desirable given that over 20 hypotheses are tested. As part of a broader forensic review of research conducted in the same region by members of the author team, we additionally assess the integrity of this study's data and documentation. Raw data were not provided because the subcontractor, according to the authors, "does not store the data for an extended period". We can nevertheless show that the study's sample overlaps with the sample of other studies that have been withdrawn or retracted in the meantime. We provide evidence that much of the study sample was likely exposed to another treatment shortly before the chess treatment - an overlap that is not mentioned in the paper. Last, we reveal several anomalies - most notably, highly unusual distributions of risk-preference choices in two schools, with responses concentrated at opposite extremes of the risk scale that amplify the only significant finding of the paper. |
| Date: | 2026 |
| URL: | https://d.repec.org/n?u=RePEc:zbw:i4rdps:340 |
| By: | Lisa Bruttel (University of Potsdam, Berlin School of Economics, CEPA) |
| Abstract: | This paper introduces a simple incentivized measure of individual attitudes towards strategic uncertainty. This measure is based on the idea that the size of the difference between the expected profit from playing a game and the value that an individual places on playing that game provides information about how much they like or dislike the strategic uncertainty involved. This measure explains trusting decisions in the trust game, where standard risk aversion measures have no explanatory power. Furthermore, it correlates across different games and predicts behavior associated with strategic uncertainty outside the laboratory. |
| Keywords: | strategic uncertainty, risk, trust, elicitation method, laboratory experiment |
| JEL: | C91 D81 D83 D91 |
| Date: | 2026–10 |
| URL: | https://d.repec.org/n?u=RePEc:pot:cepadp:107 |
| By: | Frauke Baumeister; Philipp Lergetporer; Katharina Wedel; Katharina Werner |
| Abstract: | How do adolescents respond to occupational advice from their parents? We study this question in a survey experiment with matched parent-adolescent data (N=1, 227) from Germany. We ask parents how strongly they would recommend their own occupation and that of the other parent, randomly reveal these recommendations to their child, and then measure the child’s occupa-tional preferences. Revealing parental advice does not generally increase adolescents’ interest in their parents’ occupations. Instead, adolescents often adjust their preferences in the opposite direction to parental recommendations. These non-conforming responses are particularly pro-nounced among older adolescents and when parents and adolescents envision different occupa-tional fields for the adolescent’s future. |
| Keywords: | occupational choice, advice, occupational preferences, adolescents, matched parent-child data, survey experiment |
| JEL: | J24 J62 D83 |
| Date: | 2026 |
| URL: | https://d.repec.org/n?u=RePEc:ces:ceswps:_13122 |
| By: | Sarah Cattan (Institute for Fiscal Studies); Pulkit Bajpai (World Bank); Julian Edbrooke-Childs (AFC); Emily Stapley (UCL); Monica Costa Dias (Institute for Fiscal Studies); Jenna Jacob (AFC); Emily Goodacre (AFC); Angelika Labno (AFC); Vaishnavi Dhas (IFS); Emily Orchard (AFC); Erin Mckeaveney (AFC); Grizelda Khaling (AFC); Anoushka Kapoor (AFC); Ashley New (AFC); Imran Rasul (Institute for Fiscal Studies) |
| Date: | 2026–10–28 |
| URL: | https://d.repec.org/n?u=RePEc:ifs:ifsewp:26/64 |
| By: | Angei, Fabio; Balia, Silvia; Brau, Rinaldo; Sulis, Giovanni |
| Abstract: | We study whether alternative narratives about Artificial Intelligence (AI) shape workers' support for strengthening labor market policies and whether these responses differ by workplace AI use. We conduct a four-arm survey experiment on a representative sample of 2, 105 Italian employees. Respondents are randomly assigned to a control group or to one of three information treatments framing AI as a technology that may substitute for human labor, complement workers' tasks, or increase the need for reskilling. We then elicit support for five labor market policy instruments, including both passive and active measures. Average treatment effects on policy support are small, but mask substantial heterogeneity by workplace AI use. Among non-users, the substitution narrative increases support for publicly funded training and hiring subsidies, while complementarity increases support for training. Among workplace AI users, both substitution and complementarity reduce support for unemployment insurance and short-time work schemes. Reskilling produces no statistically significant effects in either group. We interpret these heterogeneous effects using evidence that, among users, the substitution narrative increases the probability of expecting household financial losses from public intervention addressing AI's labor market effects, without a detectable change in perceived personal job-loss risk. Overall, workers' responses to the same information differ by workplace AI experience, both in the magnitude and direction of changes in support for specific labor market policies. The findings suggest that support for labor market protection may evolve as AI exposure and perceived vulnerability become more salient. |
| Keywords: | artificial intelligence, labor market policy, workers' perceptions, policy support, survey experiment |
| JEL: | J65 J08 O33 D83 C93 |
| Date: | 2026 |
| URL: | https://d.repec.org/n?u=RePEc:zbw:glodps:1856 |
| By: | Malte Kornemann (University of Bonn); Philipp Schirmer (Leibniz Institute for Financial Research SAFE) |
| Abstract: | Most economic decisions require aggregating multiple attributes into an overall evaluation. We develop and experimentally test a theory where aggregation is error-prone and, in response, attributes are weighted optimally given an environment of likely options. We characterize these decision weights, which generically differ from attributes’ impact on utility, and the bias they induce in choice. When attributes are independently distributed among likely options, higher-variance attributes are over-weighted, and this bias increases in the level of aggregation errors. We test the theory in an experiment by first manipulating between subjects which attribute varies more in choice options, and then comparing choices between options that are identical across subjects. Consistent with the model’s predictions, choice probabilities shift toward the option favored by the high-variance attribute, in our baseline treatments by about 10 percentage points. We also test the effect of aggregation complexity by lowering the number of attributes. Compared to the baseline, this reduces the shift in choice probabilities by about 72%. We apply our model to purchases where base and add-on prices may be displayed separately. A consumer reacts to a short-term change in the display as if it were a price change, while her reaction to a long-run change depends on the details of the environment. Consumer welfare is generally maximized by a fully aggregated price. |
| Keywords: | Aggregation errors, cognitive noise, multi-attribute choice, decision weights, complexity, adaptation |
| JEL: | C91 D01 D81 D91 |
| Date: | 2026–11 |
| URL: | https://d.repec.org/n?u=RePEc:ajk:ajkdps:442 |
| By: | F. Angei; S. Balia; R. Brau; G. Sulis |
| Abstract: | We study whether alternative narratives about Artificial Intelligence (AI) shape workers' support for strengthening labor market policies and whether these responses differ by workplace AI use. We conduct a four-arm survey experiment on a representative sample of 2, 105 Italian employees. Respondents are randomly assigned to a control group or to one of three information treatments framing AI as a technology that may substitute for human labor, complement workers' tasks, or increase the need for reskilling. We then elicit support for five labor market policy instruments, including both passive and active measures. Average treatment effects on policy support are small, but mask substantial heterogeneity by workplace AI use. Among non-users, the substitution narrative increases support for publicly funded training and hiring subsidies, while complementarity increases support for training. Among workplace AI users, both substitution and complementarity reduce support for unemployment insurance and short-time work schemes. Reskilling produces no statistically significant effects in either group. We interpret these heterogeneous effects using evidence that, among users, the substitution narrative increases the probability of expecting household financial losses from public intervention addressing AI's labor market effects, without a detectable change in perceived personal job-loss risk. Overall, workers' responses to the same information differ by workplace AI experience, both in the magnitude and direction of changes in support for specific labor market policies. The findings suggest that support for labor market protection may evolve as AI exposure and perceived vulnerability become more salient. |
| Keywords: | Artificial Intelligence, labor market policies, workers’ perceptions, policy support, survey experiment |
| JEL: | J65 J08 O33 D83 C93 |
| Date: | 2026 |
| URL: | https://d.repec.org/n?u=RePEc:cns:cnscwp:202617 |
| By: | Iocco, Maria Pia; Ankel-Peters, Jörg; Neubauer, Florian; Rose, Julian |
| Abstract: | Banerjee, Duflo, Imbert, Mathew, and Pande (2020) study a "just-in-time" payment reform in India, first using a randomized controlled trial in Bihar and then examining the reform's subsequent nationwide, non-randomized rollout. The RCT finds lower programme expenditure, longer payment delays, more days worked, and fewer "ghost workers, " while the difference-in-differences analysis of the national scale-up finds lower labour expenditures after implementation. This paper replicates these findings. The computational reproduction is largely successful, reproducing most tables of the original paper. However, three errors are identified: two minor, and one that shifts the significance of the ghost-worker outcome from the 5% to the 10% level. Robustness checks show that two of the five selected outcomes, programme expenditure and payment delays, are very robust. The remaining three outcomes are sensitive even to relatively modest analytical variations, several of which were already considered in the original paper or are straightforward alternatives to the published specifications. For number of days worked and the ghost-worker outcome, most alternative specifications are insignificant at the 5% level, and several remain insignificant even at the 10% level. Likewise, the national scale-up outcome, labour expenditure, is only partially robust, with most alternative specifications insignificant. These findings leave the evidence for some direct effects of the reform intact, but weaken the evidence for reduced leakage and for the reform's effectiveness when implemented at national scale. |
| Date: | 2026 |
| URL: | https://d.repec.org/n?u=RePEc:zbw:i4rdps:341 |
| By: | Giovanni Immordino (University of Naples Federico II and CSEF); Mario Macis (Johns Hopkins University and NBER); Immacolata Marino (University of Naples Federico II and CSEF); Fabrizio Panebianco (Università Cattolica del Sacro Cuore di Milano) |
| Abstract: | When a new technology promises large private benefits but may impose social costs that markets do not price, demand need not reveal how citizens want it governed. We examine this using a nationally representative U.S. survey experiment (N=5, 556) on human enhancement technologies (HET). The experiment randomizes benefit domain, mechanism, heritability, purpose, and risk across vignettes; for each respondent’s assigned vignette, we elicit stated adoption, preferred regulation, and ethical and societal concerns. Overall, about 53% would adopt. Framing the technology as enhancing rather than restorative lowers adoption by about five percentage points, as much as a severe side-effect profile. About 28% would not adopt at any benefit. This refusal is driven overwhelmingly by the enhancing framing rather than by risk, consistent with a non-compensatory constraint for a substantial subgroup. Most who would adopt still favor strict regulation, and most who would never adopt do not wish to forbid others from doing so. Productivity enhancement generates the most ethical concern of any attribute but attracts the least regulation, and respondents favor subsidizing rather than taxing its adoption, consistent with a concern about access rather than safety. Private demand is therefore an unreliable guide to the governance citizens want, and the divergence we document provides a basis for regulators seeking to align the direction of technical change with societal values and priorities. |
| Keywords: | Human Enhancement Technologies, Survey Experiment, Private–Policy Alignment, Distributional Effects and Inequality, Productivity, Moral Foundations and Ethics. |
| JEL: | I18 O33 D81 D63 J24 L51 |
| Date: | 2026–08–04 |
| URL: | https://d.repec.org/n?u=RePEc:sef:csefwp:794 |
| By: | Tiziana Assenza; Anna Mogilevskaja; Stefanie Huber; Tobias Schmidt |
| Abstract: | We use an experimental module in the Bundesbank Online Panel Households to show that the estimated link between inflation expectations and household consumption changes sign de pending on survey wording. This finding reconciles prior contradictory results and has direct implications for central bank survey design. Our experiment systematically varies elicitation framing of consumption questions along three dimensions: the reference unit (individual vs. household), the time horizon (past one, 3, or 12 months), and the question type (attitudi nal, planned, qualitative and quantitative recall-based). We find that mainly question type significantly influences the estimated relationship between inflation expectations and durable consumption. While the average association is weak, its sign and magnitude vary strongly with question design. Planned spending and attitudinal questions, such as whether it is a good time to buy, produce very similar negative associations, suggesting that respondents interpret the former as a proxy for future intentions. In contrast, quantitative recall-based questions on past spending yield a modestly positive or insignificant link. These results high light the critical role of survey design in shaping behavioral measurements, offering a novel explanation for mixed findings in the literature and guidance for both research and policy. |
| Keywords: | Inflation expectations; household consumption; survey design; question wording; framing effects; measurement. |
| JEL: | C83 D12 D84 E21 E31 |
| Date: | 2026–11 |
| URL: | https://d.repec.org/n?u=RePEc:bon:boncrc:crctr224_2025_791 |
| By: | Yang Zhong (University of Warwick) |
| Abstract: | Distractions are pervasive in modern workplaces, from frequent interruptions to noisy open-plan offices. Using an incentivized laboratory experiment, I study performance and well-being under task-switching and auditory distractions, elicit willingness to pay to avoid them, and develop a survey measure of distraction resilience that captures individuals’ willingness and ability to work under distractions. Neither distraction significantly reduces performance, but both reduce well-being. Participants are willing to pay to avoid distractions, with those experiencing lower well-being under distractions willing to pay more. I find no gender differences in the effects on performance or well-being, although women are willing to pay less than men to avoid both distractions. I validate the survey measure of distraction resilience by showing that those scoring higher experience better well-being under both distractions. I then incorporate the validated measure in a representative Dutch panel and find that greater distraction resilience is related to higher income, a greater likelihood of holding a supervisory position, and higher job satisfaction |
| Date: | 2026 |
| URL: | https://d.repec.org/n?u=RePEc:wrk:warwec:1638 |
| By: | Marie-Claire Broekhoff; Ron Berndsen |
| Abstract: | This study examines which features drive consumer adoption of new digital payment methods in the Netherlands. We conducted a Discrete Choice Experiment in the LISS panel in November 2025 among 2, 084 respondents and analyse preferences for six features: the origin of the payment service provider, the functionality, app requirements, privacy, delayed payments, and automatic payment functionalities. The results show that the origin of the provider is the strongest determinant of adoption. Relative to payment methods offered by a Dutch provider, otherwise identical methods offered by European, American or Asian providers are significantly less preferred. Convenience also matters. Respondents prefer payment methods that are integrated into an existing banking app, can be used both online and in stores, and support recurring automatic payments. By contrast, additional features such as offline use and conditional payments add limited utility. A latent class model reveals substantial heterogeneity, distinguishing a small group of reluctant non-adopters from several adopter segments that differ in their sensitivity to the attributes. The findings imply that a broad adoption of new digital payment methods is most likely when trusted governance is combined with convenient, well-integrated functionality. Policymakers, banks and other PSPs should communicate clearly why new payment methods are being developed and how their features benefit consumers in everyday use. |
| Keywords: | digital payments; consumer adoption; discrete choice experiment; payment preferences; |
| JEL: | D12 C93 G41 E58 |
| Date: | 2026–11 |
| URL: | https://d.repec.org/n?u=RePEc:dnb:dnbwpp:876 |
| By: | David M. Levinson (TransportLab, School of Civil Engineering, University of Sydney) |
| Abstract: | At 168 Sydney intersections, estimated pedestrian delay averages 39.8 seconds per crossing if everyone waits for Walk, or 29.8 seconds under observed crossing behavior, compared with 10.4 seconds per detected vehicle. We test earlier pedestrian service while holding pedestrian and vehicle volumes at their current levels. With a 30-second first-call wait target, the 95-site experiment saves 440 pedestrian-hours but adds 886 vehicle-hours, assuming full Walk compliance. At 1.2 occupants per vehicle, pedestrian waiting must count more than 2.42 times as much as vehicle-occupant waiting for the weighted time savings to exceed the added delay. In the seven-day experiment at 35 intersections, the weighted time savings exceed the added delay when pedestrian waiting counts more than 2.22 times as much as vehicle-occupant waiting. |
| Keywords: | pedestrian delay; vehicle delay; traffic signals; modal comparison; signal timing; time allocation |
| JEL: | R40 |
| Date: | 2026–11–02 |
| URL: | https://d.repec.org/n?u=RePEc:nex:wpaper:paper-2026-29 |
| By: | Sahel Iqbal; Tom Rainforth |
| Abstract: | Standard Bayesian experimental design (BED) methods assume the experimenter alone will act on the data, choosing the best action given their own beliefs and goals. In practice, data are often acted on by external parties whose procedures the experimenter cannot change and may not know, particularly when the data are released publicly. To extend BED to such settings, we introduce usage-aware experimental design, a framework in which an adaptive design policy is trained to minimise the usage risk: the experimenter's expected loss when one or more external rules act on the collected data. These rules may be non-Bayesian or black-box, and when several may apply, we either average over them or take the worst case. We show that, unlike in standard BED, more informative data can be harmful under such rules, so the usage risk cannot in general be expressed as an expected uncertainty reduction; our framework thus strictly generalises standard BED. We then use it to tackle a range of problems that are difficult to address with existing approaches. |
| Date: | 2026–10 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2610.22032 |
| By: | Patrick Arni; Bart Cockx; Andrey Launov |
| Abstract: | How do unemployed workers form beliefs about their labor-market prospects, and how do these beliefs aect behavior, policy eectiveness, and welfare? We develop and estimate a structural job-search model with quasi-Bayesian updating, combining Swiss panel data on beliefs and job-search behavior with a randomized counseling-and-coaching intervention. Job seekers enter unemployment with overly optimistic beliefs about interview arrivals, but gradually become pessimistic as past signals fade from memory. The estimated updating process combines features consistent with both motivated beliefs and recency-weighted information processing. The model replicates the observed dynamics of applications, interviews, wages, and job nding, while decomposing treatment eects into changes in search eciency, search costs, and beliefs. We nd that belief formation is the principal channel through which the intervention accelerates transitions into employment. However, despite increasing job nding, the intervention reduces participant welfare because it operates primarily through lower reservation wages. |
| Keywords: | Job search, biased beliefs, expectations, quasi-Bayesian learning, randomized control trial, structural econometric estimation, evaluation, active labor market policies |
| JEL: | C41 C51 C93 D04 D83 J64 J68 |
| Date: | 2026–10 |
| URL: | https://d.repec.org/n?u=RePEc:rug:rugwps:26/1155 |
| By: | Thomas Graeber (University of Zurich); Christopher Roth (LMU Munich, CESifo & NHH); Florian Zimmermann (University of Bonn & MPI for Behavioral Economics Bonn) |
| Abstract: | This article reviews how different types of information shape belief formation. We organize the discussion around two distinctions: whether information supplies evidence or a frame work for interpreting evidence, and whether it is qualitative or quantitative. Stories describe individual instances in qualitative terms, whereas statistics summarize many observations numerically. Narratives and quantitative models provide qualitative and quantitative frame works, respectively, for interpreting evidence. We examine how learning from evidence and learning from models interact: models shape responses to existing and subsequent evidence, while evidence can prompt model revision. We argue that qualitative information engages associative cognitive processes, such as mental simulation and memory retrieval, which tend to give it a larger and more lasting influence on beliefs than quantitative information. We then discuss implications for strategic communication and the supply of information in media markets, and identify challenges and open questions for future research. |
| Keywords: | Stories, statistics, narratives, beliefs, learning, qualitative information, models |
| JEL: | D01 D83 |
| Date: | 2026–10 |
| URL: | https://d.repec.org/n?u=RePEc:ajk:ajkdps:441 |
| By: | Saanvi Chawla; Vasilis Syrgkanis |
| Abstract: | Interference between participants in two-sided marketplace experiments can severely bias conventional A/B tests. While two-sided randomization (TSR), which randomizes both supply and demand, enables estimators that mitigate this bias, existing analyses offer limited guidance on which estimator to prefer for the balanced market conditions common in practice, as they largely focus on extreme supply- or demand-constrained regimes. We introduce the Double Diagonal (DD) estimator, a novel method for interference-robust treatment effect estimation. The DD estimator contrasts the outcomes of treated-treated and control-control pairs and uses the off-diagonal interactions (treated-control and control-treated) to correct for interference-induced bias. We provide the first theoretical comparison between the DD and the standard naive TSR (TSRN) estimator under a simplifying mean-field market model, proving two key results: (i) DD's bias is strictly smaller than the standard estimator's below a critical threshold of the outside-option parameter; in particular, DD's bias is smaller whenever the outside option is at most twice as attractive as platform options; and (ii) the two estimators exhibit biases in opposite directions, with DD underestimating and the standard estimator overestimating the true effect, respectively. Extensive numerical experiments in realistic market settings--spanning heterogeneous treatment effects, supply-demand imbalances, and varied market structures--corroborate our theoretical findings. Across a wide range of configurations, the DD estimator demonstrates the lowest bias, establishing it as a robust and often superior choice for reliable experimentation in two-sided marketplaces. |
| Date: | 2026–10 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2610.41294 |
| By: | Olivia Bertelli (DIAL-LEDa, UMR CNRS 8007 & IRD 260, Universite Paris-Dauphine, PSL Research University); Thomas Calvo (IRD, DIAL-LEDa, UMR CNRS 8007 & IRD 260, Universite Paris-Dauphine, PSL Research University); Emmanuelle Lavallée (DIAL-LEDa, UMR CNRS 8007 & IRD 260, Universite Paris-Dauphine, PSL Research University); Lydie Lazure (DIAL-LEDa, UMR CNRS 8007 & IRD 260, Universite Paris-Dauphine, PSL Research University); Marion Mercier (CNRS, DIAL-LEDa, UMR CNRS 8007 & IRD 260, Universite Paris-Dauphine, PSL Research University); Sandrine Mesplé-Somps (IRD, DIAL-LEDa, UMR CNRS 8007 & IRD 260, Universite Paris-Dauphine, PSL Research University) |
| Abstract: | Gender-based violence (GBV) is widespread across the world, and Mali is one of the countries characterized by the highest prevalence rates. Following up on a paper studying men’s justifications for and perpetration of GBV in Bamako, the capital city, we elicit women’s attitudes towards early marriage, female genital mutilation (FGM) and intimate partner violence (IPV). We administer four list experiments (LEs), whose results are then compared to those of direct questions to estimate response biases. Additionally, we question women about their second-order beliefs on men and women’s attitudes towards FGM and IPV in Bamako. Our results show that women tend to misreport their attitudes towards GBV. While they declare attitudes that are less favorable to early marriage and IPV than what the LE reveals, the reverse is true as far as FGM is concerned. Women, like men (Bertelli et al., 2025), are less conservative than what they openly declare about FGM. Besides, women’s perception of the support for FGM aligns relatively well with what men and women, respectively, openly declare. The misperception of the actual prevailing attitudes towards FGM could participate in the persistence of a practice which is, in fact, much less supported than what people declare and believe. |
| Keywords: | gender-based violence, FGM, list experiment, second-order beliefs, Mali. |
| JEL: | I12 J12 J16 O55 |
| Date: | 2026–10 |
| URL: | https://d.repec.org/n?u=RePEc:dia:wpaper:dt202607 |
| By: | Conrad, Christian; Schoelkopf, Julius Theodor; Weber, Michael; Brueckbauer, Frank |
| Abstract: | Professional forecasters' heterogeneous narratives about how inflation affects stock markets in 2023 rationalize their high disagreement regarding quantitative expectations for 12-month-ahead inflation and stock returns in December 2022. Professional forecasters causally update their return expectations in heterogeneous directions in response to information about the inflation outlook, depending on the narrative they entertain. Moreover, the narratives also affect their asset allocation in a hypothetical portfolio-choice experiment. Providing common signals does not necessarily lead to convergence of beliefs if agents subscribe to different narratives. |
| Keywords: | Narratives; Expectation Formation; Inflation; Beliefs; Monetary Policy; Subjective Models; RCT; Inflation Expectations; Return Expectations |
| Date: | 2026–11–04 |
| URL: | https://d.repec.org/n?u=RePEc:awi:wpaper:0776 |
| By: | Vincent Boucher; Semih Tumen; Michael Vlassopoulos; Jackline Wahba; Yves Zenou |
| Abstract: | We study the social integration of ethnic minority children in early childhood. We embed a randomized field experiment in a two-month Turkish preschool program for disadvantaged 5- year-old native Turkish and Syrian refugee children. Within each school, we randomly assigned children to classrooms with a low, medium, or high share of refugees, generating exogenous variation in interethnic exposure. We find that interethnic exposure fosters friendships and improves Syrian children’s Turkish language skills, with effects that persist into primary-school grades nine months later. We find no corresponding cost to the academic outcomes of native Turkish children. Our findings show that early-childhood classroom composition can shape integration in ethnically diverse early-education settings. Nous étudions l’intégration sociale des enfants appartenant à des minorités ethniques dès la petite enfance. Nous intégrons une expérience de terrain randomisée à un programme préscolaire de deux mois en Turquie, destiné à des enfants défavorisés de cinq ans, turcs et réfugiés syriens. Au sein de chaque école, nous avons réparti aléatoirement les enfants dans des classes présentant une proportion faible, moyenne ou élevée d’enfants réfugiés, générant ainsi une variation exogène de l’exposition interethnique. Nous constatons que cette exposition favorise la formation d’amitiés interethniques et améliore les compétences en langue turque des enfants syriens, avec des effets qui se prolongent jusque dans leurs résultats scolaires à l’école primaire, neuf mois plus tard. Nous ne constatons aucun effet défavorable correspondant sur les résultats scolaires des enfants turcs. Nos résultats montrent que la composition des classes dès la petite enfance peut influencer l’intégration dans des milieux d’éducation préscolaire caractérisés par une diversité ethnique. |
| Keywords: | Refugees, early childhood, friendship networks, language acquisition, Réfugiés, petite enfance, réseaux d’amitié, acquisition linguistique |
| JEL: | D85 J15 J18 Z13 |
| Date: | 2026–10–29 |
| URL: | https://d.repec.org/n?u=RePEc:cir:cirwor:2026s-23 |
| By: | Sylvia Klosin; Jonathan Roth |
| Abstract: | We develop standard errors for estimates of causal effects in settings where the distribution of treatment exposures is known but potentially exhibits complicated dependence patterns across units. Such dependence arises, for example, in settings with spillovers/interference across units or when the treatment assignment is spatially correlated. Our variance estimator generalizes several properties of the Neyman variance estimator for completely randomized experiments to general designs: it is unbiased for an upper bound on the variance from the design-based perspective; the conservativeness of this upper bound depends on the heterogeneity of treatment effects; and it is also valid under certain forms of super-population sampling of the potential outcomes. Our estimator reduces to standard approaches in simple designs, such as the Neyman variance estimator in completely randomized experiments and the cluster-robust variance estimator in cluster-randomized experiments. We show that the proposed estimator is admissible among the class of quadratic-form variance estimators, and derive asymptotic conditions under which it is consistent for an upper bound on the asymptotic variance of the treatment effect estimator. In simulations and applications based on two empirical studies, one with peer effects and another with a spatially correlated treatment, we find that the proposed variance estimator performs well, and in some cases yields standard errors that are 2-3 times shorter than those of existing alternatives. |
| Date: | 2026–10 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2610.12674 |
| By: | Rommel, Jens; Neubauer, Florian; Rose, Julian |
| Abstract: | This reproduction report reassesses the findings of Cai, de Janvry, and Sadoulet (2020), who conducted a randomized controlled trial (RCT) on insurance uptake among rice farmers in rural China. The original study found that payout experience increases future insurance adoption, particularly when combined with financial education. We reproduce the paper's main results using the publicly available data and code. The computational reproduction is successful after correcting a minor coding error that had negligible effects on estimates. We focus on four key outcomes corresponding to the paper's main claims - namely, short-term insurance uptake following payout experience with and without a full subsidy and long-term insurance uptake following payout experience conditional on financial education. We conduct an extended multiverse of robustness checks, combining analytical choices related to control variables, risk attitude definitions, and analysis samples. Across 23 alternative specifications per outcome, all paths support the original results. Effect sizes are highly consistent with the original study, and statistical significance patterns remain unchanged. However, the study's construct and external validity are constrained by a limited documentation of the survey instruments, financial training, and implementation procedures. Greater transparency about these aspects would improve the generalizability of the results. |
| Date: | 2026 |
| URL: | https://d.repec.org/n?u=RePEc:zbw:i4rdps:343 |
| By: | Christian Conrad; Julius Theodor Schoelkopf; Michael Weber; Frank Brueckbauer |
| Abstract: | Professional forecasters' heterogeneous narratives about how inflation affects stock markets in 2023 rationalize their high disagreement regarding quantitative expectations for 12-month-ahead inflation and stock returns in December 2022. Professional forecasters causally update their return expectations in heterogeneous directions in response to information about the inflation outlook, depending on the narrative they entertain. Moreover, the narratives also affect their asset allocation in a hypothetical portfolio-choice experiment. Providing common signals does not necessarily lead to convergence of beliefs if agents subscribe to different narratives. |
| JEL: | D83 D84 E31 E44 E52 E71 G11 G12 |
| Date: | 2026–10 |
| URL: | https://d.repec.org/n?u=RePEc:nbr:nberwo:35981 |
| By: | Nadja Born; Jared R. Curhan; Hannah Riley Bowles |
| Abstract: | Professionals are increasingly turning to artificial intelligence (AI) for career advice. We tested whether large language models (LLMs) provide different advice to female and male job candidates in the context of career negotiations. Across two large-scale audit experiments, including four leading models, the LLMs not only recommended lower salary targets for female than male personas, but also constructed distinct negotiation agendas, prioritizing flexibility, caregiving, and well-being more highly for female candidates, and financial rewards, mobility, and advancement more highly for male candidates. These differences closely tracked independent human ratings of employment terms associated with primary caregiver and breadwinner roles, respectively. Our findings identify a subtle pathway through which personalized AI advice may reinforce gender inequality at work: steering otherwise equivalent users toward traditional gender roles. |
| Date: | 2026–10 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2610.14036 |
| By: | Davide Domenico Pace (LMU Munich, Cesifo); Joël J. van der Weele (LMU Munich) |
| Abstract: | Information about climate change is a commonly used tool to help consumers and voters make decisions that better align with their climate concerns. Most research efforts have focused on consumers, often in the form of carbon labelling. We review recent evidence from large-sample field experiments on the impact of carbon labels on real-life consumption, and find that most effects are null or small. Moreover, carbon labels mostly work by making climate change salient, so effects are often transient. Instead, a younger literature strives to help voters understand systemic policies, building on the finding that voters often misperceive their impact and distributional effects. We review this growing literature, and we discuss the risk that information targeting consumers crowds out support for systemic policies. In light of the promising early findings, we propose that research on information-based strategies shift emphasis from consumers to voters, and focus on giving the latter a better understanding of the available options. |
| Keywords: | climate change; consumer behaviour; carbon labelling; climate policy; voters' understanding; |
| JEL: | D72 D83 D91 H23 Q54 |
| Date: | 2026–11–02 |
| URL: | https://d.repec.org/n?u=RePEc:rco:dpaper:590 |
| By: | Hörl, Moritz; Keinprecht, Michael |
| Abstract: | Rankings of social mobility depend critically on the dimension and indicator used. Societies that rank highly on one measure may rank poorly on another, yet existing research offers limited guidance for selecting among competing measures. We provide the first systematic evidence on preferences over four dimensions of social mobility (education, occupation, income, and wealth) and three indicators of income mobility: the rank-rank slope (RRS), absolute mobility (AM), and upward mobility from the bottom (CER25). Using a conjoint experiment and a within-subject rating design, we examine how a broadly representative US sample (N = 1, 002) and experts (N = 99) rate different dimensions and indicators of social mobility. We embed these preference-based approaches in a social-welfare framework to shed light on the underlying assumptions of the different indicators of social mobility. We report five sets of findings. First, each indicator is an exact welfare- sufficient statistic for a distinct social objective, and no indicator satisfies all normatively appealing axioms. Second, income is the dimension the public values most, in a ranking that is stable across demographic subgroups. Third, the RRS, the indicator most prominent in the academic literature, receives the least public support. AM and CER25 receive greater support, with roughly 70% of the gap associated with lower perceived fairness. Fourth, experts share the public's ranking of dimensions but support CER25 more and AM less. Finally, incorporating these preferences into a social mobility composite measure substantially attenuates the Great Gatsby curve and shifts US state rankings by up to 32 positions. These findings provide an empirical basis for selecting social mobility measures and designing evidence-based policy targets aligned with public preferences. |
| Keywords: | social mobility; intergenerational persistence; public preferences; Great Gatsby curve; social welfare |
| Date: | 2026–11 |
| URL: | https://d.repec.org/n?u=RePEc:wiw:wus005:92322777 |
| By: | Huiquan Zhang (Graduate School of Economics, Kobe University, JAPAN); Junyi Shen (Research Institute for Economics and Business Administration, Kobe University, JAPAN) |
| Abstract: | As environmental, social, and governance (ESG) information increasingly reaches consumers through certifications, labels, and package claims, it remains unclear whether such information translates into measurable preferences and willingness to pay (WTP). Focusing on Chin's fast-moving consumer goods (FMCG) market, this study examines consumer preferences, willingness to pay, and preference heterogeneity for ESG labels in two product categories: bottled drinks and shampoo. A discrete choice experiment is employed, and the data are analyzed using multinomial logit (MNL), random parameters logit (RPL), and latent class logit (LCL) models to estimate the marginal utilities of ESG labels across product contexts and consumer segments. The results show that consumers generally exhibit positive preferences for ESG labels and make meaningful trade-offs between label attributes and price, while substantial heterogeneity exists across labels, product categories, and consumer segments. Among the tested labels, ISO Certification shows the strongest and most consistent positive effect across both product categories and is associated with the highest WTP. The findings highlight differences in the valuation of specific ESG-related cues across product contexts and consumer groups, suggesting that how such information reaches consumers warrants attention alongside its ESG dimension. |
| Keywords: | ESG labels; Willingness to pay; Discrete choice experiment; Chinese FMCG market; Consumer preferences |
| JEL: | C35 D12 Q56 |
| Date: | 2026–10 |
| URL: | https://d.repec.org/n?u=RePEc:kob:dpaper:dp2026-29 |
| By: | Yero Samuel Ndiaye (University of Cologne & Max Planck Institute for Behavioral Economics) |
| Abstract: | Why do people supply information slanted toward their own positions? In four experiments, I investigate whether this slant reflects deliberate selection, which motives drive it, and how it affects those receiving information. People curate the truth, choosing which facts to pass on and to whom. Even with the usual explana tions ruled out, they value sharing facts that support their own political position beyond anything they expect those facts to do to others. They also want to cor rect and persuade others. However, they share less with those who hold different views on policy, even though their own expectations give them strategic reason to share more with them. Whenreceivers make real political donations, people select more on expected persuasion. Professional journalists select in a similar way, even without market or newsroom pressures. Individuals from whom facts are with held end up with less accurate beliefs and barely notice the selection. Selection among true information thus distorts what people believe without any falsehood, at a margin current policies such as fact-checking do not target. |
| Keywords: | Information Provision, Selective Disclosure, Political Information, Media Bias |
| JEL: | D83 D82 C91 D72 L82 |
| Date: | 2026–11 |
| URL: | https://d.repec.org/n?u=RePEc:ajk:ajkdps:445 |
| By: | Dang, Hai-Anh; Behrman, Jere |
| Abstract: | Nearly half of 15-year-old students in OECD countries, and more in some poorer countries, use AI weekly for learning. Generative AI makes schoolwork help cheap for connected students, but cheaper help need not yield lasting learning. We link the economics of supplementary education with learning science to examine how effort, entry costs, public provision and assessment shape AI tutoring's benefits. Randomized trials find gains and losses varying by setting, use and outcome, while lasting, equitable gains remain uncertain. We distinguish program effects from generative AI's contribution, compare full costs with alternatives, and propose decision rules and a trial design. |
| Keywords: | artificial intelligence, tutoring, learning, educational inequality, public investment, randomized trials |
| JEL: | I21 I24 I28 H52 O33 |
| Date: | 2026 |
| URL: | https://d.repec.org/n?u=RePEc:zbw:glodps:1861 |
| By: | Alex Chan; Ayush Gupta; Yetong Xu |
| Abstract: | Cadaveric organ shortages leave thousands without life-saving transplants each year. Countries differ in using opt-in (informed consent) or opt-out (presumed con sent) systems for donor registration. Using newly assembled cross-country panel data and an event-study design, this paper shows that presumed-consent laws are associated with increased organ donation only when strictly enforced and fam ily veto power is limited; weak opt-out regimes show negligible or even negative changes. A theoretical signaling model provides a plausible mechanism that de termines when opt-in or opt-out yields more donations, emphasizing the roles of donation propensity, signaling costs, and the family’s ability to overturn defaults. A large laboratory experiment further tests this mechanism, showing that opt-in generally produces equal or higher donation rates unless signaling is costly and family veto power is minimal. The results underscore that defaults alone rarely increase donations unless paired with strong institutional enforcement. |
| Keywords: | Defaults, Nudge, Organ Transplantation, Market Design, Donation, Healthcare, Giving, Behavioral Economics |
| JEL: | C92 D64 D47 H00 I11 I18 |
| Date: | 2026–11 |
| URL: | https://d.repec.org/n?u=RePEc:bon:boncrc:crctr224_2025_792 |