|
on Experimental Economics |
| By: | McKenzie, David |
| Abstract: | This paper offers practical advice on how to improve statistical power in randomized experiments through choices and actions researchers can take at the design, implementation, and analysis stages. At the design stage, the choice of estimand, choice of treatment, and decisions that affect the residual variance and intra-cluster correlation can all affect power for a given sample size. At the implementation stage, researchers can boost power through increasing compliance with treatment, reducing attrition, and improving outcome measurement. At the analysis stage, power can be increased through using different test statistics or estimands, through the choice of control variables, and through incorporating informative priors in a Bayesian analysis. A key message is that it does not make sense to talk of “the†power of an experiment. A study can be well-powered for one outcome or estimand, but not others, and a fixed sample size can yield very different levels of power depending on researcher decisions. |
| Keywords: | Take-up of government programs |
| JEL: | C93 O1 |
| Date: | 2025–07 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20466 |
| By: | Snowberg, Erik; Yariv, Leeat |
| Abstract: | This introductory chapter outlines key criteria for evaluating experimental measures, and connects these criteria to the selection of experimental parameters across various contexts. We aim for this chapter to serve as a framework for assessing the different measures, elicitations, and designs explored throughout the handbook. |
| Keywords: | Measurement; Experiments |
| JEL: | C90 C91 C93 D90 |
| Date: | 2025–07 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20449 |
| By: | Laferrière, Vincent; Montez, João; Roux, Catherine; Thoeni, Christian |
| Abstract: | We study cooperation in the presence of coordination needs in indefinitely repeated games. In a laboratory experiment, subjects play a Prisoner's Dilemma and a Stag Hunt in each round, either with different partners (single-game contact) or with the same partner (multigame contact), creating strategic interplay across games in the latter case. In theory, multigame contact can strengthen cooperation without undermining coordination as players can link the strategy in one game to their partner’s actions in the other game. In contrast, we observe that multigame contact can reduce both cooperation and coordination rates, and it lowers payoffs. While a significant share of subjects link strategies across games, a large share instead restricts punishment to the game in which the deviation occurred. Such limited deterrence fails to prevent deviations, with occasional cross-game punishment destabilizing both cooperation and coordination and leading to more frequent low-payoff outcomes. |
| Keywords: | Cooperation; Coordination; Indefinitely repeated games; Multigame contact; Experiment |
| JEL: | C72 C73 C91 C92 |
| Date: | 2025–07 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20509 |
| By: | Bigoni, Maria; Boldrini, Michela; Lomys, Niccolò; Tarantino, Emanuele |
| Abstract: | When faced with unfamiliar options, people often rely on the choices of others. We examine how this reliance affects search decisions through a laboratory experiment. Participants choose between two options whose value can only be discovered through costly sequential searches. Some search in isolation; others first observe the final choice, but not the search process, of a peer. This form of social information improves efficiency, yet behavior systematically departs from theory. When selecting which option to sample first, imitation of the peer’s choice is frequent but not universal; moreover, participants often deviate from the optimal stopping rule, with both under and over-search observed. We introduce treatments that allow participants to choose whom to observe and access signals about the reliability of their peers: these interventions increase imitation and improve welfare. Our findings underscore the role of institutional design in facilitating social search, with implications for platform architecture and information diffusion. |
| Date: | 2025–07 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20430 |
| By: | Gazzè, Ludovica; Gupta, Tanu; Huang, Allen; Londono, Valentina; Saavedra, Santiago; Toma, Mattie |
| Abstract: | There is limited evidence on the non-health impacts of air pollution, including productivity in the workplace and behavior. We examine the effect of air pollution on participation, collaboration, and feedback provision in a workplace setting. Our experiment randomly assigns air purifiers to rooms at three large academic conferences to investigate the causal impact of air pollution on participants' engagement behavior. We construct a participant engagement index based on 12 presentation-level behavioral outcomes directly measured by conference observers through an online form and weigh each behavioral outcome using weights elicited from an expert survey. Conference rooms treated with air purifiers exhibit 48% less PM2.5 concentration compared to control rooms. However, we do not find a statistically significant change in engagement. Communication in the workplace might not be a large driver of the empirical relationship between air quality and productivity, albeit more research is needed across workplaces and measures of communication. |
| Keywords: | Engagement; Field experiment |
| JEL: | Q53 J24 |
| Date: | 2025–09 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20676 |
| By: | Braghieri, Luca; Levy, Ro'ee; Trachtman, Hannah |
| Abstract: | We study the drivers of like-minded and low-reliability news-following on social media, as well as the effectiveness of interventions targeting them. In a five-week field experiment with more than 3, 000 U.S. Facebook users, we document the importance of salience-based behavioral frictions in shaping users' news portfolios on the platform. Guided by a theoretical framework, the experiment varies: (i) whether participants are prompted to re-optimize the portfolio of news pages they follow on Facebook through a platform-integrated interface that increases the salience of a balanced set of news outlets, and (ii) whether they receive personalized information about outlet slant and reliability. We find that, consistent with our salience model and in contrast to canonical models of news demand, the re-optimization interface with a salient menu of news pages induces large portfolio changes even without the provision of information; conversely, the provision of information has no effect unless paired with the re-optimization interface. Our interventions produce two main implications for users' news portfolios. First, they move users’ portfolios closer to their stated preferences, mitigating internalities. Second, they reduce the slant and increase the reliability of users' portfolios, thus potentially mitigating negative externalities for democracy. The induced portfolio changes persist for more than a month, translate into measurable changes in online news consumption, and are unlikely to be driven by experimentation motives or experimenter demand effects. |
| Keywords: | Information experiment; Media slant; Salience |
| JEL: | D90 L82 P00 |
| Date: | 2025–09 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20658 |
| By: | Jiao, Peiran; Koedijk, Kees; Xu, Yilong |
| Abstract: | Sustainability preferences are often measured at the individual level. However, it is unclear whether individual sustainability preferences can manifest in a competitive asset market setting. To address this, we conduct a series of experiments, measuring sustainability preferences at both thei ndividual and market levels. In our market experiments, participants could accept lower payoffs for assets that yield positive externalities (green) compared to those with neutral (gray) or negative (brown) externalities. We observe a robust aversion to brown assets leading to significant underpricing (a “brown discount†), but find little evidence of a green premium. When we eliminate speculative motives, green assets’ prices are even closer to the fundamental value, suggesting that any observed green premium is likely driven by speculation but not sustainability preferences. Only incentivized measures of individual sustainability preferences, rather than stated preferences, significantly correlate with participants’ market behavior. |
| Keywords: | Sustainability preference; Experimental asset markets; Green premium; Brown discount; Impact investment |
| JEL: | C92 D81 G12 G14 |
| Date: | 2025–09 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20646 |
| By: | Campante, Filipe; Durante, Ruben; Hagemeister, Felix; Sen, Ananya |
| Abstract: | We study how AI-generated misinformation affects demand for trustworthy news, using data from a field experiment by a major German outlet, Süddeutsche Zeitung (SZ). Readers were randomly assigned to a treatment highlighting the challenge of distinguishing real from AI-generated images. The treatment raised concern with misinformation (0.3 s.d.) and reduced trust in news (0.1 s.d.), including SZ. Importantly, it affected post-survey browsing behavior: daily visits to SZ digital content rose by 2.5% in the immediate aftermath of the treatment. Moreover, we find that subscriber retention increased by 1.1% after five months, corresponding to about a one-third drop in attrition rate. Results are consistent with a model where the relative value of trustworthy news sources increases with the prevalence of misinformation, which may thus boost engagement with those sources even while lowering trust in news content. |
| Keywords: | Generative AI; Misinformation; News media; Trust |
| JEL: | D12 L82 L86 |
| Date: | 2025–08 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20526 |
| By: | Weidmann, Ben; Vecci, Joseph; Said, Farah; Bhalotra, Sonia; Adharvyu, Achyut; Nyshadham, Anant; Tamayo, Jorge; Deming, David |
| Abstract: | This paper introduces and validates a novel approach to measuring management skills. In a large pre-registered lab experiment we causally identify managerial contributions by randomly assigning managers to multiple teams and controlling for differences in individual skill. We find that manager contributions matter greatly for team success, and that people who want to be in charge perform worse than randomly assigned managers. Managerial performance is strongly predicted by economic decision-making skill, but not by demographic characteristics. LinkedIn data show that participants who succeed in the lab are substantially more likely to receive real-world promotions. We also measure the skills of store managers in a large retail firm and find that they predict store sales and other correlates of productivity, which aligns with our experimental results. A one standard deviation increase in manager quality increases annual per-store sales by $4.1 million USD (25% increase). Selecting managers on skills rather than demographic characteristics or the desire to lead could substantially improve organizational performance. |
| Keywords: | Management; Teamwork; Skills; Measurement; Experiment |
| JEL: | M54 J24 C90 C92 |
| Date: | 2025–08 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20575 |
| By: | Aksoy, Cevat Giray; Dolls, Mathias; Klejdysz, Justyna; Peichl, Andreas; Windsteiger, Lisa |
| Abstract: | This paper is the first to causally identify the effect of language on attitudes and behaviors toward debt. Using observational data, randomized survey experiments, incentivized borrowing-to-invest decisions, and textual analyses, we show that the moral connotations inherent in debt-related language influence how individuals, firms, and policymakers perceive borrowing. In several languages, including German, Dutch, and Swedish, the standard term for “debt†also connotes “guilt.†Leveraging this semantic overlap, we conduct survey experiments embedded in three household surveys across seven countries and one business survey. By randomizing debt-related wording, we find that guilt-connoted language significantly reduces individuals’ willingness to borrow, their approval of public debt, and debt-financing plans among firm managers. This effect extends to an incentivized experiment, in which guilt-laden wording significantly reduces both the probability of borrowing and the amount borrowed. To complement the experimental results, we analyze three decades of German parliamentary speeches and find that politicians who oppose public debt systematically use guilt-connoted terms. A similar pattern emerges in financial advertising: guilt-laden language is concentrated in debt-relief campaigns, while neutral terminology dominates transactional promotions. Together, our findings demonstrate that moral connotations embedded in language shape economic behavior and are strategically employed in political and commercial discourse. |
| JEL: | D83 H63 G51 Z13 |
| Date: | 2025–08 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20588 |
| By: | Zara Contractor; Germ\'an Reyes |
| Abstract: | We study how generative AI affects student learning in a randomized experiment. In proctored, in-person sessions, undergraduates learn about an unfamiliar topic and write an analytical essay with or without access to off-the-shelf generative AI, then complete unaided assessments immediately and one week later. We measure learning with knowledge tests (factual and conceptual understanding) and open-ended essays (higher-order skills). AI access raises immediate test scores by 0.27 standard deviations. These gains persist one week later. Essay quality, by contrast, changes little while students have AI access but improves in style and relevance one week later, when students write unaided. These delayed gains are larger among augmentation users-who use AI to explain concepts rather than generate text-whereas automation users' short-run quality gains vanish once AI is removed. We find evidence for two mechanisms behind the learning gains: students shift time away from drafting text and toward reading and searching for information, and they report greater learning enjoyment. |
| Date: | 2026–07 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2607.08849 |
| By: | Hünewaldt, Victoria; Weinig, Max |
| Abstract: | We study how newspaper narratives shape public beliefs and policy preferences during a period of high inflation in Germany 2021–2023. We conducted an online survey experiment with a broadly representative sample of 4, 150 respondents, half of them randomly assigned to one of three inflation narrative treatments: 1) pent-up demand, 2) the energy-price crisis, or 3) corporate price gouging. We find strong baseline political polarization in support for inflation-mitigating policies, while respondents broadly agree on the distributional consequences of inflation. Pre-existing individual causal attributions about inflation correlate with inflation inequality beliefs and policy preferences but prove largely resistant to experimental updating, on average. Among right-leaning respondents, however, exposure to inflation narratives substantially increases support for redistributive and regulatory policies, reducing the baseline partisan gap by up to 80%. This effect is concentrated among respondents on the moderate right who identify with the AfD and exhibit low trust in government, suggesting that, when confronted with inflation as an economic problem, government skepticism is channeled into demand for stronger state protection. |
| Keywords: | inflation;public perceptions;polarization;survey experiment |
| JEL: | D63 D72 D83 E31 E60 |
| Date: | 2026–07 |
| URL: | https://d.repec.org/n?u=RePEc:ehl:lserod:140238 |
| By: | Krishna Dasaratha; Kevin He |
| Abstract: | In an incentivized laboratory experiment, we study how people account for and respond to others' incentives for paying attention. Participants learn a binary state from an attention task under high or low accuracy incentives. We ask subjects to predict their peers' accuracy based on the peers' incentives and to aggregate answers from multiple peers with different incentives. Most subjects fail to consistently understand that peers with stronger incentives are more accurate, and these subjects also perform worse in individual attention tasks. Subjects also participate in a social-learning task where they first learn the binary state from an attention task, then observe a peer's guess about the state in the same task, and finally make a guess themselves. We find behavior in these tasks is inconsistent with leading models of flexible costly information acquisition. In particular, subjects fail to pay more attention when paired with lower incentive peers. Overall, we find that many decision-makers do not respond to others' incentives for accuracy even when those incentives are transparent. |
| Date: | 2026–06 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2606.31850 |
| By: | Luca Braghieri; Ro'ee Levy; Hannah Trachtman |
| Abstract: | We study the drivers of like-minded and low-reliability news-following on social media, and the effectiveness of interventions targeting them. In a five-week field experiment with more than 3, 000 U.S. Facebook users, we document the passive nature of participants' news portfolio formation process and the role of behavioral frictions. The experiment varies whether participants are prompted to re-optimize the portfolio of news pages they follow on Facebook through a user-friendly platform-integrated interface, and whether they receive personalized information about outlet slant and reliability. In contrast to canonical models of news consumption, we find that the re-optimization interface induces large portfolio changes even without new information, while information alone has no effect unless paired with re-optimization. Our interventions produce two main effects: they reduce the slant and increase the reliability of users' portfolios, thus potentially mitigating negative externalities for democracy, and they move users' portfolios closer to their stated preferences, mitigating internalities. |
| Keywords: | news consumption, social media, behavioral frictions, information experiment, slant, choice architecture |
| JEL: | D90 L82 P00 |
| Date: | 2026 |
| URL: | https://d.repec.org/n?u=RePEc:ces:ceswps:_12771 |
| By: | Kevin Bauer; Andreas Grunewald; Florian Hett; Johanna Jagow; Maximilian Speicher |
| Abstract: | We study how behavioral economics and machine learning can jointly construct effective treatment-targeting rules. In a large field experiment at an online fashion retailer with approximately 500, 000 consumers, we test a loss-framed discount message. We elicit individual loss aversion in a nested incentivized behavioral measurement experiment (N=582) and use machine learning to impute it from digital footprints. Targeting based on scaled behavioral measurement yields statistically significant revenue gains and outperforms causal forests. The results show how scaling behavioral measurement can improve algorithmic treatment assignment relative to purely data-driven approaches, especially when pilot data are unavailable, noisy, or costly. |
| Keywords: | treatment targeting, behavioral measurement, machine learning |
| JEL: | C93 C55 D91 M31 L81 |
| Date: | 2026 |
| URL: | https://d.repec.org/n?u=RePEc:ces:ceswps:_12772 |
| By: | Lowes, Sara; Nunn, Nathan |
| Abstract: | Lab-in-the-field experiments, in which lab experiments are conducted in more naturalistic settings, are increasingly being implemented in developing country contexts. In this chapter we outline the conceptual and logistical challenges typically associated with lab-in-the-field experiments in non-Western settings. We describe the importance of worldviews and how researchers’ preconceptions may inadvertently shape the types of research questions that are asked. We emphasize the importance of increasing the diversity of subject pools and researchers. We also suggest a set of best practices when implementing lab-in-the-field experiments in developing countries. |
| Keywords: | Culture; Development; Lab-in-the-field experiment |
| JEL: | C9 N01 N10 Z1 |
| Date: | 2025–06 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20390 |
| By: | Giovanni Di Bartolomeo; Silvia Fedeli; Stefano Papa |
| Abstract: | We test whether minimal, non-informative messages can nudge tax compliance beyond standard deterrence. In a within-subjects lab experiment, we randomize exposure to either a reminder that leaves audit probability unchanged or an informative warningtied to higher audit probability, and estimate e¤ects on both the probability of evasion and the share of income evaded. A short non-informative reminder, holding incentives fixed, lowers the probability of evasion by about 16 percentage points, with no detectable effect on the evaded share among evaders; informative messages add at most marginal effects once audit probability is controlled for. |
| Keywords: | tax compliance; nudge; deterrence; audit; laboratory experiment |
| JEL: | H26 C91 D91 |
| Date: | 2026–05 |
| URL: | https://d.repec.org/n?u=RePEc:sap:wpaper:wp281 |
| By: | Elisabeth Grewenig (KfW, Frankfurt); Klaus Gründler (University of Kassel, ifo Institute, CESifo); Philipp Lergetporer (ifo Institute, CESifo, Technical University of Munich (TUM)); Niklas Potrafke (ifo Institute, CESifo, University of Munich (LMU)); Katharina Werner (ifo Institute, CESifo, Business School Pforzheim); Helen Zeidler (Technical University of Munich (TUM)) |
| Abstract: | Public support for policy interventions depends on citizens’ beliefs about their likely effects. We examine how individuals form such beliefs by studying their predictions of experimental outcomes in a policy-relevant setting, and why their predictions differ from expert benchmarks. We elicit forecasts from 127 professional economists and a representative sample of 6, 200 German households about a large-scale behavioral experiment on education policy (N = 3; 133). Nonexperts predict both average outcomes and treatment effects far less accurately than experts. Prediction accuracy improves with calibrated priors, self-reported effort, and the use of structured reasoning, but remains well below expert levels. We show that scalable design features, including the provision of well-calibrated numerical anchors and monetary incentives to rise effort, improve non-expert predictions, with effects comparable in magnitude to tertiary education or structured reasoning. Our findings have important implications for bridging the ‘expertise gap’ in public discourse. |
| Keywords: | expert forecasts, lay predictions, belief formation, expertise gap, policy support, behavioral experiments |
| JEL: | A11 D83 H52 I22 |
| Date: | 2026–02 |
| URL: | https://d.repec.org/n?u=RePEc:aiw:wpaper:47 |
| By: | Angrisani, Marco; Cipriani, Marco; De Filippis, Roberta; Guarino, Antonio; Kendall, Ryan |
| Abstract: | We study the behavior of professional traders in an experiment in which, endowed with private information, they trade an asset over multiple periods. Professional traders do not produce the price bubbles observed with student subjects and aggregate private information better; bubbles also do not arise in mixed markets with traders and students trading together. Professional traders exhibit higher strategic sophistication than students as measured by their Guessing Game choices. We link strategic sophistication to trading in the financial market, and show that traders with higher levels of strategic sophistication earn higher capital gains and trading profits. |
| Keywords: | Experiment; Bubbles; professional traders; strategic sophistication; Information aggregation; Informational efficiency |
| JEL: | C93 G11 G14 |
| Date: | 2025–07 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20456 |
| By: | Sudarshan, Anant (University of Warwick); Burlig, Fiona (University of Chicago) |
| Abstract: | Developing country governments routinely attempt to collect revenue using threats they cannot systematically enforce. We study how citizens assess the credibility of such empty threats in the context of payment for electricity in Madhya Pradesh, India, where state-run utilities recovered only 60 cents per dollar of power supplied. Using two field experiments covering 30, 000 households with high arrears, we show that the household response to a threat depends on the state’s choice of messenger. The first experiment randomly exposes households to reminders, threats, and enforcement action without changing incentives, state capacity, laws, or information about debt. Legal threats delivered by local linesmen– state agents with a history of ignoring non-payment have no effect. Yet identical notices sent by registered mail, bypassing linesmen, reduce arrears by 11.4 percent among recipients, a 241 percent return-on-investment. We hypothesize that choosing compromised messengers changes household beliefs about the state’s credibility, implying dynamic effects that we test with the second experiment: a year later, we randomly mail a legal notice to previously-treated households. Past treatments affect future responses. Consumers originally visited by a linesman do not respond, while those not exposed to linesmen reduce arrears. Moreover, when we in crease linesman credibility by requiring them to follow up on threats, this gap narrows. The experimental results are together consistent with a model in which consumers use the state’s choice of messenger to infer the threat’s credibility, and demonstrate that low-credibility state agents can render threats ineffective. Low-capacity governments may improve revenue collection by bypassing their agents |
| Keywords: | electricity; threats; messengers; state capacity; taxation JEL codes: O13; Q48; H26 |
| Date: | 2026 |
| URL: | https://d.repec.org/n?u=RePEc:wrk:warwec:1621 |
| By: | Dannenberg, Astrid; Lumkowsky, Marcel |
| Abstract: | The project analyzes how forgiving behavior shapes cooperation in social dilemma settings in which reciprocity is the primary sustaining mechanism. In such environments, individuals rely on signals about others’ behavior to form reputations and guide expectations about future in-teractions. Sustained cooperation depends on rewarding cooperative behavior, punishing de-fections, and restoring relationships after conflict. To study these mechanisms, the project comprised three laboratory experiments with 56 sessions and 860 participants. The experi-ments examine how cooperation emerges under different forms of reciprocity and monitoring. In contrast to much of the literature, particular emphasis was placed on leniency and for-giveness, especially on whether individuals sanction defections immediately and whether co-operation can be restored after retaliation. The first experiment uses an infinitely repeated prisoner’s dilemma with imperfect monitoring, in which signals about a co-player’s behavior can differ from actual actions. Unlike prior work, these discrepancies are endogenous, as participants can deliberately misrepresent their behavior by signaling cooperation while actual-ly defecting. The results show that the possibility of cheating reduces cooperation rates, as many participants engage in misreporting. Forgiveness also declines when cheating is feasi-ble. In the second experiment, we use a more common form of imperfect monitoring, in which false signals are generated by random errors. In this context, we introduce a new set-up in which participants’ histories from earlier supergames are observable in later interactions. We find that participants become more likely to punish defections immediately when such histories are visible, possibly because they do not want to appear exploitable. The third exper-iment focuses on normative evaluations of behavior in the context of indirect reciprocity. The findings indicate that subjects generally view cooperative behavior as good and defective be-havior as bad, regardless of the context of a particular decision. This implies that subjects express limited scope for forgiveness themselves, but also that they do not punish others for being forgiving. Overall, the project shows how signaling, reputation, and forgiveness jointly influence cooperation. The findings advance our understanding of when reciprocity succeeds or fails and provide insights for institutional design under imperfect monitoring and strategic information transmission. |
| Abstract: | Das Projekt analysiert verzeihendes Verhalten in sozialen Dilemmasituationen, in denen Re-ziprozität der zentrale Mechanismus zur Aufrechterhaltung von Kooperation ist. In solchen Situationen orientieren sich Individuen an Signalen über das Verhalten anderer, um Reputati-on aufzubauen und Erwartungen über zukünftige Interaktionen zu bilden. Kooperation beruht darauf, kooperatives Verhalten zu belohnen, Defektionen zu sanktionieren und Beziehungen nach Konflikten wiederherzustellen. Zur Untersuchung dieser Mechanismen wurden drei La-borexperimente mit 56 Sessions und 860 Teilnehmenden durchgeführt. Die Experimente un-tersuchen, wie Kooperation unter unterschiedlichen Formen von Reziprozität und Beobacht-barkeit entsteht. Im Unterschied zu großen Teilen der existierenden Literatur liegt ein beson-derer Schwerpunkt auf Nachsicht und Vergebung, insbesondere auf der Frage, ob Defektio-nen unmittelbar sanktioniert werden und ob Kooperation nach Vergeltung wiederhergestellt werden kann. Das erste Experiment verwendet ein unendlich wiederholtes Gefangenendi-lemma mit nicht-perfektem Monitoring, bei dem Signale über das Verhalten eines Mitspielen-den von dessen tatsächlichen Handlungen abweichen können. Anders als in früheren Studien sind diese Abweichungen endogen, da Teilnehmende ihr Verhalten gezielt falsch darstellen können, indem sie Kooperation signalisieren, tatsächlich jedoch defektieren. Die Ergebnisse zeigen, dass die Möglichkeit solcher Täuschung die Kooperationsraten verringert, da viele Teilnehmende falsche Signale senden. Auch Vergebung nimmt ab, wenn Täuschung möglich ist. Im zweiten Experiment wird eine verbreitetere Form nicht-perfekten Monitorings unter-sucht, bei der fehlerhafte Signale durch zufällige Fehler entstehen. In diesem Zusammenhang führen wir ein neues Design ein, in dem Verhalten aus früheren Supergames in späteren In-teraktionen beobachtbar ist. Die Ergebnisse deuten an, dass Teilnehmende Defektionen in diesem Fall häufiger unmittelbar sanktionieren, möglicherweise weil sie nicht als ausnutzbar erscheinen wollen. Das dritte Experiment befasst sich mit normativen Bewertungen von Ver-halten im Kontext indirekter Reziprozität. Die Ergebnisse zeigen, dass Teilnehmende koopera-tives Verhalten grundsätzlich positiv und defektives Verhalten negativ bewerten, weitgehend unabhängig vom jeweiligen Entscheidungskontext. Dies impliziert, dass die Teilnehmenden selbst nur begrenzt zur Vergebung bereit sind, zugleich jedoch andere nicht dafür bestrafen. Insgesamt zeigt das Projekt, wie Signale, Reputation und Vergebung gemeinsam Kooperation prägen. Die Ergebnisse vertiefen unser Verständnis darüber, wann Reziprozität gelingt oder scheitert, und liefern wichtige Hinweise für die Gestaltung von Institutionen unter Bedingungen unvollständiger Beobachtbarkeit und strategischer Informationsübermittlung. |
| Keywords: | Cooperation, Reciprocity, Social Dilemmas, Forgiveness, Laboratory Experiments |
| JEL: | C71 D91 C91 |
| Date: | 2026 |
| URL: | https://d.repec.org/n?u=RePEc:zbw:esrepo:342352 |
| By: | Burling, Fiona (Harris School of Public Policy and Energy Policy Institute (EPIC), University of Chicago, and NBER.); Sudarshan, Anant (Department of Economics, University of Warwick) |
| Abstract: | Developing country governments routinely attempt to collect revenue using threats they cannot systematically enforce. We study how citizens assess the credibility of such empty threats in the context of payment for electricity in Madhya Pradesh, India, where state-run utilities recovered only 60 cents per dollar of power supplied. Using two field experiments covering 30, 000 households with high arrears, we show that the household response to a threat depends on the state's choice of messenger. The first experiment randomly exposes households to reminders, threats, and enforcement action without changing incentives, state capacity, laws, or information about debt. Legal threats delivered by local linesmen-state agents with a history of ignoring non-payment have no effect. Yet identical notices sent by registered mail, bypassing linesmen, reduce arrears by 11.4 percent among recipients, a 241 percent return-on-investment. We hypothesize that choosing compromised messengers changes household beliefs about the state's credibility, implying dynamic effects that we test with the second experiment: a year later, we randomly mail a legal notice to previously-treated households. Past treatments affect future responses. Consumers originally visited by a linesman do not respond, while those not exposed to linesmen reduce arrears. Moreover, when we in crease linesman credibility by requiring them to follow up on threats, this gap narrows. The experimental results are together consistent with a model in which consumers use the state's choice of messenger to infer the threat's credibility, and demonstrate that low-credibility state agents can render threats ineffective. Low-capacity governments may improve revenue collection by bypassing their agents. |
| Keywords: | electricity, threats, messengers, state capacity, taxation JEL Classification: O13, Q48, H26 |
| Date: | 2026 |
| URL: | https://d.repec.org/n?u=RePEc:cge:wacage:813 |
| By: | Bénabou, Roland; Henkel, Luca |
| Abstract: | We review the economic literature on self-image, which conceptualizes identity as a set of beliefs about one’s core traits, values, goals, and social ties. Self-image concerns lead individuals to process information and make choices in non-standard ways that help affirm and protect certain valued identities. We first present the main cognitive mechanisms involved within a simple unifying framework. We then survey the extensive laboratory, online, and field experimental literature on the nature and behavioral implications of self-image concerns. We discuss in particular how they give rise to information and decision avoidance, motivated memory and beliefs, excuse-driven behavior, preferences for truth-telling, hypothetical bias, moral cleansing and moral licensing, collective identities, political preferences, and other forms of self-signaling or self-deception. We subsequently discuss common empirical strategies used to identify self-image concerns, as well as the threats to their validity and how to alleviate them. We conclude by outlining open questions and directions for future research on the belief-based approach to identity. |
| Keywords: | Self-image; Identity; Motivated beliefs; Belief-based utility; Behavioral economics; Experimental economics |
| JEL: | D64 C90 D83 |
| Date: | 2025–09 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20663 |
| By: | Graeber, Thomas; Noy, Shakked; Roth, Christopher |
| Abstract: | Information often shapes behavior regardless of its quality: unreliable claims wield influence, while reliable ones are neglected. We propose that this occurs in part because word-of-mouth transmission tends to preserve claims while dropping information about their reliability. We conduct controlled online experiments where participants listen to economic forecasts and pass them on through voice messages. Other participants listen either to original or transmitted audio recordings and report incentivized beliefs. Across various transmitter incentive schemes, a claim’s reliability is lost in transmission much more than the claim itself. Reliable and unreliable information, once filtered through transmission, impact listener beliefs similarly. Mechanism experiments show that reliability is lost not because it is perceived as less relevant or harder to transmit, but because it is less likely to come to mind during transmission. A simple associative-memory framework suggests that reliability information may be less likely to come to mind either because it is less likely to be cued by transmission requests or because attempts to retrieve it face greater interference. Evidence from our experiments, a large corpus of everyday conversations, and economic TV news supports both of these mechanisms. |
| Keywords: | Memory |
| JEL: | D83 |
| Date: | 2025–09 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20619 |
| By: | Katharina Brütt (Vrije Universiteit Amsterdam); Eve Ernst (Vrije Universiteit Amsterdam) |
| Abstract: | Paternalistic policies are pervasive, yet little is known about how relationships between decision-makers and targets shape them. This paper examines how social proximity -- the degree to which individuals share identity-defining traits – influences paternalistic interventions. In an experiment with a representative U.S. sample, we manipulate proximity and distinguish between preference-responsive paternalism, reflecting one’s own preferences, and belief-responsive paternalism, reflecting beliefs about others’ preferences. Social proximity leaves the overall frequency of restrictions unchanged but shifts their driver: low proximity fosters preference-responsive paternalism, while high proximity promotes belief-responsive paternalism. Non-religious and independents restrict least; Christians and Republicans restrict more dissimilar others. |
| Keywords: | Paternalism, Social identity, Social proximity, Experiment |
| JEL: | C91 D12 D15 D91 |
| Date: | 2025–12–05 |
| URL: | https://d.repec.org/n?u=RePEc:tin:wpaper:20250068 |
| By: | Dinarte-Diaz, Lelys; Gresham, James; Lemos, Renata; Patrinos, Harry A.; Rodriguez-Ramirez, Rony |
| Abstract: | This paper provides insights into human capital investments during wartime by presenting evidence from three experiments of an online tutoring program for Ukrainian students amid Russia’s invasion of Ukraine. Conducted between early 2023 and mid-2024, the experiments reached nearly 10, 000 students across all regions of Ukraine. The program offered three hours per week of small-group tutoring in math and Ukrainian language over six weeks, and used academic and psychosocial tools to address student challenges at different intensities of disruption. We find that the program led to substantial improvements in learning — up to 0.49 standard deviations in math and 0.40 standard deviations in Ukrainian language — and consistent reductions in stress — up to 0.12 standard deviations. We document high take-up and engagement, and identify four mechanisms driving impact: structured peer interactions, improved attitudes toward learning, enhanced socio-emotional skills, and increased student investments. A complementary experiment using information nudges to increase parental engagement highlights challenges in promoting parental investments in a conflict setting. The program was cost-effective across all experiments, with benefit-to-cost ratios ranging from 31 to 56, and scalable given its reliance on existing infrastructure and teaching capacity. |
| Keywords: | Ukraine; Wartime; Tutoring; Student achievement; Mental health |
| JEL: | I21 I24 |
| Date: | 2025–09 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20693 |
| By: | Castell, Laura; Gurgand, Marc; Imbert, Clément; Tochev, Todor |
| Abstract: | We report on two nationwide experiments with job seekers in France. We first show that a meeting with social services to assess eligibility and help with application to social benefits increased new benefit take-up by 29%. By contrast, an online simulator that gave personalized information on benefit eligibility did not increase take-up. Marginal treatment effects show that individuals who benefit the most from the meetings are the least likely to attend. Overall, without ruling out information frictions, our results suggest that transaction costs represent the main obstacle to applying for benefits or accessing government’s assistance to help apply. |
| Keywords: | Take-up of government programs; Targeting; Welfare; Marginal treatment effects |
| JEL: | H53 I38 |
| Date: | 2025–09 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20615 |
| By: | Mickeler, Maren; Zunino, Diego; Kretschmer, Tobias; Hadengue, Marine |
| Abstract: | Crowdsourcing lets firms tap into diverse sources of ideas and can be implemented either externally, by engaging customers, suppliers, or the public, or internally, by leveraging the expertise of organizational members. On top of the ideas submitted, a critical driver of crowdsourcing success is feedback by peers which helps refine and strengthen the initial submissions. Yet, peer feedback is frequently sparse in many crowdsourcing environments. We examine the role of anonymity in feedback-giving behavior among participants on a crowdsourcing platform. Running a field experiment on an internal crowdsourcing platform of a French business school, we find that anonymity increases the amount of feedback posted on the platform, but it also changes the nature of feedback: Anonymous feedback is shorter, contains less constructive, more negative content, and is less likely to be incorporated by ideators in their final idea proposals. Moreover, total effort per commenting individual is lower under anonymity. Based on our empirical findings, we propose a conceptual framework that explains peer feedback giving on crowdsourcing platforms as a function of two motivators, efficacy considerations and reputational gains, and discuss how anonymity alters the two. |
| JEL: | D23 L23 |
| Date: | 2025–06 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20382 |
| By: | Crossley, Thomas; Fisher, Paul; Levell, Peter; Low, Hamish |
| Abstract: | Different methods of eliciting the Marginal Propensity to Consume give very different distributions. Mean MPCs range from below 0.1, indicating life-cycle consumers, to over 0.5, consistent with consumers being hand-to-mouth. We conducted a randomized survey experiment to test if this difference arises because of question wording: we compare using a direct question and a filtered question. Survey wording has large effects on (i) the mean MPC, (2) the extensive margin, and (3) how MPCs vary with payment size, spending horizon, and liquidity. MPCs elicited using a filtered question are much closer to results from using a covariance restriction approach. |
| Keywords: | Survey experiment |
| JEL: | D12 D14 E21 |
| Date: | 2025–07 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20440 |
| By: | Andrea Albertazzi; Elisabetta Leni; Ennio Bilancini |
| Abstract: | Non-price interventions targeting specific household water uses are increasingly central to conservation policy, but whether end-use savings translate into lower aggregate demand remains unresolved. This paper reports evidence from a pre-registered field experiment in which 775 Finnish households were randomized to a shower timer, a water-saving shower head, or the same shower head with real-time feedback. Utility-grade water meters measure household-level effects, while shower-level data provide complementary end-use evidence for the two shower-head treatments. The shower timer has no detectable effect. In contrast, the water-saving shower head reduces daily household demand by about 5%, and pairing it with real-time feedback doubles this reduction to about 10%. The convergence between shower- and meter-based estimates shows that end-use savings largely pass through to aggregate demand rather than being offset elsewhere in the home. Cost-benefit analysis indicates that combining technological constraint with salient point-of-use feedback dominates reminder-based strategies. |
| Date: | 2026–06 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2606.23347 |
| By: | Kirchsteiger, Georg; Lenaerts, Tom; Suchon, Remi |
| Abstract: | Experimental evidence shows that in a repeated dilemma setting cooperation is more likely in small matching groups than in large ones. This result holds even if cooperation is an equilibrium outcome for all investigated group sizes. But what happens if small matching groups are merged to become large ones? Our paper is based on the idea that due to behavioral spillovers, a large group created by a merger of small groups is more likely to cooperate than a large group of similar size that is created directly. We tested this idea experimentally in the context of an infinitely repeated prisoner’s dilemma game. We compared the cooperation behavior of groups that result from mergers of smaller groups with the cooperation behavior of groups with constant group size. We found that cooperation levels were significantly higher in large groups that resulted from gradual growth than in large groups of the same size that were directly created. Looking at the individual behavior, we see that more subjects adopt lenient strategies when the group size increases than when it is already large from the beginning. Hence, our results confirm the idea that cooperation is much more likely to be achieved when groups grow from small to large than when large groups are formed directly. |
| Keywords: | Prisoner's dilemma |
| JEL: | C73 C92 D23 D90 L22 |
| Date: | 2025–06 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20326 |
| By: | Ali, Nadia; De Giorgi, Giacomo; Rahman, Aminur; Verhoogen, Eric |
| Abstract: | Many countries seek to promote exports by subsidizing market access, but evidence on such efforts has been mixed. We present the first randomized evaluation of a government financial-support program explicitly targeting exports, the Tasdir+ program in Tunisia. The program offered matching grants for fixed market-access costs but not variable costs. Tracking outcomes in administrative data, we find positive effects on exports on average. We find limited impacts on the number of destinations or exported products, which were stated policy targets. The finding that the fixed-cost subsidies expanded exports on the intensive margin but not the extensive margins of destinations or products stands in contrast to the predictions of several workhorse trade models. |
| Keywords: | Trade; Randomized control trial |
| JEL: | O1 F1 |
| Date: | 2025–07 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20398 |
| By: | Pawel Adrjan; Mária Balgova; Simon Jäger; Jonas Jessen; Jason Sockin |
| Abstract: | Discrete choice experiments are widely used to estimate workers’ willingness to pay (WTP) for job amenities under the assumption that varying an attribute does not change workers’ beliefs about other job attributes. We test this assumption by embedding an amenity with a known market price—a popular monthly public transport pass—in a largescale discrete choice experiment with German workers. Many workers, including public transport users, overvalue the ticket by more than 100%, despite WTP for other attributes aligning with the literature. A complementary belief-elicitation experiment shows that advertising an amenity, such as the pass but also common amenities like work from home, causally shifts beliefs about unlisted attributes of the job. Posted wages similarly signal unlisted attributes so that wage variation, the money metric for WTP calculation, is itself contaminated by belief spillovers—such as higher pay signaling heightened stress. These spillovers imply that discrete choice estimates capture perceived bundles rather than isolated attributes, and distort current estimates of non-wage compensation and monopsony power. |
| JEL: | C25 C83 D83 D84 D91 J0 J31 J32 J42 J64 |
| Date: | 2026–07 |
| URL: | https://d.repec.org/n?u=RePEc:nbr:nberwo:35457 |
| By: | Jizhou Liu; Azeem M. Shaikh; Liang Zhong |
| Abstract: | Randomized saturation designs are widely used to study spillover effects in clustered populations. In these designs, clusters are first assigned to treatment saturation levels, and units are then randomized within clusters according to the assigned saturation. This paper develops randomization tests for such experiments under several null hypotheses that arise naturally in spillover analysis. For a fixed pair of saturation levels, we first study two individual-level hypotheses: a partially sharp null of no spillover effect for every untreated unit and a bounded null that restricts individual spillover effects by a prespecified constant. Both hypotheses can be tested using a common conditional randomization framework, with finite-sample validity obtained by combining the same focal-unit relabeling distribution with null-specific statistics. We then study weak average-spillover nulls and show that, although these nulls do not yield finite-sample exact conditional tests, studentized relabeling statistics deliver asymptotically valid randomization-based inference. Finally, for multiple ordered saturation levels, we develop a finite-sample valid unconditional pairwise-imputation test for global monotonicity of spillover effects. Simulations and an application to the Zomba Cash Transfer experiment illustrate the finite-sample behavior and practical implementation of the methods. |
| Date: | 2026–07 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2607.04257 |
| By: | Galbiati, Roberto; Henry, Emeric; Jacquemet, Nicolas |
| Abstract: | What actions other people judge appropriate drives pro-social behavior. We show that such judgments depend on whether the observers previously faced the situation they judge (active observers) or not (passive observers). In an online giving experiment, active observers make more polarized judgments than passive ones — those who acted pro-socially judge selfish behavior more harshly and praise pro-social actions more. Moreover, active observers persistently avoid payoff-relevant information, both as dictators, likely to maintain their self-image, and then as observers. Our results imply a new link between descriptive (what most people do) and injunctive norms (what groups deem appropriate). |
| Keywords: | Injunctive norms |
| JEL: | C90 D64 |
| Date: | 2025–06 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20355 |
| By: | Bursztyn, Leonardo; Haaland, Ingar; Röver, Nicolas; Roth, Christopher |
| Abstract: | Social desirability bias (SDB) is a pervasive threat to the validity of survey and experimental data. Respondents might often misreport sensitive attitudes and behaviors to appear more socially acceptable. We begin by synthesizing empirical evidence on the prevalence and magnitude of SDB across various domains, focusing on studies with individual-level benchmarks. We then critically assess commonly used strategies to mitigate SDB, highlighting how they can sometimes fail by creating confusion or inadvertently increasing perceived sensitivity. To help researchers navigate these challenges, we offer practical guidance on selecting the most suitable tools for different research contexts. Finally, we examine how SDB can distort treatment effects in experiments and discuss mitigation strategies. |
| Date: | 2025–06 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20325 |
| By: | Katharine E. Meyer; Lindsay C. Page; Catherine Mata; Eric Smith; Brendan Tyler Walsh; Candice L. Fifield; Michelle Tyson; Amy E. Eremionkhale; Michael Evans; Shelby Frost; Eye Eoun Jung |
| Abstract: | This study provides pre-registered, experimental evidence on the use of non-generative artificial intelligence (AI) chatbots to support students in large-enrollment undergraduate courses. We find the chatbot messaging increased students’ final grades and engagement with academic supports, such as tutoring. Treatment effects were generally consistent across student demographics, with the exception of treated women in a Microeconomics course, who earned final grades that were seven percentage points higher than women in the control group. This study provides evidence that integrating AI-enabled outreach and communication to students in their college courses can enhance student engagement, learning, and course performance. |
| JEL: | I21 I23 O33 |
| Date: | 2026–07 |
| URL: | https://d.repec.org/n?u=RePEc:nbr:nberwo:35397 |
| By: | Ned Augenblick; B. Kelsey Jack; Supreet Kaur; Felix Masiye; Nicholas Swanson |
| Abstract: | Individuals may fail to recall and use information they already know when making decisions. We empirically investigate whether such “retrieval failures” distort consumption smoothing behavior among Zambian farmers, who derive their income from one annual harvest and then spend it down over the course of the year. We document that individuals underestimate upcoming spending by 50%, creating scope for under-saving. In order to improve recall, we randomize an intervention that prompts individuals to think through their future expenses associatively in categories—without providing any external information or guidance. Treated individuals increase “remembered” expenses by 36-60%; as predicted by the memory literature, effects are concentrated among small, irregular, and stochastic items. Immediate spending drops and, six weeks after the intervention, treated households hold 15% higher savings. They subsequently enter the hungry season—the final months of the year when consumption typically declines sharply—with one additional month of savings, leading to a flatter spending profile over the year. Households use the increased savings to self-finance additional farm investment, resulting in a 9% increase in the next year’s crop revenue. We replicate the intervention’s impact on beliefs among low-income Americans, suggesting that retrieval failures generalize across settings and populations. |
| JEL: | D14 D15 O12 Q12 |
| Date: | 2026–07 |
| URL: | https://d.repec.org/n?u=RePEc:nbr:nberwo:35430 |
| By: | Bhatiya, Apurav Yash; Chakravorty, Bhaskar; Imbert, Clément; Rathelot, Roland |
| Abstract: | We survey young job seekers in rural India to understand the determinants of enrollment in a government training program with guaranteed placement into urban jobs. Respondents are over-optimistic: they expect jobs that pay more and are closer to home than actual placement opportunities. We implement an RCT and provide them with objective information on the distribution of placement salaries or job locations. The intervention successfully corrects subjects’ beliefs, which affects their decision to enroll in the program. By revealed preferences, our estimates suggest that job seekers need to be paid 50% more to work outside their home state. |
| JEL: | D83 D84 R23 O15 |
| Date: | 2025–09 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20617 |
| By: | Ferrando, Mery; Katzkowicz, Noemi; Ubfal, Diego; Le Barbanchon, Thomas |
| Abstract: | This paper provides the first experimental evidence on the long-term effects of work-study programs, leveraging a randomized lottery design from a national program in Uruguay. Participation leads to a persistent 11 percent increase in formal labor earnings seven years after the program, driven by a 4 percent increase in the monthly probability of being employed and a 6 percent increase in monthly wages. Effects are significantly larger for men, while remaining positive for women. The program is highly cost-effective, outperforming most job training programs and reaching levels comparable to early childhood investments. |
| JEL: | I21 I26 J13 J24 J31 O15 |
| Date: | 2025–10 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20798 |
| By: | Bachas, Pierre; Brockmeyer, Anne; Ferreira, Alipio; Sarr, Bassirou |
| Abstract: | Can algorithms enhance bureaucrats’ work in developing countries? In data-poor environments, bureaucrats often exercise discretion over key decisions, such as audit selection. Exploiting newly digitized micro-data, we conduct an at-scale field experiment whereby half of Senegal’s annual audit program is selected by tax inspectors and the other half by a transparent risk-scoring algorithm. Algorithm-selected audits are 18 ppt less likely to be conducted, detect 89% less evasion, are less cost-effective, and don’t reduce corruption. Moreover, even a machine-learning algorithm would only have moderately raised detected evasion. These results are consistent with bureaucrats’ expertise, the task complexity, and inherent data limitations. |
| Keywords: | Algorithms; Firms; Bureaucrats; Taxation; State capacity |
| JEL: | O17 H83 H26 D73 |
| Date: | 2025–09 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20608 |
| By: | David Schönholzer; Gregory Lane; Erin M. Kelley |
| Abstract: | Road traffic injuries are the leading cause of death among people aged 5–29. Where traffic laws are weakly enforced, it is unclear whether road safety interventions can change behavior. We run a randomized controlled trial among 203 Kenyan minibuses, testing whether an incentive scheme improves safety overall or merely shifts unsafe driving to unmonitored times. Drivers reduce speeding by 29 percent and harsh braking by 13 percent, improving a safety index by 0.096 standard deviations. Labor-supply and earnings effects are modest, with little displacement across times or places. Improvements fade quickly, suggesting lasting change requires sustained enforcement or stronger incentives. |
| JEL: | K42 O12 O18 |
| Date: | 2026–06 |
| URL: | https://d.repec.org/n?u=RePEc:nbr:nberwo:35394 |
| By: | Bauer, Michal; Chytilová, Julie |
| Abstract: | Ethnic hostility can fuel violent conflicts between ethnic groups. Thus, a fundamental question is how common desire to malevolently harm ethnic out-group members is, and in what situations it is most prone to manifest. This chapter reviews advances in the study of these questions, revolving along three complementary themes. First, we focus on measurement of ethnic hostility. We describe behavioral tasks designed to identify unambiguously hostile behavior and different ways to signal ethnic identity. We also discuss methodological challenges and ways that researchers have used to address them. Second, we review evidence that tests the existence of ethnic biases in pro-social and hostile behavior. While there is a rich body of lab-in-the-field experiments measuring ethnic biases in pro-social behavior, there are still only a handful of studies measuring ethnic biases in purely hostile behavior. Third, given that many real-life inter-group conflicts are characterized by relatively sudden changes in the behavior of masses towards aggression, an active area of research has begun to explore potential triggers and magnifiers of the strength of group identity, and of group-based pro-sociality and hostility. We review studies that focus on the roles of financial scarcity, stress, exposure to war, and social influences that unravel social norms. We highlight common patterns and suggest open questions in this important research agenda. |
| Keywords: | Randomized experiments |
| JEL: | D74 D91 J15 Z12 C93 |
| Date: | 2025–07 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20402 |
| By: | Sadka, Joyce; Seira, Enrique; Woodruff, Christopher |
| Abstract: | Research shows that lawyers affect legal outcomes, but little is known about how clients choose lawyers. We conduct an experiment with plaintiffs in the Mexico City Labor Court. We compare plaintiffs’ choice of lawyers in a control group and two treatment groups: one that increases the plaintiff’s knowledge of likely case outcomes and the other that facilitates meetings with public lawyers. The knowledge treatment group contracts higher-quality lawyers, with whom they are more satisfied. The treatment halves the market share of low-quality “ambulance chasers†. The meeting treatment increases the use of public lawyers but does not increase satisfaction. |
| JEL: | K31 K41 K42 J52 J83 |
| Date: | 2025–07 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20508 |
| By: | Katharina Brütt (Vrije Universiteit Amsterdam and Tinbergen Institute) |
| Abstract: | Experimentation is at the core of innovation. This project studies collaborative experimentation in teams, focusing on the inherent two-dimensional free-riding problem induced by payoff and informational externalities. The discouraging force of observing others’ unsuccessful experimentation and the attempt to keep one’s team members optimistic theoretically result in inefficiently low experimentation in teams. In a laboratory experiment, I study how distinct elements of the experimentation environment affect this strategic experimentation. I vary (i) the observability of experimentation and (ii) whether agents work on joint or separate projects. Teams largely overcome the free-riding problem. Contrary to theoretical predictions, both the observability of experimentation and experimenting jointly increase experimentation levels. There is no lack of sophistication in updating beliefs that drives this, neither do subjects disregard their experimentation’s effect on others. Instead, the data can be best explained by joint, observable experimentation creating incentives to ‘lead by example’ and setting norms of high experimentation. |
| Keywords: | Conflicting |
| JEL: | D81 D83 D91 |
| Date: | 2025–08–29 |
| URL: | https://d.repec.org/n?u=RePEc:tin:wpaper:20250049 |
| By: | Makoto Nakakita; Teruo Nakatsuma |
| Abstract: | We study temporal assignment design for Markov switchback experiments when the reported object is a dynamic local-projection target. We develop a calibrated selector that chooses the feasible persistence minimizing the covariance, HAC, residual-bootstrap, or realized-schedule risk of the estimator and reporting object specified before the experiment. A balanced homoskedastic Markov benchmark yields a closed form because the lagged-assignment information matrix is AR(1)-Toeplitz with a tridiagonal inverse. The benchmark maps local-projection reporting weights into persistence recommendations within a prespecified first-order Markov class. Field recommendations replace the benchmark covariance with residualized, serially dependent, pilot-calibrated, or randomization-based risk. A semi-synthetic Low Carbon London evaluation uses observed half-hourly baseline dynamics and known injected responses to assess design risk. It evaluates the covariance calculations under realistic load autocovariance and identifies when calibrated covariance selection should replace the homoskedastic Markov formula. Near-boundary designs use randomization-first inference when many-spell normal approximations are unsupported. |
| Date: | 2026–07 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2607.11694 |
| By: | Sievertsen, Hans Henrik; Smith, Sarah |
| Abstract: | This paper reports results from a survey experiment comparing the effect of (the same) opinions expressed by male versus female experts. Members of the public were asked for their opinions on topical issues and shown the opinion of either a male or a female economist, all professors at leading US universities. We find, first, that experts can persuade members of the public – the opinions of individual expert economists have an effect on public opinion – and, second, that the opinions expressed by female economists are more persuasive than the same opinions expressed by male economists. |
| Date: | 2025–07 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20484 |
| By: | Ash, Elliott; Shukla, Soumitra; Sockin, Jason |
| Abstract: | Interviews allow employers to learn about workers, but do they also enable workers to learn about firms? Studying 500, 000 interview reports from Glassdoor, we find candidates for high-paying jobs are more likely to reject a job offer if they believe the interview was easy. Easy interviews appear to convey poor "fit" as those who accept offers after easy interviews are two-fifths of a standard deviation less satisfied with their jobs and 10 percent less likely to remain with their employer for at least one year. Analysis of interview narratives using large language models reveals difficult interviews signal colleague ability whereas easy interviews convey a nonselective process. In a small-scale randomized field experiment, an exogenous increase in difficulty elevated perceived difficulty and boosted applicant engagement with the vacancy. Interviews offer workers a preview of match quality, highlighting a channel through which labor markets may become less efficient if firms automate hiring with AI. |
| JEL: | J24 M50 |
| Date: | 2025–10 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20793 |
| By: | Egorov, Georgy; Guriev, Sergei; Mironov, Maxim; Zhuravskaya, Ekaterina |
| Abstract: | Why do political campaigns so often yield unexpected results? We address this question by separately estimating the direct effect of a campaign on targeted voters and the indirect effect on others in the same social environment. Partnering with a local NGO during Argentina’s 2023 presidential election, we randomized the distribution of leaflets providing an expert assessment of the consequences of the potentially dangerous proposals by the outsider candidate Javier Milei. Exploiting Argentina’s unique sub-precinct election reporting system, we show that the campaign reduced Milei’s support among directly treated voters, as expected, but increased his support among untreated voters in treated precincts, producing a backfiring, net-positive effect for Milei. A pre-registered replication confirmed these opposite-signed effects. Using theory and a survey experiment, we show that the minority of voters who disbelieved the campaign were more motivated to discuss it with peers, convincing them to support Milei. This mobilization effect appears especially likely when campaigns criticize outsider candidates. Our results highlight how campaigns aimed at anti-elite candidates can unintentionally mobilize support for them. |
| Keywords: | Political Information Campaign; Elections; Argentina; Milei; Network effects; Spillovers |
| JEL: | P00 D72 C93 |
| Date: | 2025–10 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20781 |
| By: | Graeber, Thomas; Roth, Christopher; Sammon, Marco |
| Abstract: | Most news stories contain both granular quantitative information and coarse categorizations. For instance, company earnings are typically reported as a dollar figure alongside categorizations, such as whether earnings beat or missed market expectations. We study the hypothesis that when a decision is harder, people rely more on easier-to-process signals: people still discriminate between coarse categories but distinguish less granularly within them, creating higher sensitivity around category thresholds but lower sensitivity elsewhere. Using stock market reactions to earnings announcements, we document that hard-to-value stocks are associated with a more pronounced S-shaped response pattern around category thresholds. Experiments that exogenously manipulate the problem difficulty provide supporting causal evidence in individual investor behavior. We then exploit variation in investor familiarity with earnings surprises of different sizes to show that returns exhibit greater sensitivity in regions with more historical density. |
| Keywords: | Category; Behavioral finance; Attention; Earnings announcements |
| JEL: | D01 |
| Date: | 2025–08 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20518 |
| By: | Dezső, Linda (University of Krems); Ertl, Antal; Chukaeva, Valeriia; Hajdu, Gergely (Vienna University of Economics and Busniess) |
| Abstract: | The spectator task elicits third parties’ inequality acceptance, allowing researchers to estimate population distributions of fairness-preference types. In its within-subject form, spectators make two subsequent choices in situations involving merit- and luck-based inequalities with randomized order. We document a carry-over effect: redistribution is lower in the second choice, and inferred fairness-type distributions vary with choice order. We cannot determine whether this pattern is driven by a psychological mechanism active during repeated choices or a contrast between inequality causes. Either interpretation implies a violation of causal transience and calls for caution when using this design. |
| Date: | 2026–06–26 |
| URL: | https://d.repec.org/n?u=RePEc:osf:socarx:9d5x4_v1 |
| By: | Blesse, Sebastian; Gruendler, Klaus; Heil, Philipp; Hermes, Henning |
| Abstract: | Economic narratives are pervasive in the public discourse and can shape individual behavior. But so far we know very little about whether households actually demand and value narratives as information. We combine a comprehensive expert survey with a large-scale nationally representative household sample in the U.S. to examine the demand for economic narratives in a high-stakes environment of an unprecedentedly high recession probability. We document a substantial willingness to pay for economic narratives of more than 4 USD, which is higher than for numerical forecast information. The dominant motives for acquiring narratives are intrinsic, but a smaller share of participants also lists instrumental motives. Economic narratives improve respondents’ understanding of recession drivers and shape beliefs about the economy and spending, but exert only a minor impact on quantitative expectations. Our findings underscore the potential of narratives as a tool to improve economic understanding and to foster more informed decision-making. |
| Keywords: | Narratives; Experts |
| JEL: | D83 D84 D12 E32 E71 |
| Date: | 2025–10 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20759 |
| By: | Düll, Adrian; Karle, Heiko; Martin, Simon; Schumacher, Heiner |
| Abstract: | The classic search models assume that consumers adhere to a particular method of search (sequential or non-sequential) and that they know the true price distribution. In this paper, we evaluate how well the search cost estimates from classic models predict search outcomes — the amount of search and purchase prices — when these assumptions are violated. To this end, we conduct an online experiment in which we vary searchers' information about the price distribution of a homogeneous good. For each treatment, we (i) estimate search costs, (ii) fit each model to the estimated search cost distribution to obtain in- and out-of-sample predictions about outcomes, and (iii) compare predicted and realized outcomes. We find that the prediction performance of each model is largely robust to violations of the informational assumption. Further, the prediction performance of the sequential and non-sequential search model are similar, despite the fact that the search environment strongly favors sequential search. |
| JEL: | C90 D12 D83 |
| Date: | 2025–10 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20787 |
| By: | Lucie POINET (LEMNA Ð Laboratoire d'Economie et de Management de Nantes Atlantique); Pierre-Alexandre MAHIEU (LEMNA Ð Laboratoire d'Economie et de Management de Nantes Atlantique) |
| Abstract: | Previous literature has widely explored the influence of descriptive social norms on individualsÕ pro-environmental behavior. However, despite a growing interest in the subject, the role of geographical proximity of the reference group remains unclear. Our study seeks to fill this gap by investigating the impact of a descriptive social norm at two scales: local (neighborhood), and global (city). In this aim, we incorporate descriptive norms as attributes in a discrete choice survey. Our findings reveal that only the local social norm exerts a significant influence on organic waste sorting behavior, while the global social norm does not show a significant effect at conventional statistical levels. These results highlight the importance of considering the geographical proximity of the reference group when studying descriptive social norms. The policy implications of these findings are discussed. |
| Keywords: | Descriptive Social Norm, Local Social Norm, Discrete Choice Experiment, , |
| JEL: | C99 D91 Q51 Q53 |
| Date: | 2024–10 |
| URL: | https://d.repec.org/n?u=RePEc:fae:wpaper:2024.09 |
| By: | Abhijit Banerjee; Greg Fischer; Dean S. Karlan; Matt Lowe; Benjamin Roth; Matthew Lowe |
| Abstract: | We ran a nonrandomized, market-level experiment in Kolkata vegetable markets in which we subsidized vendors in some markets to sell additional produce. The vendors earned over 60% higher profits, excluding the value of the subsidy. Nevertheless, after the subsidy ended many vendors stopped selling the additional produce. Vendors knew about the profitable opportunity and demonstrated that they were capable of exploiting it without assistance. We conclude that their behavior meaningfully diverges from profit maximization when considering take-home pay, likely due to a combination of high marginal costs of effort and feared sanctions from breaking anti-competitive norms. |
| Keywords: | microenterprises, profits, vendors, collusion, social norms, separation failures |
| JEL: | O12 D22 L21 O17 L13 L26 |
| Date: | 2026 |
| URL: | https://d.repec.org/n?u=RePEc:ces:ceswps:_12833 |
| By: | Hazan, Moshe; Tsur, Shay |
| Abstract: | We study how shifting intra-household control over resources affects fertility, exploiting a quasi-natural experiment in Israel where some Holocaust survivors began receiving substantial and unexpected reparations in 1957 and others decades later. Using a triple-difference design with heterogeneity by age, we compare fertility outcomes by timing of reparations, gender of the recipient, and age. Households where only the young female partner received reparations early had 0.25–0.4 fewer children than comparable households where only the male was treated. An event study shows that this effect is driven entirely by post-1957 fertility, suggesting a causal link to increased female resource control. |
| Keywords: | Reparations; Holocaust |
| JEL: | J13 J16 D13 |
| Date: | 2025–09 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20629 |
| By: | Bigoni, Maria; Camera, Gabriele; Gallo, Edoardo |
| Abstract: | Globalization offers unparalleled opportunities to expand welfare through cooperation across large networks of unrelated individuals. Social exclusion – permanent or temporary – and monetary exchange are institutions that in theory can incentivize cooperation. In an experiment, we evaluate their relative performance and interaction in anonymous networks of different sizes. Permanent social exclusion (ostracism) reduces long-run economic potential by leading to sparse networks. Monetary exchange and temporary social exclusion perform similarly well in small networks. In large networks, however, monetary exchange is the only institution that promotes full cooperation by crowding out ostracism and keeping the network complete. An insight is that monetary systems outperform social exclusion mechanisms in promoting cooperation in globalized social and economic networks. |
| Keywords: | Social exclusion |
| JEL: | C92 E40 D85 C73 |
| Date: | 2025–07 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20481 |
| By: | Guenther, Laurenz; Nunnari, Salvatore |
| Abstract: | Research on the rise of populism has largely overlooked the explanation populists themselves advance: that they fill political representation gaps, defined as discrepancies between mainstream parties’ policies and the “popular will.†We test this claim in an information-provision experiment conducted in the weeks leading up to the 2025 German federal election. A sample of 5, 040 German citizens was randomly assigned varying information about the immigration stance of Germany’s mainstream center-right CDU—an issue marked by a substantial representation gap. We find that perceptions of the CDU’s position significantly affect both vote intentions and incentivized behavioral measures: when the CDU is perceived as closer to the electorate’s conservative preferences on immigration, support for the right-wing populist AfD declines. Our estimates indicate that the AfD’s vote share would shrink by as much as 75% if the CDU adopted its immigration stance. These results suggest that the electoral success of populist parties is strongly linked to genuine policy preferences, rather than being driven solely by dissatisfaction with political elites or protest voting. |
| Keywords: | Populism; Elections; Immigration |
| JEL: | D72 C93 K37 |
| Date: | 2025–09 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20643 |
| By: | Gary Charness; Eugen Dimant; Daniela Grieco |
| Abstract: | We introduce the “blue lies game, ” a sender-receiver interaction where lying benefits both the liar and her ingroup member at the outgroup’s expense. Across 9, 672 U.S. participants assigned to groups based on either random allocation or political partisanship, 38–57% of senders lie. An ingroup truth-telling norm reduces blue lies, but the same norm attributed to the outgroup has no effect or backfires among strong identifiers. Adding the option to remain silent cuts blue lies sharply and raises aggregate welfare. We structurally estimate via GMM a model that separates lying motives into monetary self-interest, ingroup loyalty, outgroup disregard, and honesty costs, and use it to quantify the welfare losses that blue lies impose. These findings have implications for understanding deception in polarized environments and designing interventions to reduce misinformation. |
| Keywords: | lying behavior, in-group bias, norms, identity, silence |
| JEL: | C91 D63 D71 |
| Date: | 2026 |
| URL: | https://d.repec.org/n?u=RePEc:ces:ceswps:_12827 |
| By: | Isaga, Nsubili; Gafni, Hadar; Ndaki, Pascal Daudi; Garofalo, Orsola; Jeppesen, Lars Bo; Petro, Hawa; Zunino, Diego |
| Abstract: | Resource acquisition is a fundamental entrepreneurial task, where entrepreneurs must persuade resource provider with a pitch. While opportunity entrepreneurs in developed countries have access to multiple sources of finance, necessity entrepreneurs in developing countries find themselves much more constrained. Prosocial crowdfunding has emerged as a viable solution, but these platforms are often intermediated by microfinance institutions and other actors, who craft the pitches on behalf of entrepreneurs and drive up the cost of credit. We investigate whether necessity entrepreneurs can craft an effective pitch with minimal instructions and raise funds in a prosocial crowdfunding setting. We introduced a platform specialized in direct lending (instead of intermediated lending) in Tanzania and ran a randomized control trial across seven locations. The control group learned about the existence of the crowdfunding platform and received only basic instructions about the structure of the pitch; the treatment group was also informed about the predictors of a successful pitch according to crowdfunding research. Our results suggest that the training increased the probability of success by 15.6% and the fundraising speed by 11.5%. Entrepreneurs in the treatment group were 45% more likely to repay their debt and interact on the platform through comments. We detected differential rates in posting the pitch on the platform, which do not allow us to rule out that the results can be explained by the treatment discouraging less able entrepreneurs to seek funds. We discuss these results in terms of training how to pitch as well as governance of prosocial crowdfunding platforms. |
| Keywords: | Crowdfunding; Randomized control trials; Necessity Entrepreneurship; Entrepreneurial Pitching; Microfinance |
| JEL: | L26 O16 G21 |
| Date: | 2025–06 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20352 |
| By: | Biasi, Barbara; Cullen, Zöe; Gilman, Julia; Roussille, Nina |
| Abstract: | This paper provides the first causal evidence on how occupational wage inequality affects the labor movement, using three complementary research designs: a vignette experiment with union organizers, an information intervention during the 2023 Writers Guild of America strike, and a natural experiment following a Wisconsin reform that increased wage inequality among public school teachers. Across all studies, we find that occupational inequality undermines union strength, through multiple channels. First, workers with high individual bargaining power are more likely to withdraw support in unequal environments, preferring individual over collective bargaining. Second, union organizers strategically respond to inequality in ways that may preserve membership but limit redistribution. For instance, they shift away from campaigning on wages and choose smaller, more homogeneous bargaining units. Taken together, our findings highlight the potential for “inequality traps†, where rising inequality erodes the very institutions designed to counteract it. |
| JEL: | J31 J5 D83 D84 C93 |
| Date: | 2025–08 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20566 |
| By: | Evirgen, Sueda; Suetens, Sigrid |
| Abstract: | This paper investigates whether in Germany and France double standards are used when compensating income inequalities due to differences in productivity. To do so, controlled experiments were conducted among general-population samples in which decision makers belonging to the majority group were confronted with an unequal distribution of income between two persons of which one belonged to the majority and the other to an ethnic minority group. One of these persons performed well in a productive task and the other performed poorly. On average, less money was redistributed from rich, well-performing majority to poor, poorly-performing minority persons than from rich, well-performing minority to poor, poorly-performing majority persons, providing evidence of a double meritocratic standard. The effect is driven by those decision makers who hold a negative attitude to ethnic diversity and vote for far-right political parties or not vote at all. |
| Keywords: | Europe |
| JEL: | J15 R23 C99 |
| Date: | 2025–07 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20421 |
| By: | Fornasari, Margherita; Rasul, Imran; Rogger, Daniel; Williams, Martin J. |
| Abstract: | We study ideas generation and innovation in bureaucracies, combining qualitative and quantitative evidence on workplace cultures, workplace climate and bureaucratic performance. We study these issues at-scale in a developing country, using data from bureaucrats in all ministries staffed by the Ghanaian Civil Service. Our qualitative evidence shows these organizations have strong hierarchical cultures, where juniors feel unable to raise innovative ideas, and a lack of resources and systemic Civil Service-wide issues are cited as key bottlenecks for improving organizational productivity. Our quantitative evidence comes from a field experiment training bureaucrats how to break down problems into simple solutions and raise these new ideas with colleagues. We implemented training at the individual level, and at the division-level to bureaucrats working together day-to-day. Our key finding is that individual trainings were more effective in shifting workplace climate towards fostering new ideas, measured 6-18 months post-training. This led individuals to be more likely to raise and discuss new ideas, ultimately improving administrative processes and public service delivery. Division-level training was less effective because divisions failed to integrate in core features of the intervention in terms of the nature of innovations proposed and collective steps to implementation. Rather, division-level plans reflected pre-existing hierarchical workplace cultures that stifle bottom-up incremental innovations and instead, fall back on unrealistically aiming for resource intensive Civil Service-wide change. |
| JEL: | H11 O31 |
| Date: | 2025–06 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20388 |
| By: | Ghorbani, Khashi; Atallah, Shady S.; Gallardo, R. Karina |
| Abstract: | Antibiotic resistance in plant agriculture poses a growing threat to sustainable food production. In U.S. apple orchards, fire blight disease has long been managed primarily with an antibiotic whose efficacy is eroding due to widespread resistance. Decades of antibiotic reliance have accelerated the emergence of resistant bacterial strains. Using less antibiotics, rotating between antibiotics with different modes of actions, and using non-antibiotic alternatives are three pillars of Integrated Pest Management (IPM) strategies to control fire blight while reducing the antibiotic resistance development. Previous studies show that conserving public access resources such as antibiotic efficacy can be exacerbated by neighbors’ adoption. Additionally, state and federal level regulations can be adopted to restrict farmers’ access to resources and suppress over exploitation (i.e. antibiotic resistance). Despite increasing regulatory pressure to restrict agricultural antibiotic use, little is known about how farmers respond to prospective bans or how peer behavior shapes adoption of IPM strategies. This paper fills that gap by examining U.S. apple farmers' willingness to adopt IPM practices that reduce antibiotic dependence and manage resistance, and by investigating how expectations about future regulation and neighbors' behavior affect those decisions. We develop a theoretical model of strategic antibiotic use that generates two testable predictions. First, farmers who anticipate an antibiotic ban optimally front-load antibiotic applications before restrictions become binding, depleting efficacy more rapidly than farmers operating without regulatory expectations. Second, farmers adjust their antibiotic use in response to neighbors' resistance management decisions, exhibiting either free-riding incentives or private conservation incentives. To test these predictions empirically, we design a discrete choice experiment (DCE) embedded with a randomized information treatment and administer it to commercial apple farmers across major apple-producing states in the US. The DCE elicits preferences over fire blight management bundles that vary in antibiotic rotation, biopesticide use, streptomycin application rate, short- and long-run disease control effectiveness, neighbors' resistance management adoption rates, and cost per acre. To isolate the effect of a future ban, respondents randomly received an information treatment stating that a ban on agricultural antibiotics would be implemented in ten years, while the control group received neutral information. Both groups completed two blocks of choice tasks (one before and one after the information intervention) yielding 1, 008 unique choice observations from 126 complete and usable survey responses. We estimate preference heterogeneity using a mixed multinomial logit (MXL) model and identify distinct behavioral segments using a latent class model (LCM). The MXL results reveal that antibiotic rotation is positively valued on average, while biopesticides and reduced streptomycin use exhibit no significant mean effects but considerable variance, indicating the presence of distinct preference types. Critically, the positive and statistically significant interaction between streptomycin use and the ban information treatment confirms the theoretical prediction that exposure to a prospective antibiotic ban increases the marginal utility of antibiotic applications rather than triggering a shift toward alternatives. The ban signal induces a short-run intensification of antibiotic use which is an unintended behavioral consequence consistent with race-to-depletion dynamics documented in other common-pool resource settings. The interaction between biopesticide adoption and the ban is negative but insignificant, suggesting that policy signals alone are insufficient to accelerate the transition to non-antibiotic alternatives. In fact, the negative and significant interaction between antibiotic application and biopesticide implies that farmers consider biopesticides and antibiotics as substitutes. We also find that farmers increase antibiotic applications when they perceive higher levels of resistance management among neighbors, consistent with free-riding behavior, in which individual producers exploit the stewardship efforts of others. The LCM identifies three economically interpretable classes. The IPM adopter class (26%) strongly prefers rotation and biopesticides and reduces streptomycin use. The forward-looking class (28%) discounts antibiotic applications and prioritizes long-run disease control effectiveness and is more likely to anticipate future bans and operate under higher disease pressure. The myopic free-rider class (46%) favors using less streptomycin out of the three IPM strategies to manage resistance, emphasizes short-run effectiveness, discounts long-run outcomes, and is highly responsive to peer adoption, consistent with free-riding incentives. These findings carry important implications for policy design. Regulatory strategies that announce future bans without accompanying transitional incentives risk inducing a short-run surge in antibiotic use. Effective policies should pair prospective restrictions with instruments that reward early IPM adoption and leverage peer dynamics through targeted extension engagement with opinion leaders. Tailoring interventions to the distinct behavioral segments identified here, rather than targeting the average grower, can improve both the uptake and cost-effectiveness of strategies aimed at preserving antibiotic efficacy in specialty crop systems. |
| Keywords: | Crop Production/Industries |
| Date: | 2025 |
| URL: | https://d.repec.org/n?u=RePEc:ags:aaea25:404853 |
| By: | Agustina Affonso; Esteban M. Aucejo; Stephanie El Khoury; Tomás Larroucau |
| Abstract: | We study whether information frictions about the relative effectiveness of study methods distort students’ effort allocation. Standard models treat effort as a single input whose level, rather than composition, determines outcomes. We instead model effort as an allocation across methods with heterogeneous effectiveness. Combining panel surveys, administrative records, digital activity logs, and a randomized information intervention with over 2, 000 undergraduates, we test whether imperfect information leads to effort misallocation and quantify consequences for performance. At baseline, students hold divergent beliefs, and many devote time to passive strategies such as rereading, despite evidence that active retrieval is more effective; these choices predict lower performance and larger self-assessment errors. We randomize treated students into general-feedback and personalized-feedback arms, benchmarking habits against peers in the latter. General feedback has little effect. Personalized feedback widens the perceived effectiveness gap between active and passive methods by 42 percent, shifts time toward effective practices, and raises exam scores by 0.05–0.08 standard deviations. A dynamic model with learning-by-doing and convex allocation costs shows that, for students with the weakest baseline beliefs, personalized feedback is equivalent to a 16.5 percent reduction in adoption costs. These results identify information frictions about multidimensional effort as a determinant of academic performance. |
| JEL: | I2 I21 I23 |
| Date: | 2026–07 |
| URL: | https://d.repec.org/n?u=RePEc:nbr:nberwo:35436 |
| By: | Nicola Gennaioli; Matthias Rodemeier; Gregory Sun |
| Abstract: | On a major e-commerce platform, we randomize linear promotions applying a δ% discount and nonlinear ones applying the same discount above a spending threshold of M€. We unveil a striking fact: nonlinear schemes create bunching at M but no global self-selection above M relative to linear schemes with the same discount δ. This pattern rejects optimization frictions and standard biases. We offer a salience account: upon seeing (M, δ) some people overly focus on M and cut spending above M (where the threshold should not matter), others overly focus on δ and seek discounts below M (where they do not qualify). Using browsing data on attempts to redeem promotions below M, we confirm the model's predictions and show its ability to match the spending anomalies. |
| Keywords: | revealed preference, salience, WARP, welfare, nunching, field experiments |
| JEL: | D01 D12 D40 D60 D91 |
| Date: | 2026 |
| URL: | https://d.repec.org/n?u=RePEc:ces:ceswps:_12769 |
| By: | Po Han Teo |
| Abstract: | Large Language Models (LLMs) are increasingly used as stand-ins in behavioural games. These stand-ins rely on the assumption that the LLM's distribution of choices meaningfully matches how humans play the same game. This study tests that assumption through two games. The first is a p-beauty contest, and the second one is a public goods game. The study first investigates five local-model settings within the same model family. These settings are varied together in a 360-cell factorial, which balances temperature, scale (0.5-32B), quantisation, instruct vs base, and framing. Each cell's distribution is then compared against whole choice distributions in published human data. Each deployment setting, except for quantisation, governs a different aspect of fidelity. Mechanically, while the dispersion of human players can be somewhat recovered through deployment settings, the strategic process behind it cannot. Through the lens of the level-k cognitive theory, we find that LLMs act as static, category-retrieved level-k players, where k is set by the model scale. The models also do not run within-game belief-updating or backward induction throughout multiple-round horizon settings. While human contributions decayed in the public goods game, LLMs stayed flat or rose at every scale. When the horizon test was administered, LLMs were more cooperative under an indefinite horizon compared to a finite one. However, LLMs ignore their relative round position, so no last-round defection was displayed. This implies that LLMs retrieved levels relative to the horizon category rather than working out iteratively from the specific game setting. |
| Date: | 2026–06 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2606.27845 |
| By: | Mengsi Gao; Peng Ding |
| Abstract: | For analytic convenience, existing statistical frameworks either assume random or fixed regressors. However, it is a little awkward that they do not cover the practical case of estimating the average treatment effect in experiments with randomized treatments and non-randomized, fixed pretreatment covariates. We unify the literature by providing the theory for regressions with mixed regressors that contain both random and fixed components. Importantly, our theory allows for misspecification of the regression functions. We first establish general results for estimating equations with both random and fixed components and then use it to analyze misspecified linear regression, with applications to completely randomized experiments. We focus on the causal interpretation of the regression coefficients and standard errors even when the models are wrong. We start with the theory for independent data and then extend the discussion to clustered data. |
| Date: | 2026–07 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2607.09536 |
| By: | Asen Kochov |
| Abstract: | The paper shows that in a dynamic setting with multiple sources of uncertainty, recursive ambiguity-averse preferences may imply a preference for late resolution of uncertainty. In particular, the simultaneous resolution of two sources can offer hedging benefits that are negated when one source is resolved earlier and uncertainty is evaluated recursively. The paper illustrates the mechanism in a simple dynamic urn problem. In a special case, an equivalent static formulation involving composite bets on two urns suggests a direct experimental test. |
| Keywords: | ambiguity, recursive utility, timing attitudes |
| JEL: | D81 |
| Date: | 2026 |
| URL: | https://d.repec.org/n?u=RePEc:ces:ceswps:_12804 |
| By: | Elia Lapenta; Anthony Strittmatter; Pedro Vergara Merino |
| Abstract: | This study proposes a formal, computationally efficient nonparametric omnibus test for treatment-effect heterogeneity that is compatible with a broad class of estimators, including modern machine-learning methods. The test is designed for settings in which identification can rely on high-dimensional controls while heterogeneity is assessed with respect to a low-dimensional subset of covariates. We derive the test statistic's asymptotic null distribution and develop a bootstrap procedure that is efficient because it avoids re-estimating nuisance parameters in each iteration. The testing approach applies to multiple empirical designs, including randomized experiments, selection-on-observables, difference-in-differences, and instrumental-variables settings. Monte Carlo simulations show that the test attains near-nominal size under the null and exhibits good power against heterogeneous alternatives. We further illustrate the procedure using two empirical applications on retirement savings and trade liberalization. |
| Date: | 2026–07 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2607.06412 |
| By: | Kutai, Ari; Saporta-Eksten, Itay; Schlosser, Analia |
| Abstract: | Digital hiring tools challenge older workers accustomed to traditional job search methods. Using a large-scale RCT, we evaluate a job search training program for high-skilled older unemployed individuals focused on the use of online platforms and social networks. The program increased earnings and job stability for men, especially in jobs requiring complex matching, suggesting improved match quality, but had no positive impact on women's earnings. Gender differences cannot be explained by observable characteristics or differential compliance but appear to derive from treated women's increased reservation wages and lower returns to online search tools. |
| JEL: | J08 J14 J16 |
| Date: | 2025–07 |
| URL: | https://d.repec.org/n?u=RePEc:cpr:ceprdp:20494 |