This research investigates the impact of social media reviews on corporate environmental performance, with particular emphasis on the influence of reviewers’ professional expertise. We propose a two-stage social learning model to capture consumer product selection, the dissemination of reviews and firms’ decisions regarding environmental investments. Theoretically, our analysis demonstrates that social media reviews help bridge information gaps left by traditional product information sources and incentivize firms to enhance their environmental performance. Notably, reviews authored by professional reviewers have a more pronounced effect on motivating such investments. Empirically, we validate these theoretical insights using an eight-year dataset comprising daily observations from 125 publicly traded companies, which includes 294,727 expert reviews and 2,677,228 non-expert reviews sourced from online platforms. The empirical findings corroborate the theoretical predictions. Additionally, we observe that increased consistency and transparency in firms’ disclosed environmental values mitigate information asymmetry and reduce dependence on potentially biased expert evaluations.

Cheap, disposable online identities make abuse easier to externalize. Users can harass, evade bans, amplify content through fake accounts, or abandon a damaged reputation at low cost, while other users, moderators, and platforms bear the consequences. This paper examines verified pseudonymity as an institutional response to that problem. First, I model online communities as club-governed informational commons in which incivility degrades the shared environment and raises enforcement costs. The model shows that conduct can improve when sanctions attach to a persistent pseudonymous identity and when users have future access, reputation, or governance rights at stake. Second, I compare verified pseudonymity with open pseudonymity, real-name mandates, centralized know-your-customer verification, algorithmic moderation, and no intervention. Decentralized identifiers, verifiable credentials, proof of personhood, and non-transferable standing credentials matter because they can separate authentication from public identification. Third, I add a community-currency layer that separates access to scarce attention from governance rights. The result is a governance framework in which accountability depends less on public naming than on durable standing, credible sanctions, and reusable privacy-preserving credentials.

This paper investigates how much time people spend on retirement planning information available on the online portal of the largest pension fund for government and education sector employees in the Netherlands. The portal records the time participants devote to reviewing their retirement information on a daily basis and at the individual level. This dataset offers a fine-grained view of pension information use across all age groups. On average, participants spent only 833 seconds (about 14 minutes) in the portal during a 13-month observation period – highlighting that they make little use of it overall. When participants do devote significantly more time, it is only when concrete retirement options become available. We exploit a 2019 reform of the statutory retirement age in a Tobit regression discontinuity design to show, causally, that greater clarity about retirement timing substantially increases the time participants spend in the portal.

We build an experiment to uncover the bandit-like nature of consumer behavior usually masked by the product and time aggregation of consumption data. Subjects make repeated choices between musical styles (either all familiar or unfamiliar), and post-choice satisfaction is observed. We estimate Bayesian bandit models of learning taste by consuming with satiation. Our best model features decreasing random exploration, with openness being associated with higher exploration. Early exploration is more intensive in the unfamiliar treatment and persists throughout the experiment in both treatments. Overall, subjects make choices that deviate from their best prediction 61.5% of the time in the unfamiliar treatment versus 44.7% in the familiar treatment. Our model offers a rational interpretation of random utility in discrete consumer choices which does not rest on perception and/or decision errors.

We experimentally investigate preferences for clumping-versus-separating information in the gain and the loss domains, and also preferences for timing. Our design is motivated by the idea that information preferences may depend on reference points. Subjects participate in two monetary lotteries and choose how to receive the outcome information. For half of the subjects, the lotteries are framed as two gain lotteries; for the other half, as two loss lotteries. Based on Thaler (1985) one can expect that people want to learn the outcomes of the gain lotteries separately and the outcomes of the loss lotteries clumped together (cf. hedonic editing hypothesis). On the other hand, a different reference dependent model by Koszegi and Rabin (2009) relies on expectations-based reference points, and predicts that subjects should prefer clumped information irrespective of the frame.

The results of our experiment show a preference for separating information about gains, and no preference for clumping or separating information about losses. Regarding timing, we find a weak overall preference for receiving information sooner. These findings provide new insights into information preferences. We conclude by discussing policy implications, as well as our additional contributions to related literature.

This paper provides a new theoretical framework and a criterion to model the choice between democratic, hybrid, and epistocratic modes of political governance. From a normative perspective, we claim that the specificity of information should guide the choice between these modes of political governance because of its impact on costs of political governance. Any issue has a degree of information specificity that determines costs of political governance, which are combined in a Social Costs Function. Therefore, the model helps to assess the relative efficiency between democratic, hybrid, and epistocratic decision-making procedures to reach collective choices. The last section proposes extensions of the model by discussing how political, cultural, and epistemic institutions as well as polycentric governance modify the parameters of the model.

This research explores the factors that influence the adoption of barrel-aging techniques by US-based craft brewers from 2008 to 2014. Particular focus is placed on the importance of influence from geographically close peer breweries as a way to understand the effects of local influence or knowledge spillovers from agglomeration. Combining data on brewery-level production and estimates of the timing of the release of barrel-aged (BA) beers, I find evidence that nearby releases of BA beers increase the likelihood of a brewery introducing its first BA beer. However, national trends appear to be a stronger influence. These effects are robust to estimating on subsamples of brewery and metro sizes and controlling for a local demand proxy.

The prevalence of false and misleading news has become an issue of great concern in recent years. Academic researchers, policymakers, and social media firms all continue to seek effective solutions to reduce the sharing of misinformation. In this paper, we evaluate the effectiveness of two policies in particular: competition among media firms and fact-checking of published news articles by independent organizations. We first develop a theoretical model that predicts the effect of each policy and then conduct a behavioral experiment to test those predictions. Our experimental findings indicate that media competition is most effective at nipping misinformation in the bud because media firms spend significantly more resources on improving the accuracy of their news when readers obtain news from multiple sources. We also find that fact-checking improves the overall quality of news available to viewers; however, it does not incentivize firms to improve the accuracy of their own news articles. Last, our results from an interaction treatment suggest that under competition, fact-checking adversely affects firms’ investment in news accuracy.

In a game with costly information acquisition, the ability of one player to acquire information directly affects her opponent’s incentives for gathering information. Rational inattention theory then posits the opponent’s information-acquisition strategy is a direct function of these incentives. This paper argues that people are cognitively limited in predicting their opponent’s level of information, and hence lack the strategic sophistication that the theory requires. In an experiment involving a real-effort attention task and a simple two-player trading game, I study the ability of subjects to (1) anticipate the information acquisition of opponents in this strategic game, and (2) best respond to this information acquisition when acquiring their own costly information. I study this by exogenously manipulating the difficulty of the attention task for both the player and their opponent. Predictions of behavior are generated by a novel theoretical model in which Level-K agents can acquire information à la rational inattention. I find an out-sized lack of strategic sophistication, driven largely by the cognitive difficulties of predicting opponent information. These results suggest a necessary integration of the theories of rational inattention and costly sophistication in strategic settings.

Using a model, we explain why propaganda in autocracies can be blatantly false and unconvincing. We model two news outlets that report on a hidden state of the world, motivated by the ex-post beliefs of the audience about the state of the world. News outlets face a tradeoff when making egregiously false statements. On the one hand, such statements are easily verifiable as false. On the other hand, a demonstrably false report reduces the credibility of the report made by the competing outlet. This is especially true for audiences in autocracies that are characterized by high media cynicism and are prone to making sweeping generalizations about the self-serving nature of all media.

Air pollution remains a major challenge, especially in developing countries, requiring joint efforts from governments and society. This study examines how mass media, through its emotional tone, functions as an informal regulator of air pollution in China’s “war on air pollution”. Using daily data on media sentiment, air quality and related variables across Chinese cities, we find that negative emotional tones in environmental news are significantly associated with lower pollution levels. We identify mechanisms through which media influence public awareness, trigger government responses and pressure firms to reduce emissions. Our findings highlight the media’s role beyond information dissemination to shape agendas and social norms, even in contexts with restricted press freedom. This study offers new insights into how emotional framing in mass media contributes to environmental governance in developing countries.

This paper analyzes individual behavior in multi-armed bandit problems. We use a between-subjects experiment to implement four bandit problems that vary based on the horizon (indefinite or finite) and the number of bandit arms (two or three). We analyze commonly suggested strategies and find that an overwhelming majority of subjects are best fit by either a probabilistic “win-stay lose-shift” strategy or reinforcement learning. However, we show that subjects violate the assumptions of the probabilistic win-stay lose-shift strategy as switching depends on more than the previous outcome. We design two new “biased” strategies that adapt either reinforcement learning or myopic quantal response by incorporating a bias toward choosing the previous arm. We find that a majority of subjects are best fit by one of these two strategies but also find heterogeneity in subjects’ best-fitting strategies. We show that the performance of our biased strategies is robust to adapting popular strategies from other literatures (e.g., EWA and I-SAW) and using different selection criteria. Additionally, we find that our biased strategies best fit a majority of subjects when analyzing a new treatment with a new set of subjects.

This study presents a comparative evaluation of sentiment analysis models applied to a large corpus of expert wine reviews from Wine Spectator, with the goal of classifying reviews into binary sentiment categories based on expert ratings. We assess six models: logistic regression, XGBoost, LSTM, BERT, the interpretable Attention-based Multiple Instance Classification (AMIC) model, and the generative language model LLAMA 3.1, highlighting their differences in accuracy, interpretability, and computational efficiency. While LLAMA 3.1 achieves the highest accuracy, its marginal improvement over AMIC and BERT comes at a significantly higher computational cost. Notably, AMIC matches the performance of pretrained large language models while offering superior interpretability, making it particularly effective for domain-specific tasks such as wine sentiment analysis. Through qualitative analysis of sentiment-bearing words, we demonstrate AMIC’s ability to uncover nuanced, context-dependent language patterns unique to wine reviews. These findings challenge the assumption of generative models’ universal superiority and underscore the importance of aligning model selection with domain-specific requirements, especially in applications where transparency and linguistic nuance are critical.

We present a simple and robustly incentive-compatible price list methodology to elicit quantiles of a subjective real-valued belief. These elicited quantiles can be employed to approximate a subject’s complete subjective distribution, and we establish that the distribution maximizing entropy while adhering to the elicited quantiles is piecewise linear. Using this approach, our methodology extends to estimating arbitrary unobserved attributes of the subjective distribution, such as mean and variance, which are otherwise challenging to elicit. We provide a proof-of-concept for our framework through an experiment involving the elicitation of participants’ beliefs regarding the mathematical abilities of their peers.

Demographic change is one of Germany’s most pressing social and economic challenges. Using data from a representative telephone survey, we analyze how well informed respondents are about the magnitude of demographic change and what factors influence the accuracy of their beliefs. We find that respondents tend to overestimate the old-age dependency ratio when considering the current and long-term demographic situation separately. However, their beliefs regarding the change of the old-age dependency ratios over the considered period are not far from the projected change. A better understanding of the German statutory pension insurance plays an important role for more accurate beliefs.

We report the results of an experiment on selective exposure to information. A decision maker interested in learning about an uncertain state of the world can acquire information from one of two sources that have opposite biases: when informed on the state, they report it truthfully; when uninformed, they report their favorite state. A Bayesian decision-maker is better off seeking confirmatory information unless the source biased against the prior is sufficiently more reliable. In line with the theory, subjects are more likely to seek confirmatory information when sources are symmetrically reliable. On the other hand, when sources are asymmetrically reliable, subjects are more likely to consult the more reliable source even when prior beliefs are strongly unbalanced and this source is less informative. Our experiment suggests that base rate neglect and simple heuristics (e.g., listen to the most reliable source) are important drivers of the endogenous acquisition of information.

In this paper, we adopt an evolutionary model to describe the coevolution of technological transition and pollution in a country, where the choice of technology does not only give firms access to cleaner (but more expensive) or dirtier (cheaper and illegal) forms of production, but also access to social groups and information. Firms’ activity may be harmful to the environment and, due to the existence of ambient pollution charges, economic activity is affected by the level of pollution in the country. Our analysis describes how the evolution of the transition to clean technology and pollution generates a rich set of possible equilibria, which include stable pure strategies (where all firms choose the same technology) and inner equilibria (where both technologies could be adopted in the long run). We also observe more complex behavior and coexistence of different attractors as well as highlight the importance of initial conditions and uncover how the regulator may face possible pollution traps.

In the presence of a default option, the optimal search rule for an agent with a reference-dependent utility and a search cost predicts: (i) the default increases the reservation utility due to the reference effect, leading to a better choice, and (ii) those with higher reservation utility will self-select into search and are more likely to find a superior option. Our experiments document the presence of both effects. Those who reject the default are likely to find higher-ranked options in their active search, supporting the self-selection effect. Even when the self-selection channel is shut down, the reference effect remains.

We compare different forms of communication in the context of cheap talk sender-receiver games. While previous experiments find evidence supporting the comparative statics prediction that more preference divergence leads to less information transmission, there is also a consistent pattern of overcommunication and exaggeration, not predicted by theory, in which subjects convey more information than predicted in equilibrium. The latter of these findings may be due to the restricted nature of the message space in most experimental cheap talk games, encouraging subjects to engage in exaggeration artificially, rather than allowing it to emerge naturally. We tested this hypothesis with an incentivized lab experiment, and found evidence both phenomena persist with natural language (text-based) communication. Moreover, we probe the consequences of this expanded message space for outcomes, showing that senders benefit more than receivers, but that the most notable effect is that text messages improve efficiency.

Social scientists are paying attention to the role that knowledge plays in economic phenomena. This focus on knowledge has led to exploring two challenges: first, its governance to reap positive externalities and solve social dilemmas, and second, how we can craft institutions to match the intangible nature of ideas with adequate property rules. This article contributes by elaborating on the different knowledge property regimes and the elements contributing to their classification. This paper first taxonomises knowledge governance regimes based on Ostrom’s work on institutional analysis. Second, it examines why governance structures for managing knowledge production vary across industries, according to (1) the characteristics of knowledge, (2) the attributes of the organisations, and (3) the different rules-in-use to enforce property rights. This is the first study at the intersection of institutional analysis and political economy that highlights the knowledge features, incentive structures, and mechanisms undergirding knowledge governance in different property regimes.