No full-text available
To read the full-text of this research,
you can request a copy directly from the author.
A Simple Model of Herd Behaviour
Abstract
The author analyzes a sequential decision model in which each decisionmaker looks at the decisions made by previous decisionmakers in taking her own decision. This is rational for her because these other decisionmakers may have some information that is important for her. The author then shows that the decision rules that are chosen by optimizing individuals will be characterized by herd behavior; i.e., people will be doing what others are doing rather than using their information. The author then shows that the resulting equilibrium is inefficient. Copyright 1992, the President and Fellows of Harvard College and the Massachusetts Institute of Technology.
... The author in [1] asserts that a centrality measure is useful in quantifying influence in a social network only if it arises from the underlying social process. While the opinion dynamics models proposed in the literature [2]- [5] capture various aspects on social interactions, the Friedkin-Johnsen (FJ) model [4] is widely popular due to its analytical tractability and proven empirical validity. In the FJ framework, the agents are heterogeneous with varying degrees of stubbornness towards their internal biases. ...
... Several works on social power [8]- [10] demonstrate that an increase in the stubborn behaviour of an agent results in the agent achieving a higher social power in the group. However, the stubborn behaviour of an agent is often its inherent property, which may not 1 Aashi Shrinate is a research scholar, 2 Aravind Seshadri is an undergraduate student and 3 Twinkle Tripathy is an Assistant Professor in the Control and Automation specialization of the Department of Electrical Engineering, Indian Institute of Technology Kanpur, Kanpur, Uttar Pradesh, India, 208016. Email: aashis21@iitk.ac.in , aravinds21@iitk.ac.in and ttripathy@iitk.ac.in. ...
... (1), the final opinions satisfy x * = (I n − β)W x * + βx(0). Additionally, it follows from eqn. (2) and (3) that average final opinion denoted byx satisfiesx = 1 T n x * /n = x(0) T c. We use the following equations to construct a signal flow graph G s = (V s , E s ) [20]. ...
In this paper, we consider two stubborn agents who compete for `influence' over a strongly connected group of agents. This framework represents real-world contests, such as competition among firms, two-party elections, and sports rivalries, among others. Considering stubbornness of agents to be an immutable property, we utilise the network topology alone to increase the influence of a preferred stubborn agent. We demonstrate this on a special class of strongly connected networks by identifying the supporters of each of the stubborn agents in such networks. Thereafter, we present sufficient conditions under which a network perturbation always increases the influence of the preferred stubborn agent. A key advantage of the proposed topology-based conditions is that they hold independent of the edge weights in the network. Most importantly, we assert that there exists a sequence of perturbations that can make the lesser influential stubborn agent more influential. Finally, we demonstrate our results over the Sampson's Monastery dataset.
... That is, a herding player chooses an action, which has been chosen by the majority of the players before its turn. Clearly, the latter behavior is common, as we often observe individuals choosing a hyped restaurant, buying a rumoured (or well-discussed) product, or participating in a viral trend, etc. (see, e.g., [6,15]). ...
... Recall that the proportion is simply the fraction of players who choose action 1 and its updates can be re-written in the form of a SA scheme as shown in (6). It is well known in SA-based literature ( [9,14]) that the SA-scheme (and thus, -iterates) typically converges to some specific 'zeroes' (say ∞ ∈ [0, 1]) of the conditional expectation of the function (·), see (6): ...
... This can continue forever, with the dynamics crossing ∞ continuously. It is not difficult to guess that the overshoots/undershoots around ∞ decrease, as time ( ) progresses because 1/ decreases to zero (see (6)). In view of these observations, one can anticipate that any ∞ satisfying, ...
In classical game theory, the players are assumed to be rational and intelligent, which is often contradictory to reality. We consider more realistic behavioral game dynamics where the players choose actions in a turn-by-turn manner and exhibit two prominent behavioral traits --- α-fraction of them are myopic who strategically choose optimal actions against the empirical distribution of the previous plays, while the rest exhibit herding behavior by choosing the most popular action till then. The utilities are realised for all, at the end of the game, and each player gets to play only once. Our analysis focuses on scenarios when players encounter two possible choices, common in applications like participation games (e.g., crowd-sourcing) or minority games.
To begin with, we derive the almost sure mean-field limits of such dynamics. The proof is constructive and progressively narrows down the potential limit set and finally establishes the existence of a unique limit for almost all sample paths. We argue that the dynamics at the limit is captured by a differential inclusion (and not the usual ordinary differential equation) due to the discontinuities arising from the switching behavioral choices. It is noteworthy that our methodology can be easily modified to analyse the avoid-the-crowd behavior, in place of herding behavior.
We conclude with two interesting examples, named participation game and routing game, which encapsulate several real-life scenarios. Interestingly, for the first game, we observe that the game designer can induce a higher level of participation in an activity with smaller reward, by leveraging upon the presence of herding players.
... For example, when someone considers purchasing a product, they weigh both the features of the product and the reviews of other customers [2]. This process can be modeled within the framework of sequential learning to understand how individual decisions evolve over time based on the actions of their predecessors [3,4,5]. ...
... The first question examines whether, as observations accumulate in the long run, individuals collectively make the correct decision. The seminal works of Banerjee [3], Bikhchandani et al. [4] show that in herding and information cascade models, agents can ignore their private signals and imitate their predecessors, leading to inefficient learning where incorrect beliefs persist indefinitely. Subsequent research explores the conditions under which asymptotic learning can be guaranteed, where individuals eventually converge to the correct belief [6,5]. ...
... First, our research is closely related to sequential learning, particularly in understanding whether asymptotic learning will eventually occur and how fast it happens when it does. The first question has been a central topic since Banerjee [3] and Bikhchandani et al. [4] introduced the concepts of information cascades and herd behavior. They found that when private signals are binary, a cascade occurs with probability one, preventing further learning. ...
In a differentially private sequential learning setting, agents introduce endogenous noise into their actions to maintain privacy. Applying this to a standard sequential learning model leads to different outcomes for continuous vs. binary signals. For continuous signals with a nonzero privacy budget, we introduce a novel smoothed randomized response mechanism that adapts noise based on distance to a threshold, unlike traditional randomized response, which applies uniform noise. This enables agents' actions to better reflect both private signals and observed history, accelerating asymptotic learning speed to , compared to in the non-private regime where privacy budget is infinite. Moreover, in the non-private setting, the expected stopping time for the first correct decision and the number of incorrect actions diverge, meaning early agents may make mistakes for an unreasonably long period. In contrast, under a finite privacy budget , both remain finite, highlighting a stark contrast between private and non-private learning. Learning with continuous signals in the private regime is more efficient, as smooth randomized response enhances the log-likelihood ratio over time, improving information aggregation. Conversely, for binary signals, differential privacy noise hinders learning, as agents tend to use a constant randomized response strategy before an information cascade forms, reducing action informativeness and hampering the overall process.
... Pioneering works in economic theory have provided theoretical foundations for sequential social learning in which agents act in a predetermined sequence [7], [8]. The classical model demonstrates the phenomenon of information cascadesabsorbing states in which agents' beliefs about the state of the world become so strong that they overpower all private information and social learning halts. ...
... H 1 = (b 1 ). As in the classic model of [7] and [8], (b i ) i∈N is a Markov process. ...
We introduce a model of sequential social learning in which a planner may pay a cost to adjust the private signal precision of some agents. This framework presents a new optimization problem for social learning that sheds light on practical policy questions, such as how the socially optimal level of ad personalization changes according to current beliefs or how a biased planner might derail social learning. We then characterize the optimal policies of an altruistic planner who maximizes social welfare and a biased planner who seeks to induce a specific action. Even for a planner who has equivalent knowledge to an individual, cannot lie or cherry-pick information, and is fully observable, we demonstrate that it can dramatically influence social welfare in both positive and negative directions. An important area for future exploration is how one might prevent these latter outcomes to protect against the manipulation of social learning.
... Herd behaviour is the tendency of individuals to imitate the behaviour of a larger group or community, often instead of their own individual knowledge or instincts (Banerjee, 1992). ...
... In their studies, Bikhchandani, Hirshleifer, and Welch (1992) explain this situation based on the preferences of individuals within the framework of an economic model. In addition, Banerjee (1992) explains how the herd effect occurs in economic decisions. In their study, Campbell and Kyle (1993) explain how individual preferences are shaped in financial markets, how herd behaviour emerges, and how investor decisions are affected. ...
This study explores the impact of herd behaviour and peer influence on restaurant choices among Kassel University students. Through a controlled experiment involving 51 students, this research divides participants into two groups: a control group that makes restaurant choices independently and a treatment group that receives additional information such as Google review scores and peer preferences. The results, analysed using the Z test for proportions, show a significant influence of peer recommendations on the treatment group's restaurant choices compared to the control group. While the study is limited by a small sample size and potential self-reporting biases, it highlights the importance of social factors in consumer behaviour. These findings are mainly related to understanding the dynamics of decision-making processes among university students and could inform future research and marketing strategies in the food industry.
... This enables a unique environment where the information about the quantity available for rare offerings may gear toward a different direction than conventionally expected. There are instances where scarcity does not always produce perceptions of rarity or exclusivity but communality instead [62], because of the social cues scarcity may generate based on the economic theories of bandwagon effects and herd behavior [63]. For example, if a special edition produces a rare pair of virtual sneakers, those consumers who take note of the rarity of the offering may feel a unique camaraderie with others who could own the same pair of virtual sneakers. ...
... Perhaps the most promising aspect for future research is derived from the lack of insight into process measures to investigate the motivation behind the positive impact of the need for uniqueness. In order to delve deeper into the process of why the need for uniqueness became a prominent player in this context, future researchers may wish to 13 Human Behavior and Emerging Technologies examine variables that investigate social and emotional aspects of metaverse retailing, particularly pertaining to the need to belong to a unique community, yet juggling the equally influential concern of wishing to stand out from the majority of the population, that is, studying the social aspects of scarcity, whether from the counterconformity motivation [72], seeking novelty [57] or variety [101], or automatically seeking herd behavior for comfort [63]. Specifically, it will be helpful to collect emotional measures such as anger [43], aggression [52], or sense of fairness [102] in order to investigate further the process within which the need for uniqueness operates. ...
This research explores consumer attitudes and behavior in a metaverse retailing environment, mainly focusing on how perceptions of scarcity and rarity influence consumers' views of purchasing virtual wearables. Our findings diverge from preconceived notions about scarcity in physical/online retail, opening the door to a new understanding of how metaverse citizens may perceive scarcity of products. While it may appear simple to assume that physical-world strategies can seemingly be exported to virtual worlds, we uncovered a more complex story. The influence of the supply (availability) information on consumer attitudes in the metaverse is mediated by consumers' need for uniqueness. Specifically, seeing the virtual offerings as relatively abundant increased consumers' need for uniqueness, which improved the likelihood of purchase, a puzzling result. The mystery is better understood when considering how all items in exclusive collections in the metaverse can preserve their rare status, thereby fully separating scarcity and rarity. Unlike in physical retail environments, our findings indicate an interaction: high product availability (low scarcity) increases the likelihood of purchasing only when product rarity is high. These surprising results provide novel insights for academics and practitioners to consider the combinatorial effects of availability information and product rarity, as well as the virtual customers' characteristics, particularly their need for uniqueness as a mediator to their attitudes toward virtual products. Summary • Metaverse retail dynamics differ significantly from physical retail. • Product abundance in the metaverse increases con-sumers' need for uniqueness. • Need for uniqueness has a positive impact on meta-verse consumer attitudes. • Scarcity cues alone seem not to factor in virtual con-sumers' attitudes. • Rarity can be separated from scarcity (availability) perceptions. • Availability increases purchase likelihood only for highly rare offerings.
... Due to the easy access and widespread dissemination of public information such as macroeconomic data or policy changes, investors often make decisions based on the collective response to this information, overlooking any unique insights they may have. Banerjee shows that upon observing the actions of others, an individual may elect to follow suit even if their private information suggests a divergent course of action, under the presumption that the former possess more accurate information, thereby engendering herd behavior [2]. Furthermore, the perspective of behavioral homogeneity further uncovers the consistency aspect within the herd effect, signifying the high degree of uniformity in investors' decision-making and trading activities. ...
The monetary policies of the Federal Reserve have a bigger effect on the financial markets of many countries because the decision to lower interest rates often starts a chain reaction. This is particularly true when the Fed decides to decrease interest rates, as this typically sets off a chain reaction. On September 18, 2024, the Federal Reserve announced an interest rate cut, immediately drawing widespread attention. Therefore, the paper aims to explore the impact mechanism of the Feds interest rate cut policy on the herd behavior in the A-share market, and analyze how it affects the behavior and psychological drivers of investors therein. It adopts the theoretical framework of behavioral finance to explore the changes in the decision-making patterns and herd behavior of investors in the A-share market after the announcement of the interest rate cut by analyzing relevant literature and market transaction data. The results indicate that the Feds interest rate cut policy prompts investors to exhibit typical herd behavior by influencing market sentiment and expectations, thus exacerbating the herd effect in the A-share market. This behavioral pattern increases market volatility in the short term, leading to irrational fluctuations in asset prices, which in turn increases market risk. In addition, the herd behavior may affect the effective discovery of market prices, resource allocation and exacerbate the instability of financial markets.
... The relationship between network usage and public sentiment about air pollution on social media will be further discussed here since our previous research has already studied other socioeconomic factors such as population, economic conditions, and education level (Ye et al. 2022). Herd behavior, describing the phenomenon where a user discounts their own information initially formed and imitates others after observing predecessors' behavior (Banerjee 1992;Sun 2013), was found in social media. For instance, the observation of a "like" from a strongly tied predecessor may lead to a view-through intention on Facebook (Mattke et al. 2020). ...
Public response (including sentiment and sensitivity) to air pollution, one of the most serious health threats in recent years, can be monitored in real-time across China through social media channels. However, few studies have long-term investigations into both public sentiment and sensitivity towards different air pollutants while considering socioeconomic factors, resulting in an incomplete understanding of public reactions to air pollution and less effective policy measures. In this paper, we employed sentiment analysis to classify Weibos with positive and negative sentiments, then explored the relationship between the concentrations of six air pollutants (PM2.5, PM10, CO, NO2, O3, and SO2) and Weibos related to air pollution during 2017–2021 across China. The results show that residents in China exhibit the greatest sensitivity and express the most negative sentiments toward PM2.5, both at the national level and for individual provinces. After filtering out positive Weibos by sentiment analysis, there would be a stronger relationship between the number of negative Weibos reflecting public sentiment about air pollution and PM2.5 concentrations. A threshold effect has been identified where public reaction plateaus or wanes at high pollution levels. Socioeconomic factors, including education level, economic conditions, and network usage, are found to influence public sentiment towards air pollution. This study highlights the critical role of targeted policy interventions and the application of sentiment analysis in effectively understanding and addressing public concerns about air pollution, particularly PM2.5, which is essential for enhancing environmental health strategies across China.
... Such settings have often been modeled formally via the following general framework (Banerjee 1992;Bikhchandani, Hirshleifer, and Welch 1992;Acemoglu et al. 2008 there is a set of options being considered; a sequence of agents must decide, one by one, which option to accept; and each agent receives a signal from the environment that provides probabilistic information about which option is better. We imagine that there is a "correct" or "best" option, and each agent receives a payoff if they choose this best option. ...
A rich line of theoretical work has modeled scenarios in which a set of agents make decisions sequentially, based on observing a growing mix of public and private signals that are revealed as these decisions occur. Here, we study a second crucial dimension, which is the way in which strategies can depend on crowding. In particular, consider a setting in which agents must sequentially decide which of several options to invest in, each based on a public signal that they receive. One of these options will ultimately be revealed to be valuable; but crucially, all the agents who selected this option must divide the value that comes from it. As a result, when a given agent j goes to make a decision among the options, the decisions of earlier agents convey information about the payoff that j will receive in any eventual division of the value. When many earlier agents have chosen a specific option, the greater crowding on this option means it must be divided more finely, resulting in lower payoffs. To simulate large games when signals are public, we define a polynomial-time algorithm to compute equilibrium strategies. We show that even in this case of public signals, the interaction of crowding with informational effects leads to complex non-monotonicities in the resulting sequential decisions, with agents sometimes choosing options with lower expected levels of crowding --- and hence a better split of the potential value --- over options with better informational or current crowding properties.
... Herding behavior refers to the tendency of individuals to align with the majority of decision-makers in a given environment, rather than relying on their own information when making decisions (Banerjee, 1992). Kumar and Goyal (2015) describe situations in which rational individuals begin to behave irrationally by imitating the judgments of others. ...
Purpose
This study investigates herd behavior in the Fan Tokens market, comparing it with the non-fungible tokens (NFTs) and traditional cryptocurrency markets.
Design/methodology/approach
This study investigates herding behavior by examining the relationship between the cross-sectional dispersion of asset returns and overall market returns, utilizing five distinct model specifications. To enhance the robustness of the findings, the regressions are re-estimated using the GARCH model, ensuring more reliable parameter estimates and capturing the impact of volatility on herding behavior.
Findings
The analysis reveals strong evidence of herd behavior in the Fan Token market, particularly during bearish conditions, heightened volatility, and low trading volume. Positive news was found to amplify volatility more than negative news. In contrast, no statistically significant herd behavior was identified in the NFT and traditional cryptocurrency markets, where investors showed a more cautious response to market conditions.
Practical implications
Understanding the unique dynamics of Fan Tokens can help investors, regulators, and market participants make informed decisions and develop strategies to mitigate risks associated with herd behavior and volatility in this rapidly evolving market.
Originality/value
This study highlights the unique characteristics of Fan Tokens, emphasizing their strong ties to fan sentiment and sports outcomes, as well as the role of uninformed investors in shaping market dynamics. The findings contribute to the literature on digital asset markets and investor psychology, offering novel insights into this emerging asset class.
... Such constructions are widely applied in decision-making, learning, and inference tasks for single or multi-agent systems in static environments. In multi-agent networks or social networks, the sequential process where agents infer the true state by observing the decisions of peers and signals generated by the environment, is referred to as (non-)Bayesian social learning [1]- [8]. These models are widely used in economics, political science, and sociology to model the behavior of financial markets, social groups, and social networks [9], [10]. ...
Hidden Markov Models (HMMs) provide a rigorous framework for inference in dynamic environments. In this work, we study the alpha-HMM algorithm motivated by the optimal online filtering formulation in settings where the true state evolves as a Markov chain with equal exit probabilities. We quantify the dynamics of the algorithm in stationary environments, revealing a trade-off between inference and adaptation, showing how key parameters and the quality of observations affect performance. Comprehensive theoretical analysis on the nonlinear dynamical system that governs the evolution of the log-belief ratio over time and numerical experiments demonstrate that the proposed approach effectively balances adaptation and inference performance.
... Social learning theory also offers valuable insights. According to this theory, individuals acquire social behaviors predominantly through observation and imitation (Banerjee, 1992). Childhood is an important stage in individual cognitive shaping and ideological development; therefore, children are more inclined to seek learning standards and experiential knowledge from communication with their parents. ...
Objective
This study explored the developmental pathways of smartphone dependence among rural children in China, focusing on the interplay between parental supervision, children’s self-control, and parent–child relationships.
Methods
In-depth interviews were conducted with 20 rural Chinese children and their parents. A Qualitative Comparative Analysis was employed to examine the conditions and mechanisms underlying smartphone dependence from both children’s and parents’ perspectives.
Results
Three distinct pathways to smartphone dependence were identified. Path 1: Children who were not left behind exhibited low self-control, lacked supervision and guidance, and had introverted personalities, and were more susceptible to smartphone dependence. Paths 2 and 3— Children who were left behind in rural areas and lack self-control were prone to developing smartphone dependence, regardless of whether they had a distant or harmonious parent–child relationship and an introverted or extroverted personality.
Conclusion
Children’s self-control and parental supervision were critical factors influencing the participants’ smartphone dependence. The children’s sex, age, academic performance, parents’ smartphone use duration, and primary caregivers’ parenting skills moderated these influencing paths. Interventions should focus on enhancing children’s self-control through skill-building and equipping parents and primary caregivers with effective supervision, communication, and boundary-setting strategies to foster healthier technological habits.
... Again according to Banerjee (1992) when the average payoff decreases as more people choose it, herding tends to decrease. However, investors can expect increased returns by following the choices of others, which in turn can lead to a resurgence in herding behaviour. ...
Numerous stock market crashes and bubbles demonstrate that human decision-making in financial markets is not always rational. There exists a compelling influence on market investors' decision-making processes, which begs the question: what is it? This influence is none other than behavioural biases, currently one of the most captivating topics in economics and finance. These biases act as a supernova factor shaping market outcomes. Past research indicates that investors often deviate from rational decision-making when selecting investments, particularly during uncertain times or when under stress. They tend to follow the crowd without critically analysing information, resulting in a phenomenon known as herding bias. Our current research has undertaken an extensive literature review to explore how investors are affected by this herding tendency.
... Again according to Banerjee (1992) when the average payoff decreases as more people choose it, herding tends to decrease. However, investors can expect increased returns by following the choices of others, which in turn can lead to a resurgence in herding behaviour. ...
Numerous stock market crashes and bubbles demonstrate that human decision-making in financial markets is not always rational. There exists a compelling influence on market investors' decision-making processes, which begs the question: what is it? This influence is none other than behavioural biases, currently one of the most captivating topics in economics and finance. These biases act as a supernova factor shaping market outcomes. Past research indicates that investors often deviate from rational decision-making when selecting investments, particularly during uncertain times or when under stress. They tend to follow the crowd without critically analysing information, resulting in a phenomenon known as herding bias. Our current research has undertaken an extensive literature review to explore how investors are affected by this herding tendency.
... This herding phenomenon amplifies the tendency to conform, creating a cascade effect that fosters pledging decisions based on incomplete information processing. Such decisions often lack comprehensive analysis and critical evaluation, as individuals prioritize social validation over independent assessment (Banerjee, 1992). Thus, the following hypothesis is proposed: ...
Purpose
This study aims to examine the transformative impact of crowdfunding on entrepreneurial ventures, focusing on backers’ decision-making on reward-based platforms. It investigates why crowdfunding is popular among creators and backers, despite the inconsistent success rates of projects. The study empirically analyzes the factors influencing backers’ pledging decisions, particularly past crowdfunding experience, herding behavior and product innovativeness. By exploring these dimensions, the research aims to provide insights into backers’ decision-making dynamics in crowdfunding.
Design/methodology/approach
The study used logistic and ordered logistic regression models to assess the effects of key determinants, including prior crowdfunding experience, herding behavior and perceived product innovativeness. A total of 402 valid responses were collected from individuals with experience on a reward-based crowdfunding platform. The model specifications tested both the likelihood of changes in backers’ initial decisions and their final pledging intentions.
Findings
Backers’ crowdfunding experience significantly increases the likelihood of changing their initial decisions. The number of backers positively influences impulse purchases. However, product innovativeness does not have a statistically significant impact. Interestingly, the interaction between crowdfunding experience and time pressure – measured by the number of remaining days – reveals that backers are more likely to alter their decisions in the early stages of campaigns, contrary to expectations.
Originality/value
This research examines how crowdfunding transforms entrepreneurial ventures by analyzing backers’ decision-making on reward-based platforms. Despite its popularity, project success rates vary, highlighting the need to understand decisions made with incomplete information. Using a multi-step approach that includes literature review, campaign analysis and online surveys, the study explores factors influencing impulse purchases. Findings show that backers’ crowdfunding experience and campaign dynamics significantly affect decision changes, offering implications for campaign management. This study provides valuable insights into backer behavior and practical guidance for creators and platform operators to optimize crowdfunding campaigns.
... Social and public choice theory often employ equilibrium-oriented models (Goodman and Porter 2004;Roger B Myerson et al. 2013). Similarly, collective behavior models in economics, such as studies in opinion formation on networks or studies of 'herd' behavior, focus on mathematically tractable dynamics that converge to specific equilibria (Banerjee 1992;Golub and Jackson 2010). But equilibrium models naturally limit the scope of analysis because often they either do not specify explicit dynamic processes, which is critical when understanding how collective decisions unfold (Pangallo 2024), or, if they do, assumptions about the behavior of agents are purposefully so restrictive that the dynamic process exhibits an analytically tractable character. ...
Prevailing top-down systems in politics and economics struggle to keep pace with the pressing challenges of the 21st century, such as climate change, social inequality and conflict. Bottom-up democratisation and participatory approaches in politics and economics are increasingly seen as promising alternatives to confront and overcome these issues, often with utopian overtones, as proponents believe they may dramatically reshape political, social and ecological futures for the better and in contrast to contemporary authoritarian tendencies across various countries. Institutional specifics and the associated collective human behavior or culture remains little understood and debated, however. In this article, I propose a novel research agenda focusing on utopian democratisation efforts with formal and computational methods as well as with artificial intelligence - I call this agenda Artificial Utopia. Artificial Utopias provide safe testing grounds for new political ideas and economic policies in-silico with reduced risk of negative consequences as compared to testing ideas in real-world contexts. An increasing number of advanced simulation and intelligence methods, that aim at representing human cognition and collective decision-making in more realistic ways, could benefit this process. This includes agent-based modelling, reinforcement learning, large language models and more. I clarify what some of these simulation approaches can contribute to the study of Artificial Utopias with the help of two institutional examples: the citizen assembly and the democratic firm.
... When selecting virtual reality technology as an intervention for their child's ADHD, parents often turn to online communities for advice and recommendations to aid in their final decision-making. In the past, parents solely relied on doctors for treatment plans, which was due to inducements from medical market providers and information asymmetry between doctors and patients (Banerjee 1992;Raafat et al. 2009). With the increasing prevalence of online communities, parents are more frequently involved in community discussions to decide on treatment plans for children with ADHD. ...
Digital health interventions (DHI) using virtual reality (VR) technologies have been developed to treat attention deficit hyperactivity disorder (ADHD). While previous studies have mainly evaluated the feasibility of VR as an ADHD intervention, there is a dearth of research examining the decision-making psychology and influencing factors among parents of ADHD patients regarding the adoption of such emerging VR intervention techniques, which carry inherent risks. Building on the principles of Prospect Theory, this study highlights preference structures, belief characteristics, and community participation. The study selected 23 explanatory variables, including parents’ comprehension of VR treatment, level of trust, information sources, time and financial costs. A self-designed questionnaire was used to collect data on the willingness of parents of ADHD children to opt for VR treatment. By constructing a binary logistic regression model, we examine the preference structure, belief characteristics and decision readiness of parents of children with ADHD when choosing a virtual reality intervention policy. Parents’ choices of VR interventions for their children are complex. While parents consider the therapeutic benefits of VR, the time investment required for children’s treatment, and knowledge on VR interventions from online communities, their decisions are not always made objectively like an agent would. Instead, they frequently make choices based on a willingness to take risks, placing greater emphasis on relative rather than absolute values. Their decision-making is often swayed by online community information, resulting in choices that may not optimize benefits and sometimes disregarding financial and time costs related to their children’s health. Overall, parents of children with ADHD have demonstrated acceptance of the innovative VR intervention technique. Through examining the factors that impact preference selection, the implementation and promotion of VR intervention in ADHD treatments can be facilitated, thereby advancing the development of Digital Health Interventions (DHI). This can provide valuable insights for developing effective ADHD intervention strategies.
... This implies that in absence of a clear preference or knowledge, social proof is highly successful in persuading and guiding people's decision-making (Hugh et al. 2022). In the context of financial markets, this phenomenon, where actions of predecessors are intentionally imitated and individual research and own beliefs ignored, is often referred to as herding behavior (Banerjee 1992) which has been increasingly linked to finfluencers (e.g., Gupta and Goyal 2024;Anshori and Restuningdiah 2024). Katona's (1951) early work illustrates how individuals can learn from groups and from mechanical 'stamping in' of heuristics or simple rules of thumb, which trade deeper understanding of processes for trust and reliance on others for information. ...
Influencer marketing has become a prominent strategy for promoting utilitarian products and services. In the financial sector, the demand for financial literacy has fostered the emergence of a specialized group of financial influencers. These influencers disseminate content, share personal experiences, and offer recommendations on financial decision‐making via social media platforms. This research aims to evaluate when financial influencers recommend stocks and cryptocurrencies and to document the returns when individuals would have invested accordingly. The study utilizes a hand‐collected dataset with 453 recommendations by 21 different Dutch financial influencers, each with more than 1000 followers, pertaining to 243 different stocks and 61 different cryptocurrencies. The investigation is the first that focuses on pre‐recommendation returns and thereby considers the timing of the influencers' endorsements. Findings indicate that financial influencers tend to recommend financial assets that have demonstrated strong performance in the recent past. However, the study reveals that the returns on these recommended stocks and cryptocurrencies are typically negative post‐recommendation. The research highlights a trend where financial influencers' endorsements are driven by overreliance on social heuristics, thus suggesting a potential adverse impact for investors who act on these recommendations. This underscores the risks for investors following finfluencer advice, suggesting the need for caution and stricter regulatory oversight to ensure transparency and to protect the financial well‐being of consumers.
... Potential backers are also more inclined to support projects they perceive as having a high chance of success due to time and resource constraints. This tendency is often influenced by herding behavior, where people act based on what they believe others will do (Banerjee, 1992), even when making funding decisions. Research shows that investment rates increase as the perceived likelihood of a project's successful funding increases (Agrawal et al., 2010). ...
... From a social psychology standpoint, Banerjee (1992) postulated that the peer effect is a phenomenon whereby individuals in a group imitate and derive insights from the behaviour of their peers, thereby influencing their own decision-making processes. Amid a complex and evolving external economic environment, researchers have extended the concept of peer effect to the governance of micro-firms to determine the extent to which corporations emulate their peers' strategic actions. ...
... Theoretical work on sequential (or simultaneous) choice has shown that social influence can have dramatic consequences, as people might be led astray by the already established popularity of an option and eventually settle on options of inferior quality [1,2,4,6,12]. Large scale empirical studies on ranking and upvoting/downvoting interfaces [16,17] have also demonstrated that social influence can substantially alter the social dynamics and the long-term outcomes of online systems, influencing the evaluations and/or popularity of the curated content. ...
Theoretical work on sequential choice and large-scale experiments in online ranking and voting systems has demonstrated that social influence can have a drastic impact on social and technological systems. Yet, the effect of social influence on online rating systems remains understudied and the few existing contributions suggest that online ratings would self-correct given enough users. Here, we propose a new framework for studying the effect of social influence on online ratings. We start from the assumption that people are influenced linearly by the observed average rating, but postulate that their propensity to be influenced varies. When the weight people assign to the observed average depends only on their own latent rating, the resulting system is linear, but the long-term rating may substantially deviate from the true mean rating. When the weight people put on the observed average depends on both their own latent rating and the observed average rating, the resulting system is non-linear, and may support multiple equilibria, suggesting that ratings might be path-dependent and deviations dramatic. Our results highlight potential limitations in crowdsourced information aggregation and can inform the design of more robust online rating systems.
The stock exchange serves as a critical source of corporate financing and as a platform for individuals to invest their savin gs, attracting a substantial amount of domestic capital in recent years and playing a pivotal role in the country's economic growth and development. This study investigates herding behavior in the Tehran Stock Exchange (TSE) under various economic and social conditions, including peri ods before and after exchange rate fluctuations, prior to and following the COVID-19 pandemic, and during both bullish and bearish market phases from April 2015 to March 2023. The analysis is based on the overall stock price index and the price index of the top 50 companies, employing Ordinary Least Squares (OLS) regression and quantile regression methodologies. The findings indicate that herding behavior is significantly evident in the TSE throughout the entire study period and across most quantiles (from the 0.05 to the 0.75 quantile) for both indices. Specifically, herding behavior is prominently observed prior to significant exchange rate fluctuations; however, it becomes unconfirmed afterward due to the insignificance of the resulting coefficients. Furthermore, herding behavior was found to be present in the TSE before the COVID-19 outbreak, with a notable decline after the pandemic, and in some cases, a reversal was observed. The analyses also demonstrate that herding behavior persists in both bullish and bearish market conditions, especially within the lower market quantiles.
The results reveal significant evidence of herding in the overall, bearish, and extended crisis market phases during extreme downturns, while the magnitude of market returns in the tail distribution is considered. Asymmetric herding behavior is more pronounced and prevalent, conditioned by market dimensions like return direction, trading volume, and volatility, with CSSD proving more effective than CSAD in detecting asymmetric patterns. Notably, herding strongly appears in the COVID-19 market during periods of abnormally high market volatility, reflecting heightened market sentiment. Applying Dow Theory to delineate bull and bear market phases significantly improved the methodological complexity and analytical depth related to herding behavior. These findings suggest policy implications for regulators and market participants in minimizing herding effects to create an efficient market environment through enhanced market surveillance, improved investor education, and the use of advanced technologies.
Purpose – This paper aims to elucidate the factors influencing individuals’ cryptocurrency adoption within the distinctive socio-economic context of a developing country. It argues that encouraging financial innovation and understanding cryptocurrency adoption is crucial for businesses, regulators and researchers.
Design/methodology/approach – It investigates factors affecting cryptocurrency adoption by integrating the Theory of Planned Behavior (TPB), Innovation Diffusion Theory (IDT) and the Trust-Risk Framework. It employed Structural Equation Modeling (SEM) on data collected from 370 respondents.
Findings – Results indicate that attitude significantly influences intention to use cryptocurrency, with trust being the most influential factor shaping attitude. Relative advantage and compatibility positively influence attitudes toward cryptocurrency, whereas complexity has a negative effect. Observability was not a significant predictor of attitude. Interestingly, perceived risk positively affects trust in cryptocurrencies.
Research limitations/implications – Findings reflecting the context of Türkiye may not be generalizable to countries with different cultural or legal contexts. The high education level of participants could introduce bias due to their likely higher technology literacy.
Practical implications – The findings inform various stakeholders about cryptocurrency adoption in emerging countries.
Social implications – Understanding dynamics influencing attitudes and intentions contributes to socio-economic discussions on integrating cryptocurrencies into mainstream financial systems.
Originality/value – This research contributes to theory and practice by providing empirical insights into the dynamics of cryptocurrency adoption in emerging countries. By synthesizing established theories, this study advances understanding and enriches comprehension of the factors influencing adoption in emerging economies.
This paper investigates the potential influence of herd behavior on companies' capital structure choices. While existing capital structure theories often overlook psychological and social factors, this study explores the role of herding, an area under-examined in the literature. The analysis is conducted on publicly listed companies across multiple developed European markets, including the United Kingdom, France, Germany, and Switzerland, over the period from 2000 to 2023. Using the cross-sectional absolute deviation (CSAD) methodology, the results reveal that herd behavior is more pronounced in specific countries and sectors. The United Kingdom exhibits the most prominent herding, particularly in materials, industrials, consumer discretionary, communication services, and information technology sectors, often centered around industry mean capital structures. Switzerland shows herding in health care and industrials, possibly driven by demographic and technological shifts, while France sees herding in the health care sector. Sectors like energy and consumer staples demonstrate resistance to herding, likely due to their need for stable, individualized capital structures. Additionally, the degree of financial leverage influences herding, with firms in sectors like consumer discretionary and communication services aligning with industry benchmarks to avoid negative market perceptions. These findings enhance understanding of behavioral factors impacting capital structure decisions and offer implications for both academic research and practical policy.
In financial markets, the expression "herding behaviour" portrays investors' affinity to mimic the exercises of others, which habitually brings about them venturing into risky investment decisions without cautiously gauging the important risk-reward ratio or relevant information. This phenomenon can make financial frameworks more unsound, enhance market volatility, and make the general economy more volatile. This review of the literature offers an exhaustive analysis of herding behaviour, underlining its monetary, social, and psychological causes as well as its effects on market outcomes. The review assesses the social components and psychological biases leading to herding affinities among institutional and individual investors by joining the empirical evidences from different studies. The models stowed away this idiosyncrasy are figured out in this review through the discussion of theoretical frameworks like informational cascades, social learning models, and behavioural finance theories. This review also attempts to unveil the unfavourable outcomes of herding behaviour, for instance, the distortion of market efficiency, the escalation of market volatility, and the formation of speculative bubbles. This article tries to give a comprehensive information on herding behaviour, its commitment in financial market disappointments, and reasonable methodology for mitigating its debilitating consequences on market efficiency through an analysis of empirical evidences and theoretical frameworks.
We investigate whether the herding behaviour of bond funds influences corporate bond issuance. Using quarterly data from 1998 to 2018, we find that only buy herding significantly increases bond issuance activity. Further analysis supports an informational channel, suggesting that ‘investigative herding’ among funds enhances credit information efficiency more significantly than demand‐driven price impacts. Our main identification strategy employs an instrumental variable approach based on the share of inexperienced managers among fund bondholders.
Social learning provides a fundamental framework in economics and social sciences for studying interactions among rational agents who observe each other's actions but lack direct access to individual beliefs. This paper investigates a specific social learning paradigm known as Word-of-Mouth (WoM), where a series of agents seeks to estimate the state of a dynamical system. The first agent receives noisy measurements of the state, while each subsequent agent relies solely on a degraded version of her predecessor's estimate. A defining feature of WoM is that the final agent's belief is publicly broadcast and adopted by all agents, in place of their own. We analyze this setting both theoretically and through numerical simulations, showing that some agents benefit from using the public belief broadcast by the last agent, while others suffer from performance deterioration.
In dieser Arbeit wurde untersucht, ob, und in welcher Form, Digital Nudging im E-Mail-Marketing von Non-Profit-Organisationen (NPO) in der Not- und Katastrophenhilfe (NKH) dabei unterstützen kann, Einmal- und Mehrfachspendende unter geringem Ressourcenaufwand in Dauerspendende zu konvertieren. Dazu wurden mithilfe der Design Science Research Methodologie und der Digital-Nudge-Design-Methode vier Artefakte entwickelt, die den vier Digital-Nudge-Mustern Anker, Voreinstellungs-Anker, Phantom-Köder und Identifikations-Kohärenz entsprechen.
This study examines electoral forecasting in volatile party systems, focusing on factors contributing to deviations between poll predictions and actual election outcomes. Using Italy as a case study, it identifies biases in polling data and proposes a method to enhance estimator accuracy in a context of stable institutions and volatile electoral dynamics. Data from three Italian general elections are analyzed to evaluate discrepancies between pre-electoral polls and results, assessing key factors such as timing of data collection, survey methodology, sample size, and party system fragmentation. Employing a Bayesian inference process via a Markov chain Monte Carlo (MCMC) adaptive Metropolis-Hastings (MH) algorithm, the study demonstrates that pre-electoral estimates can be significantly improved using the Two-Stage Model (TSM). By consistently outperforming traditional poll predictions, the TSM offers a robust framework for addressing polling biases. These findings advance political forecasting by improving accuracy in both consolidated democracies and volatile electoral contexts, while emphasizing the need for future research on dynamic polling methods and fundamentals-based models.
Project environments are increasingly characterized by uncertainty and, for this reason, require project managers to make frequent, quick decisions with poor and short-term information. In this context, project managers must shift from a rational and structured decision-making process to a simplified and less structured one leveraging heuristics. Heuristics are beliefs, experiences, mental models, frames, and other cognitive shortcuts for replacing parts of the rational and analytical process underpinning decisions. They are adequate support and very useful in many cases; nonetheless, they induce distortions in evaluations and estimates, with significant impacts, particularly in the context of projects. Heuristics are not free of the project manager’s influence (“bias”). The challenge for project managers in today’s uncertain project environments is to boost the use of heuristics while managing and controlling bias. Addressing and mitigating biases becomes crucial to improve decision-making effectiveness and outcomes. Artificial intelligence (AI) emerges, in such a context, as a potential support harbinger of tangible benefits for the project managers’ decisions.
This study explores the dynamics of customer loyalty in ride-hailing aggregation platforms, a burgeoning segment of the sharing economy that consolidates multiple service providers into a unified interface. Leveraging perceived value theory, we analyze the “double-edged sword” effect of same-side network externalities—how perceived customer network size impacts customer loyalty in a non-linear, inverted U-shaped pattern. Data from 543 aggregation platform users reveal that moderate levels of perceived customer network size optimize loyalty, balancing the positive effects of increased utility and trust with the negative impacts of network congestion. Moreover, perceived value mediates this curvilinear relationship, while perceived driver network size—a measure of cross-side network externalities—intensifies its curvature, highlighting the critical role of interdependencies between platform participants. Our arguments and results contribute to the literature on customer loyalty on platforms, and can inform the strategic decisions of platforms that rely on same side network externalities for value creation.
Purpose
This study examines how default bias, driven by the default positive review (DPR) rule – which automatically classifies reviews not provided by consumers within a specified period as positive – and rebate bias, associated with the conditional rebate strategy (CRS), where sellers offer rebates exclusively to consumers who submit positive reviews, distort the distribution of online product reviews over time and impact consumer satisfaction.
Design/methodology/approach
A key aspect of our method lies in developing latent variable models that capture the relationship between biased online reviews and consumer satisfaction levels. By applying our models to a panel dataset from Taobao – a leading Chinese e-commerce platform – and using insights from online consumer feedback surveys, we assess the extent of the biases introduced by DPR and CRS in a given feedback system. A hierarchical regression model was employed to investigate the impact of the proposed biases on consumer satisfaction.
Findings
Consumers who have previously written online reviews experience satisfaction outcomes 72.9% of the time with DPR and up to 81.3% when CRS is included. Implementing DPR may boost product sales to some extent, but it would significantly amplify consumer dissatisfaction, whereas offering a rebate could effectively alleviate consumer discontent, even though the rebate is conditional.
Originality/value
Our findings reveal the extent of biases introduced by CRS and DPR in online reviews and inform the consumer satisfaction debate regarding the phenomenon of excessive positive reviews resulting from these practices.
Given the global prevalence of COVID-19 affecting financial markets, including cryptocurrency markets, examining trader behavior during the pandemic is particularly intriguing. This research employs the Cross-Sectional Absolute Dispersion (CSAD) method and Ordinary Least Squares (OLS) technique to analyze herd behavior in cryptocurrency markets across five distinct periods: the overall period, pre-COVID-19, during the COVID-19 pandemic, bullish markets, and bearish markets. Daily price data for four major cryptocurrencies by market capitalization (Bitcoin, Ethereum, Tether, and Ripple) from January 2019 to December 2021 are used. The study results indicate strong herd behavior throughout the entire period, suggesting that investors tend to follow market trends during this time. Additionally, herd behavior is not observed in the pre-COVID-19 period, whereas the COVID-19 pandemic period is associated with significant herd behavior in cryptocurrency markets. The findings also reveal that herd behavior is significantly confirmed only in bullish markets, implying that traders in cryptocurrency markets during the COVID-19 pandemic and in bullish markets tend to make similar trading decisions.
Keywords
Herd Behavior COVID-19 Cryptocurrency Bullish Market Bearish Market
This study investigates the adaptive nature of herding behaviour in global energy markets, focusing on the USA, Europe and Asia, from January 2013 to December 2023. Utilizing advanced sentiment analysis tools, including Thomson Reuters MarketPsych Indices, Bloomberg Sentiment Analysis and the Hedonometer, this study captures real-time fluctuations in market mood. By employing Cross-Sectional Absolute Deviation models, static regression, rolling regression and quantile-on-quantile regression, this study examines the impact of investor sentiment, news sentiment and happiness on herding behaviour under varying market conditions. The findings revealed significant regional variations and underscored the dynamic and context-dependent aspects of herding behaviour. In the USA, higher levels of investor happiness and positive market sentiment are linked to increased herding behaviour, whereas negative news sentiment is linked to reduced herding behaviour. Similar but less pronounced patterns are observed in Europe, whereas Asian markets show weaker correlations, highlighting the critical role of market volatility and trading volume. These insights offer valuable implications for portfolio management, regulatory frameworks and investment strategies, thus enhancing the understanding of herding behaviour in volatile and sentiment-driven markets.
There are often benefits to consumers and to firms from standardization of a product. We examine whether these standardization benefits can "trap" an industry in an obsolete or inferior standard when there is a better alternative available. With complete information and identical preferences among firms the answer is no; but when information is incomplete this "excess inertia" can occur. We also discuss the extent to which the problem can be overcome by communication.
The view that individual choice is fundamental to the analysis of modern reproductive behavior has been largely unchallenged. This article proposes that certain characteristics of demographic change in Western Europe between 1870 and 1960 suggest a significant role for "others"--kin, friends, and neighbors--in accounting for demographic behavior. Moreover, demographic as well as linguistic patterns suggest that while in the past the relevant "others" were members of local community, in the present the relevant community is largely national.
An economic theory of the process of diffusion of innovations is developed and illustrated. In the theory, adoption is determined
by comparative advantage considerations. An innovation is first adopted by skilled and experimenting entrepreneurs and then
“diffuses” down the skills scale. If the innovation affects supply substantially, prices may decline, profits eliminated,
and early, skilled (and high labor opportunity cost) producers may exit from the affected line of production—hence, an “innovation
cycle.” The theory implies that technological change is affected by the distribution as well as by the average level of skills.
This paper demonstrates that the existence of public opinion polls amplifies the effects of shifts in the distribution of the public's preferences over the issue space on the election's results. Voters evaluate candidates by their positions on the issue space and by a valence or general ability dimension. Some individual's (the informed) have more precise information about the relative abilities of candidates than others (the uninformed). Since public opinion polls reflect the information of the informed, high approval for a candidate at the polls signals to the uninformed that he is more likely to be abler. However, high approval may also reflect changes in the distribution of the public on the issue space. As a result the uninformed partly confuse the two effects and public opinion polls tend to reinforce the effects of shifts in the public's ideological preferences on election results. A shift to the right that shows up as more support for the right-wing candidate at the polls is partly interpreted as a higher efficiency of this candidate. As a result he wins with a higher margin than the margin he would have won with in the absence of polls or under perfect information.
This paper examines the dynamics of allocation under increasing returns within a model where agents choose between technologies competing for adoption and where each technology improves as it gains in adoption. It shows that the economy, over time, can become locked-in, by "random" historical events, to a technological path that is not necessarily efficient, not possible to predict from usual knowledge of supply and demand functions, and not easy to change by standard tax or subsidy policies. Rational expectations about future agents' technology choices can exacerbate this lock-in tendency. It discusses the implications for economic history, policy, and forecasting. Copyright 1989 by Royal Economic Society.
