ABSTRACT. ==== Rationale ==== Humans are an especially cooperative species (Bowles & Gintis, 2011). We are also remarkably honest, even when tempted to lie to secure personal profit (Abeler et al., 2019; Gerlach et al., 2019). A recent body of work demonstrates that when honesty and cooperation are at odds (Weisel & Shalvi, 2015, 2022; Leib et al. 2021) honesty often gives way to corrupt collaboration, suggesting that honesty and collaboration are tradeable moral currencies. In the current project we explored how the balance between individual (dis)honesty, collaborative (dis)honesty, and individual cooperativeness depends on the cultural setting, connecting to a growing body of research demonstrating significant cultural variation in ethical thinking and behavior (Gächter & Schulz, 2016; Schulz et al., 2019).

==== Method ==== We conducted an online study in 20 countries (N≅11,500). The study included measures of individual (dis)honesty (using a variation of a die-rolling task), collaborative dishonesty (using a variant of the dyadic dishonesty task, Leib et al., 2021), cooperativeness (using a public good game), and a host of additional individual measures (cognitive reflection task, honesty-humility scale, trust, family ties, tightness, socioeconomic stratification, categorization type, individualism-collectivism).

==== Results and Interpretation ==== At the individual level, the tendency to engage in corrupt collaboration is positively related to individual dishonesty and negatively to individual cooperativeness. This result suggests that although a successful, mutually beneficial relationship with a partner, which is based on joint dishonesty, clearly requires cooperation, a cooperative disposition, which is related to a more general tendency toward moral behavior, seems to hinder such joint, corrupt enterprises. At the country level, the main result is that the tendency to switch from being honest when acting alone to being dishonest in collaboration with others is positively related to the Prevalence of Rule Violations in the country (PRV; a measure of political fraud, tax evasion and corruption, Gächter & Schulz, 2016). For example, the proportion of participants who are honest in the individual task, but dishonesty in the collaborative task is lowest in Sweden (a low PRV country) and highest in Nigeria (a high PRV country). PRV is thus not only related negatively to individual dishonesty (Gächter & Schulz, 2016; also current data), but also to the strength of the temptation that corrupt collaboration poses to otherwise honest individuals.

==== Conclusions ==== Our results are the first to demonstrate how individual and collaborative dishonesty co-vary across societies. Previous work already established that societal variation in individual (dis)honesty is related to the prevalence of rule violations in a country, and that collaborative settings pose a temptation that many individually honest people find hard to resist. Here we show that country level PRV is not only related to individual dishonesty, but also to peoples’ inability to resist the temptation to engage in corrupt collaboration. In other words, we show that the degree to which honesty and collaboration are tradeable moral currencies varies systematically—is positively related country level PRV—across societies.

==== References ==== Abeler, J., Nosenzo, D., & Raymond, C. (2019). Preferences for truth‐telling. Econometrica, 87(4), 1115-1153. Gächter, S., & Schulz, J. F. (2016). Intrinsic honesty and the prevalence of rule violations across societies. Nature, 531(7595), 496-499. Gerlach, P., Teodorescu, K., & Hertwig, R. (2019). The truth about lies: A meta-analysis on dishonest behavior. Psychological bulletin, 145(1), 1. Leib, M., Köbis, N., Soraperra, I., Weisel, O., & Shalvi, S. (2021). Collaborative dishonesty: A meta-analytic review. Psychological Bulletin, 147(12), 1241. Schulz, J. F., Bahrami-Rad, D., Beauchamp, J. P., & Henrich, J. (2019). The Church, intensive kinship, and global psychological variation. Science, 366(6466), eaau5141. Weisel, O., & Shalvi, S. (2015). The collaborative roots of corruption. Proceedings of the National Academy of Sciences, 112(34), 10651-10656. Weisel, O., & Shalvi, S. (2022). Moral currencies: Explaining corrupt collaboration. Current opinion in psychology, 44, 270-274.

ABSTRACT. Humans have the propensity of reacting to unfair situations even when they are not directly affected by the violation and when there is no future reward for themselves by doing so (Fehr& Fischbacher, 2004b; Gintis, 2000). According to many evolutionary scientists, intervention of third-parties to violations of cooperation or equal sharing norms plays a key role in maintaining cooperation in humans (Boyd et al., 2003; Tomasello, 2016). Past research have shown that third parties are even willing to incur a cost to intervene after norm violations, and this behavior has been identified as an altruistic behavior (Fehr & Gächter, 2002; Shinada et al., 2004). Third parties may intervene after norm violations by either compensating the victim or punishing the transgressor (Gummerum et al., 2016; Liu et al., 2018). We aimed to understand the dynamics of third-party interventions by differentiating these two different types of third-party behaviors and investigating their proximate mechanisms. We treated social value orientation (SVO) as an individual-level proximate mechanism, which refers to individuals’ tendencies of considering the interests of others or themselves in social interactions and cooperation (Van Lange et al., 1997). Three different types of SVO have been identified; prosocial orientation (i.e., preference for joint outcomes), individualistic orientation (i.e., preference for own outcome) and competitive orientation (i.e., preference for a maximum difference between own and other’s outcome). We investigated whether third-party punishment and compensation differ from each other in terms of their relationships with prosocial orientation. In addition to considering individual level differences in prosocial orientations, we investigated the effect of relatedness motivation orientation which was experimentally manipulated, since past research has shown that highlighting motivation orientation increased prosocial intentions and behaviors (Pavey et al., 2011). This study was pre-registered. The study sample consisted of 306 participants (48 male, 249 female, 6 nonbinary, 3 other; Mage=18.9; SD= 2.04). Participants completed the Nine-Item Triple Dominance Scale (Van Lange et al., 1997); answered a set of self-affirmation questions related to the experimental condition they randomly assigned (relatedness, autonomy, or competence; adapted from Deci and Ryan, 2000; taken from Pavey et al., 2011; Reed & Aspinwall, 1998) and played a Third-Party Intervention Game (adapted from Fehr & Fischbacher, 2004; taken from Gummerum et al., 2016). In order to check the effect of manipulation three one-way Analyses of Variance (ANOVAs) were conducted. The results showed that the experimental manipulations were successful in highlighting the three basic needs. In order to test research questions, the Linear mixed-effects models were run in RStudio with participants’ punishment and compensation expenditure as continuous variables. The main effects of altruism, relatedness, offer by Player A and decision type (punishment vs compensation), and all possible two-way, three-way, and four-way interactions were entered as fixed effects. Subject identities were entered as random intercepts. Firstly, the results showed that participants were more likely to spend their points for punishment (M =. 79, SD = 1.15) than compensation (M =. 59, SD = .99; B = -.81, p< .001), and the expenditures to either punish the transgressor or compensate the victim decreased as the offer size increased (B = -.40, p< .001). This demonstrates a preference for incurring costs to intervene following norm violations in unfair conditions, with an inclination toward punishment. Secondly, the interaction effect of Decision type X Prosocial tendency was significant (B = -.04, p< .001), showing that whether one prefers to either punish or compensate in response to unfairness depends on their level of prosocial orientation. According to this, participants high in altruism spent similar amounts on punishment and compensation while participants low in altruism spent more points on punishment than compensation. Thirdly, the interaction effect of Decision type x Relatedness was significant (B = -.97, p< .05). Participants spent more points on compensation in the relatedness (M = .65, SD = 1.10) than all other conditions (M = .58, SD = .95), while there was no difference in points spent on punishment in the relatedness (M =. 80, SD = 1.13) and all other condition. This indicates that the effect of relatedness manipulation on third-party intervention behavior is valid for compensation but not for punishment. Finally, the interaction effect of Relatedness x Prosocial tendency x Decision type was not significant (B = -.02, p> .05), showing that prosocial tendency does not interact with relatedness motivation orientation in influencing the preferred type of third-party intervention behaviour. Overall, the findings show that individuals tend to spend their resources to respond to unfairness, they prefer punishment over compensation, and this inclination is modulated by prosocial tendencies as well as relatedness motivation orientations. In conclusion, prosocial tendencies and relatedness cues might be more proximate mechanism for third-party compensation than third-party punishment.

ABSTRACT. Aim Press companies sometimes sell newspapers using an honour system, which involves placing the papers at unmonitored distribution points and expecting customers to pay honestly when they purchase papers. The system is useful for selling papers during weekends and holidays. Yet, without control mechanisms or payment enforcement, many exploit the system by underpaying or taking papers without payment, causing substantial economic losses. The current research sought to examine the effectiveness of text-based interventions in increasing payment morale in an honour system in a large-scale, longitudinal field experiment in Austria, targeting newspaper customers (daily circulation of about 240,000) in two major Austrian cities (i.e., Vienna and Graz, with a total population of around 2.3 million).

Method We partnered with an Austrian print production company for this research. As the initial step, we crowdsourced intervention proposals from behavioural science communities. Based on the responses we received, we designed 19 interventions and had them evaluated by a representative online sample from Germany recruited from Prolific (n ~= 900). Participants predicted how effective these interventions would be in the exact context where they would be implemented. As the number of independent observations was limited in the field study, we selected the two most promising interventions to implement. One intervention suggested that the printing company would donate 1,000€ to a local children’s hospital in the respective city if the payment rate exceeds 80% (Donation condition). The other invited customers to participate in an “honesty contest” to find out which neighbourhood was the most honest in each city (Honesty Contest condition). We assigned distribution points with comparable payment rates to one of the intervention conditions or to a control condition (without intervention) and gathered payment rate data, which included eleven, six, and two observations for each payment counting unit from the pre-intervention, intervention, and post-intervention periods. During the intervention period, intervention texts were positioned right above the newspapers and were thus expected to be seen clearly by customers.

Results Overall, both interventions substantially increased payment rates (by approximately 40%, on average) compared with the control, and the effect persisted in the post-intervention period. While this increase is substantial, it should be noted that the pre-intervention payment rate was low (though we cannot disclose the exact number due to a confidentiality agreement with the company). Nonetheless, the total revenue still benefited remarkably from the interventions due to the scale of the implementation. The Honesty Contest intervention was slightly more effective than the Donation intervention. Furthermore, we observed considerable heterogeneity in the effect. The effect was mainly driven by observations from Graz (where the payment rate was twice as high as in Vienna before the intervention).

Conclusions Text-based behavioural interventions are effective in increasing payment morale in an honour system. The effectiveness of interventions may depend on the strength of existing social norms. More research is needed to understand better whether interventions that exploit people’s tendency to be competitive in being good and moral are generally more effective than interventions that appeal to individuals’ prosociality.

ABSTRACT. Summary How a business seeks consent from customers, so that the customer makes an informed choice about the interaction and its knock-on effects, can be hard to do well. By conducting behavioural science research in a large corporate, we found that consent-seeking touchpoints could be improved using behavioural design to both increase consent rates and uplift the customer experience.

Context and Background On a regular basis, consumers are asked by companies to share their personal data. Indeed, the digital economy model is based on the premise of exchange of personal information (consumer data) for business services. For example, consumers may be asked for consent to share their cookie history, to use their data to pre-fill fields, or to share their email address to receive marketing communications.

Consumer consent to this use of their data provides notable benefit to the company, who can then analyse behaviour, tailor the interaction, and/or send marketing material to the customer. Consumers, on the other hand, have a more complicated relationship with consent. Their ‘privacy calculus’ about whether to accept or decline the sharing of their information is complex and shaped by many situational factors (Culnan & Armstrong, 1999; Kehr et al., 2015). This can result in the ‘privacy paradox’ (Acquisti & Grossklags, 2003), a form of an intention-action gap, where customers may say that they care about their privacy but don’t take commensurate actions to protect it. Some scholars (Martin 2019) suggest that this is more an error and expectation gap than a paradox.

So, while the benefit to business is clear, the benefit to consumers is sometimes less so. And while businesses may have a short-term commercial incentive to design consent in a way that simply increases the ‘yes’ rate, there are ethical, regulatory, and reputational risks in doing so without consideration for how well a customer understands and values the exchange. Simply put, responsible businesses need to balance business and customer outcomes when obtaining marketing communication consent.

Research This context prompts the question of how to best seek consent from consumers that will result in a mutual benefit, and how to test this prior to investing in any major change to existing business process.

We designed and ran a series of RCTs that attempted to address the question of how to ask customers for consent, to best meet both business and customer objectives. Broadly, we used a kitchen sink / nudge stacking approach to test a set of behavioural mechanisms with different types of consent (primarily accepting to receive marketing communications) in different channels (digital and telephone contact centre). Some of the studies used a novel approach to test behavioural interventions using online framed field experiments (behavioural simulators).

We found that using behavioural insights in the design of how we seek consent from customers resulted in better outcomes for both our business and our customers. For example, some of the best performing interventions generated improvements in consent rate in the range of 14-50 percentage points, improved customer comprehension as well as attitude toward the brand.

Impact and discussion Overall, our research has valuable practical application, demonstrates the effective use of a novel research method, and provides the basis for discussion points that would be of interest to businesses, policy makers, and academics alike.

In terms of practical application, the behavioural insights generated by our research can be used in the design of customer consent interaction points across multiple channels, and can help organisations avoid costly fines or reputational damage associated with inadequate design of consent experiences. Using a framed-field research approach allowed us to test a sensitive topic in a low-risk way prior to any direct customer or commercial impacts. Finally, we hope this research will spark discussion about a range of related topics: optimising for informed choice; privacy calculus, paradox, and expectations; relative benefits of testing in market, framed field, in lab, or elsewhere; and creating win-win value exchanges between businesses and consumers more generally.

ABSTRACT. Intellectual capital exerts a significant influence on entrepreneurship. This study delves into one vital component of intellectual capital crucial for entrepreneurial endeavors: relational capital (RC). RC encompasses the value derived from trust, respect, familiarity, collaboration, and interpersonal bonds. The notion of RC, is rooted in the understanding that the success of entrepreneurial ventures hinges significantly on the relationships they cultivate within their milieu. Indeed, a substantial body of literature suggests that mastering social skills constitutes a pivotal aspect of the entrepreneurial process. In accordance with the intellectual capital literature, relational capital (RC) encompasses interactions with stakeholders, customers, suppliers, associates, and partners, constituting one of the three fundamental components of intellectual capital. The other two components are human capital, which pertains to the value derived from the collective stock of tacit and explicit knowledge possessed by individuals, and structural capital, which refers to the value stemming from the organizational structure, processes, and cultural knowledge internalized within the business entity. RC serves as an intangible asset crucial for fostering innovation, enhancing efficiency, fostering cooperation, forming alliances, and boosting overall performance. Furthermore, RC plays a pivotal role in various aspects of business operations. It can lower negotiating costs within alliances, facilitate favorable payment terms, provide strategic insights such as information on potential competitors or distribution channels, enable the establishment of a more robust alliance network encompassing finance and investment opportunities, and even offer invaluable emotional support from friends and. Entrepreneurial development relies heavily on extensive interactions and relationships with a spectrum of stakeholders, including potential partners, investors, bankers, customers, suppliers, employees, advisors, family members, and other collaborators. Nevertheless, RC is deemed to confer a competitive advantage, particularly in the nascent phases of business development. Given the consensus between the entrepreneurial and intellectual capital literature regarding the pivotal role of relationships in entrepreneurship, it prompts the inquiry into whether distinct personality factors influencing the caliber and extent of relationship cultivation can be identified and empirically scrutinized for their impact on entrepreneurship. In this paper, I propose a theoretical model that underscores a potentially significant personality driver of relational capital (RC), namely disgust sensitivity, and examine its moderating influence on entrepreneurial tendencies —which previous research has shown to significantly predict actual entrepreneurship. Disgust sensitivity represents a distinctive personality trait known to profoundly impact both physical interactions and psychological distancing in interpersonal relationships, thus playing a pivotal role in relationship building. The entrepreneurial tendency comprises several facets of the entrepreneurial personality pivotal for the initial phases of entrepreneurial development, such as awareness, creativity, opportunism, and vision. Research has demonstrated that this tendency significantly predicts actual engagement in entrepreneurship, thus positioning it not only as a personality trait but also as a robust proxy for determining entrepreneurial career choices. According to the proposed model, disgust sensitivity acts as a moderator for the direct impact of various broad categories of personality traits—known to influence relationship quality and satisfaction—on entrepreneurial tendency and through it on entrepreneurship. Alongside theoretical elaboration, the paper will provide empirical evidence to substantiate the proposed theoretical framework. In contributing to this existing literature, the paper introduces a crucial and often overlooked personality trait, namely disgust sensitivity, which governs both physical interactions and psychological distancing. This trait significantly impacts the interconnected self of an entrepreneur, who must actively engage in the social realm and cultivate extensive relationship capital. Specifically, I propose that individuals exhibiting high levels of disgust sensitivity experience moderating effects on several Big Five general personality traits crucial to entrepreneurship, in contrast to individuals characterized by low disgust sensitivity. Additionally, this paper bears close relevance to the economic entrepreneurial literature, which examines various facets of social capital and their influence on entrepreneurs' capacity to cultivate supportive relationships, particularly trust and networking. Furthermore, it intersects with the economic psychology literature, which has shown a keen interest in exploring the role of personality characteristics in economic behaviors, preferences, and outcomes. According to the Affect-as-Information theory, individuals implicitly inquire, "how do I feel about this?", with these feelings serving as sources of information during decision-making. Within this framework, disgust serves as information regarding trust in others and influences harsher moral judgments. High disgust sensitivity can consequently prompt individuals to withdraw from interactions and pose significant obstacles to relationship building. In contributing to this body of literature and addressing the growing interest in the impact of negative affect on entrepreneurship, the paper provides empirical support for the potential effect disgust sensitivity. It demonstrates how disgust sensitivity may influence not only simple behaviors and attitudes but also moderate the impact of personality traits on a more complex cognitive-motivational structure—namely, the inclination to become an entrepreneur, which is highly contingent on social contacts and relationships. To summarize, this study develops a theoretical model and offers empirical validation, positing that disgust sensitivity moderates the relationships between entrepreneurial tendency—which previous research has shown to significantly predict actual entrepreneurship—and the Big Five personality traits: Extraversion, Neuroticism, and Agreeableness. These traits are recognized for their impact on relationship quality and satisfaction. Furthermore, the model suggests that heightened disgust sensitivity mitigates the entrepreneurial gender gap, which partly arises from differences in networking behaviors between genders. Women tend to utilize their networks for relationship-building, whereas men often leverage networks for strategic and instrumental gains. The findings underscore the intertwined effects of disgust sensitivity and personality traits on a multifaceted, higher-order entrepreneurial tendency. In light of the prevailing focus on human capital accumulation and resource enhancement in entrepreneurship endeavors, this paper underscores the significance of personality traits in fostering social networking and relationship building, thereby constituting vital supplementary resources for entrepreneurship. Importantly, these factors also influence the entrepreneurial gender gap.

ABSTRACT. Introduction The foundation of behavioral insights research is robust measurement. This study focuses on measuring Exponential Growth Bias (EGB) in financial decision-making. EGB is the tendency to underestimate the effects of compound interest, which can significantly impact financial decisions. Accurate measurement of EGB is crucial for understanding its prevalence and developing interventions to mitigate its effects (Messick, 1995; Huston, 2010; Kurach et al., 2021; Lusardi, 2019; Schmeiser & Seligman, 2013). This research addresses the following question: How do baseline EGB measures, domain variations, and response patterns extend our understanding of EGB? Background A valid measure of a concept is essential to understand its magnitude and prevalence (Messick, 1995). Accurate measurement of EGB is key to better understanding its impact on financial decisions and developing effective interventions to reduce it (Huston, 2010; Kurach et al., 2021; Lusardi, 2019; Schmeiser & Seligman, 2013). However, measuring EGB presents unique challenges. Measurement Development The development of novel EGB measures in this research was guided by existing literature and a thorough review of financial literacy measures related to compound interest. EGB spans saving and borrowing scenarios, ranging from basic to complex, with explicit or implicit interest calculations. Relevant domains were identified and differentiated by complexity. Scenarios were classified as "basic" if they could be solved through single-step or mental calculations, characterized by lower interest rates, short time horizons, and lump sum principal amounts. "Complex" scenarios required multi-step calculations, more involved algebraic manipulation, or the use of financial formulas or calculators, characterized by high interest rates, long time horizons, and annuity structures. Most measures explicitly state the terms of interest and label it as compound interest. However, implicit scenarios involved questions where interest compounding was embedded within the transaction structure without explicitly labeling it as compound interest. These scenarios required participants to recognize that compounding logic applies, even though specific terminology was not used. The design process emphasized four key principles for effective measurement outlined in financial literacy research: relevance, simplicity, brevity, and capacity to differentiate (Lusardi & Mitchell, 2021; Huston, 2009). The development process began with 12 questions, eight adapted from established measures (Lusardi, 2010; Goda, 2019; Soll et al., 2013), and four newly developed items addressing common financial scenarios absent in prior research. Response Options Multiple-choice questions were used, with response options reflecting common errors in calculating compound interest. Errors were categorized as follows: • Ignoring Interest: Interest is not considered. • Less Than Simple Interest: Simple interest is underestimated. • Simple Interest Error: Interest is accounted for but not compound interest. • More Than Simple Interest: Compound interest is recognized but underestimated. These categories represent distinct levels of bias, illustrating a progression in magnitude. Correct responses and overestimation of interest are not categorized as errors because EGB is defined as the tendency to underestimate interest. Scoring EGB The scoring system for EGB measures was designed to differentiate between individuals who slightly underestimated and those who significantly underestimated compound interest. Correct answers received a score of 0, indicating no bias, while incorrect responses were assigned scores ranging from 1 to 3 to reflect increasing levels of underestimation. Results Baseline Scores Baseline scores revealed varying EGB levels across different compound interest calculations. The Total EGB composite measure showed a mean of 3.6 (SD = 1.9), with individual scores ranging from 0 to 9 out of a maximum possible of 13. Basic computational tasks yielded lower bias scores (M = 1.0, SD = 1.0) than complex calculations (M = 2.6, SD = 1.6). Borrowing scenarios showed marginally higher bias (M = 1.9, SD = 1.2) than savings scenarios (M = 1.7, SD = 1.2). Individual Questions Detailed response information for each individual question revealed distinct patterns of underestimation of interest. For example, in the basic saving question, 66% of participants correctly identified the correct answer. In contrast, only 27% correctly identified the future value in a complex savings scenario. The payday loan scenario revealed the most severe comprehension challenges, with only 11% correctly recognizing the final amount. Discussion Research Question 1: How do baseline EGB measures, domain variations, and response patterns extend understanding of EGB? Baseline EGB measures revealed systematic patterns that align with and extend previous research findings. EGB increases systematically with computational complexity and varies across financial contexts, which may relate to product familiarity. Responses predominantly defaulted to simple interest calculations when incorrect, consistent with prior research (De Bock et al., 2002; Putarek & Vlahović-Štetić, 2019). A clear progression in task performance emerged across the complexity spectrum. Participants demonstrated the highest comprehension of basic savings questions, with performance declining substantially for complex scenarios. The implicit interest question involving a payday loan scenario highlighted significant comprehension challenges, suggesting a dual cognitive barrier: recognizing implicit compounding and grasping the rapid escalation of debt under high interest rates. Participants' real-world experience with financial products also influenced their ability to estimate compound interest accurately. Familiarity with financial products mirrored performance patterns, with particularly poor performance in the payday loan scenario where product unfamiliarity compounded cognitive challenges. Conclusion This study underscores the importance of accurate measurement in behavioral insights research. The findings indicate that as financial products become more complex and less familiar, participants increasingly default to simple interest calculations or linear logic. These insights are crucial for developing effective interventions to reduce EGB and improve financial decision-making.

ABSTRACT. There are many reasons why young people should save money, including future financial security, emergency preparedness, financial independence, debt avoidance, more deliberate spending, and wealth accumulation. However, people in general, and young people in particular, often struggle with saving. Several factors hinder their ability to save, such as impulsive and unplanned spending, which is further exacerbated by marketing strategies both online and offline, low financial literacy, peer pressure, and social media influences (e.g., Huang, 2024; Lusardi, 2019; Penman & McNeill, 2008). To address this issue, we developed and tested four different interventions aimed at increasing individuals' willingness to save. First, we designed a three hour long behavioral education based on the recommendations of Lučić and Uzelac (2024). This intervention focused on teaching practical tools to encourage saving and reduce impulsive spending in young people, such as how to keep personal budget, how to save money and when and how to borrow money. Second, we created a 10-minute educational video presentation designed to make saving and investing more meaningful and attractive for young people. In video, drastically different futures of two young people are contrasted, Marija who spent prudently and saved/invested part of her income and Ivan who spent all his income on short-term wants and luxuries, where, unlike Marija, Ivan lacks any financial security in adulthood and cannot plan for things like buying a home or starting a family. Third, drawing from the literature on goal-setting and implementation intentions (e.g., Epton et al., 2017; Sheeran et al., 2024; Soman & Zhao, 2011), we developed a 10-15 minute online goal-setting module. This module guided participants through five steps: identifying a relevant savings goal, making it attainable and specific, forming "if-then" implementation intentions, and modifying their environment to facilitate adherence to their savings plan. The final intervention was an AI-based approach involving interaction with ChatGPT. We developed an online chatbot in which participants first stated a specific savings goal. This input was fed into ChatGPT, which was prompted to act as a financial advisor. Its task was to "help the students create a plan that would enable them to save enough money within the desired timeframe to achieve their stated goal." Participants engaged in at least two rounds of conversation with ChatGPT until they were satisfied with their personalized savings plan. A total of N = 476 high school and college students participated in our study in both waves of data collection (M = 21.76 years, SD = 3.14; 71% female). Participants were randomly assigned to one of the four intervention conditions or a control group. Because the behavioral education intervention was conducted live in schools, whereas the other interventions were administered online, the response rates for the online interventions were lower, leading to an uneven distribution of participants across conditions. Specifically, there were N = 155 participants (32%) in the behavioral education condition, N = 93 (19%) in the goal-setting condition, N = 77 (16%) in the educational video condition, and N = 58 (12%) in the ChatGPT condition. Additionally, there were N = 100 participants (21%) in the control group, which did not receive any intervention. We employed longitudinal research design, collecting data at two time points. The first wave of data collection took place in 2023. Approximately one year later, participants underwent one of the four interventions, followed by a second wave of data collection one to one-and-a-half months later. Our primary variables of interest were financial self-efficacy (e.g., "I believe I am capable of achieving my financial goals in the next three months"), perceived behavioral control (e.g., "My financial situation and the achievement of my financial goals are completely under my control"), and behavioral intentions (e.g., "I believe I will start following my own financial plan in the next three months"). To assess the effectiveness of our interventions, we conducted a mixed-design ANOVA, comparing each intervention to the control group separately. The main effect of interest was the interaction between experimental condition (Intervention vs. Control) and time (T1 vs. T2). If an intervention was effective, we expected this interaction to be significant, indicating that the dependent variables increased from T1 to T2 for the intervention group but remained unchanged or decreased for the control group. Our results indicate that behavioral education was the most effective intervention, as evidenced by significant time x condition interactions for all three dependent variables: financial self-efficacy (F(1, 241) = 14.61, p < .001), perceived behavioral control (F(1, 246) = 8.19, p = .003), and behavioral intentions (F(1, 232) = 6.76, p = .010). Unlike the control group, which showed little change over time, participants in the behavioral education condition exhibited increased scores on all three measures. The goal-setting intervention also showed promise, yielding a significant time x condition interaction for financial self-efficacy (F(1, 191) = 7.46, p = .007) and a marginally significant interaction for behavioral intentions (F(1, 191) = 3.66, p = .057), though it did not significantly impact perceived behavioral control. In both cases, scores in the intervention group increased, whereas those in the control group remained stable or declined slightly. Finally, neither the educational video nor the ChatGPT intervention significantly impacted the dependent variables. In conclusion, two of our four theoretically and empirically informed interventions — behavioral education and goal-setting — demonstrated potential for increasing young people's propensity to save. Further research is needed to refine these approaches and explore additional strategies to enhance financial decision-making among young consumers.

ABSTRACT. Research on goal setting and goal pursuit shows that people who have set a (financial) goal are more motivated to perform behaviors that lead to goal attainment (e.g., Gollwitzer and Brandstätter 1997). For example, people who set saving goals are more likely to actually save (e.g., Ye at al. 2025). In the financial advice industry, not unsurprisingly a big trend is “goal-based” advice, where the advisor makes a financial plan based on the client’s goals and goal prioritization.

The academic literature is remarkably silent when it comes to studying goal setting, in that, it starts from given goals and then looks at subsequent downstream decision making and behaviors. As a result, the focus of many studies is on increasing the motivation to achieve these financial goals. Other studies ask participants to formulate financial goals as part of the experimental procedure, such as getting them to think about a savings moment and the amount they want to save for this occasion (Ülkümen and Cheema 2011). However, setting financial goals as a mandatory response in an experiment says little about the relevance of these goals and their impact on financial behavior in real life. Anecdotal evidence from practice, however, shows that when people are asked about their financial goals, many are not able to mention at least one financial goal.

Therefore, we look into how people set financial goals and if the methods to support goal setting impact the number and types of goals set. First, in Study 1 we perform qualitative in-depth interviews with a diverse sample of consumers (N=10). We find that the majority of subjects has no financial goals “top of mind” and only when asked further and supported with showing lists of potential goals some subjects seem to have financial goals or new goals emerge. In Study 2 we then test quantitively based on a survey experiment (N=1,222, nationally representative) how different ways to elicit goals impact goal setting. Subjects were randomly assigned to 1 of 4 groups being asked about their financial goals either as open answer field or by showing a list of 5, 10, or 15 of in practice common goals (always with the option to add more). We find statistically significant evidence that longer lists of possible goals led to subjects reporting having more goals (and sometimes different goal), although these additional goals were often ranked less important by subjects. Moreover, we find a significant interaction with subject’s propensity to plan in that subjects with a high propensity to plan are the least susceptible to the goal list treatment effect.

Our findings allow for two interpretations. Either longer lists of goals help consumers to remind them of “forgotten” goals or, especially for non-planners they “create” potentially even not relevant goals. Our research thus provides valuable insights for designing decision guidance and (automated) financial advice, where careful consideration must be given to how goals are elicited to ensure they are authentic and relevant to the consumer.

References

Gollwitzer, P., & Brandstätter, V. (1997). Implementation Intentions and Effective Goal Pursuit. Journal of personality and social psychology, 73(1), 186-199.

Ülkümen, G., & Cheema, A. (2011). Framing goals to influence personal savings: The role of specificity and construal level. Journal of Marketing Research, 48(6), 958-969.

Ye, Post, Zou, Chen (2025), Savings goals matter - Cognitive constraints, retirement planning, and downstream economic behaviors, working paper.

ABSTRACT. This study investigates the spending and saving motives of digital bank customers, focusing on the use of mental accounting. Drawing on survey data from 1,979 Norwegian adults, we examine the prevalence and patterns of digital mental accounting, its relationship with financial literacy and financial self-efficacy, and its role in mitigating financial vulnerability.

Having control over personal finances can be difficult, and previous studies have shown that households have a tendency to underestimate expected future spending (Peetz and Buehler 2009). In the financial climate of today, this can be particularly challenging. As a result, an increasing amount of consumer debt and loan defaults can be observed among consumers (Reserve 2020, Finanstilsynet 2021, Finanstilsynet 2024). A way of combatting this problem could be by using some form of plan for one’s finances, as researchers have found that people who exhibit planning behavior arrive at retirement age with much more wealth than non-planners (Rabinovich and Webley 2007). One tool that can be used to help with financial planning is metal accounting, which refers to cognitive operations used to organize and keep track of financial activities (Thaler 1999). In our study we investigate whether setting specific labels or goals on digital bank accounts can be used as a form of digital mental accounting. Mental accounts can in many cases be seen as derived as a function of consumers’ goals, and setting up these types of accounts can be helpful in achieving these savings goals (Soman and Ahn 2011). According to Sui, Sun et al. (2021), households that overspend their income rarely set savings goals, which implies that setting these types of goals could give people more control over their spending behavior.

Data for this study has been collected through a survey by Kantar TNS and is representative of the Norwegian population. Norway is an ideal country to study digital mental accounting seeing as it is one of the most digitally advanced countries in the word, along with being one of the countries furthest along in becoming a cashless society (Mumtaza, Nabillah et al. 2020, EC 2023, NETS 2023). We built on work by Xiao and Noring (1994), for the dependent variables of categories for digital mental accounts: Daily: “Daily expenses”, Purchase: “Savings for e.g. travel, celebrations (e.g. weddings or birthdays), purchasing things (e.g. furniture, car, white goods, etc.), renovations, etc.” Emergency: “Emergencies and unforeseen expenses”, Retire: “Pension savings (except savings through the employer)”, Child: “Savings for children and/or grandchildren”’ Grow: “Savings for leisure, e.g. concerts, sports events, hobbies, etc.”. In addition we created our own category of Home: “Save for a home or a holiday home”, and one for Other: “Other purposes not mentioned above”. Binary logistic regressions have been used in the analysis.

Our findings highlight several key insights into how individuals leverage mental accounting in a digital context to manage their finances. We confirmed that digital mental accounting is a widely used tool among the participants, with distinct accounts serving as earmarked funds for specific purposes such as daily expenses, purchases, emergencies, and retirement. Younger participants were more likely to use multiple accounts, reflecting a generational shift in financial behavior influenced by digitalization. Furthermore, gender differences in account usage suggest that traditional norms and priorities continue to shape financial strategies, with women favoring accounts for emergencies and children, while men were more inclined toward growth and daily expenses. Additionally, the study underscored the importance of financial literacy and financial self-efficacy in promoting effective financial management. Participants with higher financial literacy and self-efficacy were significantly more likely to have saved for a rainy day and to engage in purposeful mental accounting. This reinforces the critical role of financial education in fostering resilience and long-term planning among individuals. The analysis also revealed how financial vulnerability, including difficulties in paying bills and holding high-interest debts, negatively impacts individuals’ ability to save for emergencies. However, the use of mental accounting and structured financial planning could potentially alleviate some of these challenges by enabling individuals to better allocate their resources.

References EC (2023). Digital Economy and Society Index (DESI) 2022 Norway. E. Commission. https://digital-strategy.ec.europa.eu/en/policies/desi-norway, European Union.

Finanstilsynet (2021, April 28th 2021). "Redusert utlånsvolum og fortsatt høy andel misligholdte forbrukslån. Flere inkassosaker knyttet til forbruksgjeld, og økt andel eldre krav." Retrieved January 3rd, 2022, from https://www.finanstilsynet.no/nyhetsarkiv/nyheter/2021/redusert-utlansvolum-og-fortsatt-hoy-andel-misligholdte-forbrukslan.-flere-inkassosaker-knyttet-til/.

Finanstilsynet (2024). RAPPORT OM UTVIKLINGEN I FORBRUKSGJELD. finanstilsynet.no, Finanstilsynet.

Mumtaza, Q. M. H., et al. (2020). "Worldwide mobile wallet: a futuristic cashless system." Bulletin of Social Informatics Theory and Application 4(2): 70-75.

NETS (2023). Nordic Payment Report 2023. NETS. https://insights.nets.eu/publication/nordic-payment-report-2023, NETS: 34.

Peetz, J. and R. Buehler (2009). "Is there a budget fallacy? The role of savings goals in the prediction of personal spending." Personality and Social Psychology Bulletin 35(12): 1579-1591.

Rabinovich, A. and P. Webley (2007). "Filling the gap between planning and doing: Psychological factors involved in the successful implementation of saving intention." Journal of Economic Psychology 28(4): 444-461.

Reserve, U. F. (2020). Report on the economic well-being of US households in 2019-May 2020. 2020. https://www.federalreserve.gov, US Federal Reserve: 68.

Soman, D. and H.-K. Ahn (2011). "Mental accounting and individual welfare." Perspectives on framing: 65-92.

Sui, L., et al. (2021). "An assessment of the effects of mental accounting on overspending behaviour: An empirical study." International Journal of Consumer Studies 45(2): 221-234.

Thaler, R. H. (1999). "Mental accounting matters." Journal of Behavioral Decision Making 12(3): 183-206.

Xiao, J. J. and F. E. Noring (1994). "Perceived saving motives and hierarchical financial needs." Financial Counseling and Planning 5(1): 25-44.

ABSTRACT. Introduction

This study explores the social and emotional pressures consumers face when making financial decisions during the pre-Christmas period. Specifically, we examine the associations between gift anxiety, materialism, gift motivation, and the propensity to spend and borrow at Christmas. Building on research by McNair, Ranyard, and Nyhus (2024), which found that high materialism predicts both a higher propensity to spend and borrow, and that an obligation gift motive predicts a higher propensity to spend, we extend this research by investigating whether gift anxiety is another psychological factor that increases the willingness to spend and borrow at Christmas. This is in addition to materialism (Richins & Dawson, 1992) and gift motives (Wolfinbarger & Yale, 1993).

Our study is inspired by the work of Sherry, McGrath, and Levy (1993), Wooten (2000), Otnes, Kim, and Lowrey (1992), and Reshadi and Givi (2023), who have examined the negative aspects of gift-giving. Traditionally, gift-giving is analyzed from a social exchange perspective, where gifts are seen as tools for maintaining and strengthening social relationships. However, some studies have revealed that gift-giving can lead to negative emotional outcomes, such as anxiety, and that some gifts are given out of a sense of obligation rather than positive motivation.

Sherry et al. (1993) found that gift exchange can cause high levels of anxiety and lead to interpersonal conflicts. Respondents described situations where gifts were used as weapons to make the receiver feel uncomfortable. Otnes et al. (1992) found that some receivers are perceived as more difficult because they express expectations that are hard to meet or determine for the giver. Wooten (2000), found that situations where it is important for the giver to elicit a certain reaction from the receiver, while simultaneously perceiving the likelihood of doing so as small, are particularly anxiety-inducing. Relatedly, Reshadi and Givi (2023) have shown that givers spend more on gifts for wealthy receivers to avoid imbalance in the gift exchange and to signal higher financial status, aligning with Wooten’s perspective that gifts are linked to self-presentation. These studies suggest that concerns about the gift receivers’ reactions may motivate givers to increase spending and, if necessary, borrowing to finance Christmas-related expenses.

This Study

Data were collected in Norway (n=656) and the UK (n=345) in the week before Christmas in December 2023. Respondents from the UK were recruited through Prolific, while Norwegian respondents were recruited through posts on social media (Facebook and LinkedIn). All respondents were over 18 years old.

The questionnaires included questions about the respondents’ age, sex, education level, occupation, number of children, personal and household income. Additionally, the questionnaire included measures based on previously validated scales: enjoyment and obligation gift motives (10 questions) (Wolfinbarger & Yale, 1993), the materialism dimensions of success, centrality, and happiness (9 questions) (Richins & Dawson, 1992), financial hardship (3 questions) (Lempers, Clark-Lempers, & Simons, 1989), money management practices (9 questions) (Garðarsdóttir & Dittmar, 2012), propensity to spend (8 questions) and borrow (8 questions) (McNair et al., 2024), five financial literacy questions, positive and negative feelings (12 questions) (Diener et al., 2010). We included 12 questions to measure gift anxiety, based on Wooten (2000). In both surveys, items were subjected to correlational analysis followed by principal components analysis. Factor structures were similar in both countries, allowing scores for each psychological factor to be calculated in the same way.

Preliminary Results

Hierarchical regression analyses for each of the two outcome variables were conducted for each survey separately. The sociodemographic predictors were entered in the first step, followed in the second by economic hardship and proactive money management, with the psychological factors entered in the third step. Preliminary results show that gift anxiety and materialism are significant predictors of the propensity to spend in both countries. We also find that the enjoyment gift motive predicts the propensity to spend in both countries, while the obligation motive is only a significant predictor in Norway.

We find that a higher materialism score (happiness dimension) predicts a higher propensity to borrow in both countries, while the obligation gift motivation is a positive predictor of the propensity to borrow in Norway. The increase in explained variance when adding psychological variables to the equation was much higher for the propensity to spend than for the propensity to borrow. Financial variables and money management skills were the most important predictors for the propensity to borrow.

References

Diener, E., Wirtz, D., Tov, W., Kim-Prieto, C., Choi, D. W., Oishi, S., & Biswas-Diener, R. (2010). New well-being measures: Short scales to assess flourishing and positive and negative feelings. Social Indicators Research, 97(2), 143-156.

Garðarsdóttir, R. B., & Dittmar, H. (2012). The relationship of materialism to debt and financial well-being: The case of Iceland’s perceived prosperity. Journal of Economic Psychology, 33(3), 471-481.

Lempers, J. D., Clark-Lempers, D., & Simons, R. L. (1989). Economic hardship, parenting, and distress in adolescence. Child Development, 60(1), 25-39.

McNair, S., Nyhus, E. K., & Ranyard, R. (2024). Propensity to spend and borrow at a time of high pressure: The role of the meaning of Christmas and other psychological factors. Frontiers in Behavioral Economics, 3, 1385609.

Otnes, C., Lowrey, T. M., & Kim, Y. C. (1993). Gift selection for easy and difficult recipients: A social roles interpretation. Journal of Consumer Research, 20(2), 229-244.

Reshadi, F., & Givi, J. (2023). Spending the most on those who need it the least: Gift givers buy more expensive gifts for affluent recipients. European Journal of Marketing, 57(2), 479-504.

Richins, M. L., & Dawson, S. (1992). A consumer values orientation for materialism and its measurement: Scale development and validation. Journal of Consumer Research, 19(3), 303-316.

Sherry, J. F., McGrath, M. A., & Levy, S. J. (1993). The dark side of the gift. Journal of Business Research, 28(3), 225-244.

Wolfinbarger, M. F., & Yale, L. J. (1993). Three motivations for interpersonal gift giving: Experiential, obligated, and practical motivations. Advances in Consumer Research, 20(1).

Wooten, D. B. (2000). Qualitative steps toward an expanded model of anxiety in gift-giving. Journal of Consumer Research, 27(1), 84-95.

ABSTRACT. Introduction Nudging, defined as subtle changes in choice architecture to influence behaviour while maintaining freedom of choice, is widely recognized as a promising tool for public health promotion. Since its conceptualization by Thaler and Sunstein in 2008, research on nudge interventions has grown substantially, leading to numerous literature reviews across various health domains. However, a comprehensive synthesis of applied nudging methods and effective interventions for public health is still lacking. This overview seeks to fill this gap by summarizing findings from recent literature reviews and providing a foundation for future research in this field.

Methodology The literature search was restricted to reviews published between January 2018 and May 2024, sourced from Scopus, WoS, Science Direct, and EBSCO. Search terms included keywords related to nudging and various health behaviours or healthcare delivery. Findings were grouped into thematic domains, describing tested interventions, effectiveness, and research gaps and limitations.

Results A total of 46 literature reviews, including 35 systematic reviews, were analysed. Studies examined nudging health behaviours, including lifestyle changes and disease prevention and management. Additionally, studies explored how health professionals can be nudged to improve healthcare delivery. Figure 1 presents the detailed thematic domains along with corresponding references. Studies have demonstrated positive but heterogeneous results, with general findings for each thematic domain outlined below.

1. Lifestyle changes Nudging interventions show promise in promoting healthy food choices by lowering effort-related barriers (e.g., increasing accessibility of healthier options), setting healthier defaults, and providing simplified (e.g., traffic-light) food labelling. Small positive effects have been observed with variations in plate, portion, or cutlery size, and social norms have been found to influence choices when dining out. To encourage physical activity, effective strategies include point-of-choice prompts, such as banners promoting stair use in public spaces. Wearable activity trackers enhance conscious exercise behaviours, including step counts and moderate-to-vigorous activity. Gamification, commitment lotteries, and social incentives also show positive but modest results. Research on sleep-related nudges remains insufficient for broad conclusions. For addressing addictive substance consumption, school-based interventions show minimal impact, while strategies for tobacco and alcohol consumption prevention in adults by increasing salience of information or incentives, providing feedback, or combining nudges demonstrate some effectiveness. Active approaches, such as portion adjustments with immediate measurement of outcomes, yield positive results, whereas passive nudges in complex environments are ineffective. Gain-framed smoking cessation messages show a slight but not statistically significant advantage over loss-framed ones, warranting a reevaluation of traditional public health messaging strategies.

2. Disease prevention and management To increase vaccination rates, reminders, informational nudges, and combinations of nudges are commonly used. Evidence highlights the success of nudges that incentivize parents and healthcare professionals, enhance information salience, and utilize trusted messengers. To increase screening uptake, techniques such as framed information, default bias, and authority bias have been identified, though ethical concerns about transparency and the evaluation of benefit-harm ratios have been raised. For reducing healthcare appointment no-show rates reminders via SMS, phone calls, or mail are the most common and effective interventions. In disease management, tested methods for supporting the self-management of patients with chronic conditions include reminders, planning prompts, small financial incentives, and feedback. While nudging positively impacts self-care behaviours, such as medication adherence, there is a lack of evidence supporting improvements in broader disease control outcomes, such as reduced hospitalizations. For diabetes management, the use of reminders, gamification, and social modelling through group meetings has demonstrated particularly positive effects. Regarding HIV management, opt-out defaults, active-choice defaults, and lottery incentives have effectively increased HIV testing uptake, with lottery incentives also enhancing HIV prevention and antiretroviral therapy adherence. For tropical disease prevention in low- and middle-income countries, nudging primarily focuses on general health behaviours (e.g., handwashing) with limited attention to specific disease control measures (e.g., medication adherence). In non-specific disease prevention, such as hand hygiene and mask use, informative interventions using messages, posters, visual cues, environmental modifications, and improved access have demonstrated positive effects.

3. Nudging health professionals to improve healthcare delivery Nudging interventions effectively assist healthcare professionals in clinical decision-making, including ordering medications, laboratory tests, imaging, prescribing medications, and enhancing implementation and adherence to clinical guidelines. Strategies that influence clinical behaviour include framing information, providing feedback (e.g., comparing antibiotic prescriptions among peers), modifying default options (e.g., automatic substitution for generic medications), enabling active choices, delivering reminders, adjusting options, and minimizing effort for desired choices. Digital platforms, such as electronic health records, have proven effective in implementing nudges. These systems enable the delivery of nudging information or adjustments to how options are presented. Passive nudges that require minimal user action (e.g., default prescription changes) are particularly effective. In line with the 'nudge ladder theory', stronger interventions, such as default settings, consistently outperform less intrusive approaches like information provision.

Discussion While nudging shows promise for public health promotion, its effectiveness varies due to contextual, population-specific, and methodological factors. The reviewed studies highlight significant research gaps and limitations. Key challenges include insufficient evidence on long-term impacts, particularly for interventions fostering sustained behavioural changes like habitual physical activity. High heterogeneity, inconsistent outcome measures, and limited data from low-income and crisis settings further constrain applicability. Researchers emphasize the need for rigorous, well-designed studies, including preregistered trials to reduce publication bias, and greater focus on diverse populations, such as children and low socioeconomic groups. Future research should also explore multicomponent nudging strategies combined with complementary approaches for improved outcomes. Additionally, standardizing terminology, evaluating economic aspects, and addressing ethical concerns, particularly for passive nudges, remain critical. Addressing these gaps could enhance the scalability and impact of nudging public health.

Conclusions Nudging shows considerable potential to improve public health, though its effectiveness is described as moderate and heterogeneous. While positive results have been reported across various health-related domains, the variability in effectiveness underscores the need for context-specific approaches, better conceptual clarity, and further well-designed research on long-term impacts and scalability.

ABSTRACT. INTRODUCTION AND MOTIVATION Improving the quality of care is a key objective of healthcare management and policy. To achieve this goal, policy makers are encouraged to look beyond traditional financial incentives by focusing on both extrinsic and intrinsic motivations of health care providers. Shedding light on this important topic, our study analyzes the interplay of both kinds of motivations regarding their impact on physicians’ medical service provision and the quality of care. This is an underexplored research area, as mostly the impact of extrinsic factors, such as payment schemes, on providers’ health care provision are studied, whereas the effect of intrinsic factors, such as personality traits, is mostly lacking. The extrinsic factors we are studying are the basic payment systems capitation (CAP), and fee-for-service (FFS), as well as blended CAP-based and FFS-based pay-for-performance (P4P) and mixed (Mix) payment schemes. All of them are known to influence health care providers’ behavior. For instance, CAP incentivizes underprovision, whereas FFS induces overprovision. Blended systems are assumed to mitigate these undesired effects as incentives of physicians and patients are better aligned. Controlled experimental studies provide support for this theoretical result. However, field evidence is less conclusive, in mixed systems and under performance pay. Despite its extended use, the impact of P4P on the quality of care appears moderate, if at all, with substantial heterogeneity in providers’ responses. A deeper understanding of what drives this heterogeneity is missing. Medical care providers have been found to be motivated by factors beyond financial incentives. Personality traits appear to cause heterogeneity in behavior as they relate to individuals’ intrinsic motivation. Personality traits have been found to explain behavior and responses to incentives in various contexts other than health care and relate to, for example, prosocial behavior, productivity and self-selection into the public sector. In health economics, evidence on how the interplay between intrinsic and extrinsic motivations drives physician behavior is scarce. The few results suggest a negative correlation between performance-pay effects and neuroticism as well as conscientiousness. Lab experiments on provider payments typically do not consider personality traits at all or do not analyze them separately. What is missing so far is a thorough understanding of the link between personality traits, physicians’ (intrinsic and extrinsic) motivations and outcomes. RESEARCH QUESTIONS The basic question is whether dissimilarities in personality traits affect provider behavior differently already under a given (basic) payment system. And if so, do we observe a different effect in CAP compared to FFS due to the diverging incentives of the two schemes? If it were the case that certain personality traits are related to more patient-regarding behavior than others, or that people with higher intensities in the same personality trait provide higher treatment quality, then there might be less room for improvement under additional financial incentives such as P4P. In this case, it is not the additional performance incentive that may fail, but it is the respective person’s characteristic(s) that is (are) relevant for an already higher treatment quality under the basic payment scheme. It thus appears that a systematic analysis on whether, and if so, how personality traits may contribute to understanding health care providers’ responses to financial incentives, is mostly lacking. In particular, insights on the potential of personality traits to explain heterogeneity in health professionals’ behavior are missing as is a comprehensive understanding of how and under which conditions personality traits may moderate the effect of blended payment systems. Our study contributes to filling these gaps by analyzing the following research questions: (1) How do personality traits relate to individual physicians’ behavioral responses to the basic payment systems CAP and FFS, and (2) how do personality traits interact with blended CAP- and FFS-based pay-for-performance and mixed payment schemes? METHODS To answer our first research question, we analyze laboratory and field data, as well as survey data (N = 659), based on behavioral experimental studies we have been conducting between 2011 and 2015. All studies follow the same experimental setting, including the decision situation, framing, and ex-post questionnaire. We link the data on personality traits (BFI-10) with the behavioral data of medical students (prospective doctors), practicing doctors, and nonmedical students who are randomly assigned to either a CAP or an FFS payment scheme. To answer our second research question, we focus on a smaller sample (N = 277). These are conditions where participants are confronted with a blended payment system in part 2 of the experiment. We study how personality traits interact with CAP- or FFS-based mixed payment schemes or with CAP- or FFS-based pay-for-performance systems. We compare the results with responses in the related basic non-blended scenarios at a within-subject level. RESULTS We find that Agreeableness significantly relates to the supply and the quality of care. Under CAP and FFS, more agreeable individuals are more patient-regarding by providing higher quality of care. Other traits do not significantly relate to providers’ behavior in the basic systems. While we find that by introducing mixed and performance pay systems, the quality of care under CAP and FFS is significantly improved, some personality traits moderate the incentive effect — but not under all blended payment schemes. In FFS-based and CAP-based mixed systems, it is Agreeableness that relates to a higher quality of care. No other personality trait does so. In CAP+P4P, it is higher Agreeableness and Conscientiousness that relates to a higher quality of care. Yet, more conscientious and more agreeable individuals respond significantly less to this payment scheme. Under FFS+P4P, personality traits are apparently not behaviorally relevant. CONCLUSION In essence, we observe only in the CAP-based performance-pay system that the incentive effect is moderated, namely by Agreeableness and Conscientiousness. Our findings seem informative for better incentivizing physicians — in particular, the tailoring of financial incentives based on personality traits, as well as the respective necessary organizational changes.

ABSTRACT. Within behavioural public policy, the psychology of poverty seeks to produce generalizable, mechanistic, and policy-relevant knowledge, and has already made seminal contributions to the understanding of poverty from a behavioural and experimental perspective. This methodological talk evaluates the field's aim by asking what consequences there are for the mechanistic extrapolation and generalization of causal effects, if one evaluates the paradigmatic experimental designs with the concept of construct validity. With the concept I emphasize its theoretical side of connecting high-order constructs with empirical operationalizations and inspecting experimental interventions as operationalizations. I argue that the field's aim of producing robust policy-relevant causal generalizations and extrapolations is in tension with the current practices of construct validation around the interventions and mechanisms studied by the field. The practices and their downstream consequences to experimental design and resulting causal inferences do not easily allow generalizations and extrapolations made.

Building on a behavioural economics perspective, the field began with two seminal experimental studies a decade ago that produced bold causal generalizations about the negative psychological effects of poverty and resource scarcity. Since the field has expanded significantly, and the paradigmatic experimental interventions include priming, payday, harvest, and laboratory strategies (with some field experiments). Now, the main construct, scarcity, is operationalized flexibly and is often linked to “universal” cognitive mechanisms that apply across contexts. However, when flexible operationalizations are left implicit, challenges arise. Cognitive mechanisms may vary or interact with other kinds of mechanisms across contexts. The heterogeneity of scarcity can produce varying effects depending on duration and intensity—as could be in different welfare-state settings, as some evidence points out—which broad generalizations may fail to capture. Moreover, the nature of the experimental interventions—what they changed and what they did not—along with their relations to the higher-order construct of scarcity, is not always clear. Tendency to overlook construct validity leads to an incomplete understanding of how interventions, assumed mechanisms, and outcomes relate to the various behaviours and (choice) environments of the poor, thus hindering building generalizations. Lack of construct validity can also partly explain the mixed evidence-base of the field.

The psychology of poverty holds great value, with experiments remaining central. The field has shown that poverty has cognitive effects through causal mechanisms, even though much work remains to be done on the specifics. To get further, instead of applying the paradigmatic cognitive mechanism schema by default to new studies, researchers could consider more how empirical operationalizations reflect the phenomenon with its social-cultural aspects, treating interventions as local and context-dependent, and embracing—sometimes messy—mixed-methods triangulation to untangle how interventions are actually implemented and how mechanisms seem to work. Thus, many issues outlined do not have a straightforward statistical solution, but require better use of theory, as was emphasized in the early stages of psychometrics. Moreover, insights from the mechanistic approach to explanation can help balance the trade-off between granularity and generality in causal claims and thus recalibrate the aim of the field more realistic. Collaboration between behavioural researchers and poverty scholars could further help by developing policy-relevant constructs of poverty and scarcity, even though there is no need nor possibilities for a “grand theory” of poverty.

ABSTRACT. This paper studies how key economic preferences are distributed in a society and how this variation relates to individuals’ early health endowment and family background. To answer these questions, we (i) present a novel dataset that includes incentivized measures of competitiveness and key economic preferences (social preferences, risk preferences, time preferences) as well as survey data from nearly 6000 Finnish individuals aged from 15 to 30 years, (ii) link these data with respondents’ medical birth records, medical histories, and a broad variety of administrative records containing educational and labor market outcomes for respondents and their family members, (iii) systematically explore a large set of potential developmental and demographic correlates of economic preferences and assess their relative importance in explaining individual and geographical variation in preferences, and (iv) investigate how children’s early health endowment and parental inputs may jointly shape economic preferences. The key idea is to link incentivized preference data with rich individual- and population-level datasets to expand the scope of research into important domains of economic behavior and decision-making that are not easily studied using behavioral experiments or administrative datasets alone.

The evidence presented in this paper suggest that (i) competitiveness, risk taking, patience and altruism increase in an age interval from 15 to 30 years, (ii) heterogeneity in preferences across geographical regions tends to outweigh individual variation in preferences, (iii) individuals’ early health endowment predicts their competitiveness in adolescence and early adulthood, and (iv) individuals’ personality traits tend to be more accurate predictors of their economic preferences than a wide range of demographic and health-related variables.

ABSTRACT. We investigate the effects of regional fragmentation and heterogeneity in the ethnic composition of groups on resource extraction. Using novel and unique data from resource-rich, ethnically diverse Guinea, this paper presents a lab-in-field experiment on a social dilemma featuring multiple generations. The research questions are: 1) What is the impact of social identity on natural resource extraction behavior? and, 2) Are there ethnic differences in sustaining resources for future generations?

With respect to environmental economics literature, this study will contribute to limited experimental literature on common-pool resources in a dynamic, inter-temporal environment. Most are laboratory work involving complex designs not adaptable to the field (Fisher et al, 2004). Outside experimental literature, it is related to theoretical and macro-econometric research on the “natural resource curse” whose empirical validity remains unclear. While Holder (2006) and van der Ploeg and Rohner (2012) state that natural resource dependence intensifies conflict, Brunnschweiler and Bulte (2009) argue the opposite. With ethnic groups in the resource-rich country of Guinea as case study, our work provides a new micro-perspective through social science experiments. It also gives a novel behavioural analysis on how culture and ethnic diversity plays into the “natural resource curse”. In relation to experimental economics literature, this paper is related to experiments on “social identity theory”. It states that membership of a social group means individuals display a higher concern for the welfare of fellow “in-group” members than outsiders. Our paper’s hypothesis is that dynamic resource extraction might be lower in ethnically homogenous groups than heterogenous groups.

Multiple sessions of pen-and-paper behavioural experiments were conducted in coordination with two big universities in Conakry, Guinea in Sub-saharan Africa: L'université Général Lansana Conté de Sonfonia-Conakry and Universite de Simbaya. Guinea is a West African country characterized by heavy economic dependence in natural resources (NRGI, 2015). It is the world’s top producer of bauxite (unprocessed aluminum). Its vast resources co-exists with a population that is one of the most ethnically heterogenous in the world. The unique demography of its capital Conakry offers convenient access to two of its most dominant ethnic groups, the Fulani and Malinke When recruiting participants, we focus on these groups because they differ in their cultural beliefs and history of resource use. Majority of the Guineans, 41%, belong to the Fulani tribe. They have a tradition of sedentary agriculture where resources are cultivated in one land. They have a unique code of behavior called “pulaaku” characterized by patience, honesty, and self-control. Comparably, 33% of the population belong to the Mandinka tribe. They are Muslims who adhere to strict rules of socially stratified communities. They follow a subsistence system where they use natural resources, esp. agricultural crops, based on their current needs only. Furthermore, the Fulani and Malinke have a history of ethnic tensions and politically motivated violence (WA, 2018).

Again, using laboratory-in-the-field experimental methods, we investigate the effects of regional fragmentation and ethnic heterogeneity on non-renewable resource extraction. Different past literature, this study consists of players across multiple generations, i.e. five generations (periods). The dynamic game has a new generation of two players every period. The amount of natural resource available for future generations is affected by the extraction rates of past generations. The game is also characterized by shocks endogenously caused by players’ extraction decisions. After a given threshold (i.e. “tipping point”), the natural resource suddenly drops to lower values.

The predictions of the above dynamic game is then tested on a two-player, field experiment. It is conducted on a sample of Fulani (ethnic majority group from the Middle Guinea region) and Malinke (minority group from the Upper Guinea region) participants. We frame instructions in the context of bauxite, the natural resource that Guinea’s economy is heavily dependent on. The treatment variable in the experiment is the ethnically-inclined region of origin of the two players across generations. We vary the degree of ethnic heterogeneity of groups who extract the non-renewable resource. There are 4 treatments: (1) CONTROL (i.e. No mention of ethnicity of the two players in the game); (2) HOMOGENEOUS: Fulani Only (2 players for each generation are from Middle Guinea); (3) HOMOGENEOUS: Malinke Only; and (4) MIXED: Mixture of 1 Fulani and 1 Malinke players. All treatment groups play the same non-renewable resource extraction game. The only difference is that players are reminded of the regional origin of the other player they are paired with (i.e. Priming on their ethnicity compared to their partner’s ethnicity). For example, for the HETEROGENOUS treatment, they are reminded with this phrase: If you are from High Guinea (Middle Guinea), your partner will be from the other region Middle Guinea (High Guinea). The hypothesis is that extraction will be least in homogenous group because they are more likely to care about the welfare of future generations. Preliminary results & statistical analyses suggest that regional fragmentation significantly affects the behavior of the majority group, the Fulani. Across time, observed data implies that the Fulani significantly choose lower extraction rates, especially at the latter periods. Homogenous groups that only consist of Fulani members are less likely to deplete the natural resource, and are implicitly more concerned of future generations. This is more prevalent when they are paired with a player of the same ethnic identity as theirs.

---

ABSTRACT. Abstract for IAREP 2025 Interactive Session Proposal Title: We Know the World is Hard: How Do We Avoid ‘Dumb’ Behavioural Public Policy?*

Behavioural science has become an essential tool for governments and organisations grappling with pressing challenges. However, as Michael Hallsworth’s Manifesto for Applied Behavioural Science (2023)1 highlights, the field faces critical limitations, including over-simplification, mechanistic approaches, limited representation of target audiences, and a failure to grapple with complexity and systemic change. Encouragingly, teams adjacent to government like BehaviourWorks Australia have applied a Method that brackets more common behavioural science deep dives with systems informed problem definition, and scale up2 in over 400 projects, and Bursara have integrated systems thinking and agent based modelling into their work3. The EU Joint Competency Centre have compellingly detailed how they integrate behaviour and systems in supporting policy development4. While such developments provide a compelling roadmap for improvement, they do not document and build on the methodological and practical innovations already being used by applied behavioural science practitioners working inside the public sector. This interactive session aims to surface and document how behavioural science practitioners are navigating complexity and deploying systems thinking in their work. Rather than treating complexity as an intractable barrier, many BSci professionals working in government, industry, and policy units are actively finding ways to engage with systems thinking and systemic change, blending behavioural insights with broader governance, regulatory, futures/foresighting and institutional approaches. These real-world adaptations often outpace academic discussions, providing a valuable source of learning for both scholars and practitioners. They also tend be never reported in academic literature, obscuring important contributions from practice back into research. * (with apologies to Richard Thaller for the misquote). Session Goals Interrogate the critiques and opportunities related to applying behavioural science within complicated and complex, adaptive systems. Identify practical strategies and innovations that practitioners are already using to integrate behavioural science and systems approaches in their engagement with policy, administration and governance. Explore tensions, challenges and opportunites that arise when behavioural public policy engages with systemic approaches. Document and synthesise these insights to contribute to a broader research agenda on behavioural science and systems change. Key Discussion Themes Relevance of critiques and opportunities: How do practitioners interpret and respond to concerns? Are they valid? Are they overstated? What is missing? Engagement with systems thinking and systemic change: Are behavioural science practitioners integrating systems approaches? If so, how? Methodological and disciplinary tensions: What happens when behavioural insights intersect with disciplines such as complex systems, and system change in governance, and public administration? What are the conflicts or synergies? Practical workarounds and innovations: What real-world strategies have practitioners developed to navigate complexity, overcome barriers, and integrate behavioural science and systems approaches effectively? Skills and capacity-building: Do practitioners have the necessary skills, capabilities and tools to engage with complexity? How well do they collaborate with cognate public policy ‘trades’ like design, innovation, and across disciplinary divides? If not, what training or institutional support is needed? Session Format This session is an interactive workshop rather than a traditional panel or presentation. Participants will be engaged in structured activities designed to elicit and synthesise insights from diverse experiences. Framing Talk (10-15 min): Overview of challenges and opportunities of relating behavioural and systems approaches How can we do behavioural public policy that is not ‘stupid’? Live Polling & Real-Time Feedback (5 min): Which critiques resonate most? Where have you encountered successful applications of behavioural science within complex policy environments? Small Group Breakout Discussions (25 min): Groups discuss and document: What strategies and methodologies they use to navigate complexity. Key barriers and tensions in their work. Practical case studies of behavioural science in complex systems. Plenary Synthesis & Insights Harvesting (15-20 min): Common themes and divergences are mapped during report-back. Discussion of gaps, challenges, and opportunities for further collaboration and research. Next Steps & Research Agenda (5-10 min): Exploring a collaborative rsource to document and share practitioner insights. Identifying opportunities for further methodological development. Who Should Attend? This session is designed for: Behavioural science practitioners working in government, industry, and consulting. Researchers interested in applied behavioural public policy. Nudge units, design thinkers, public sector innovators and policy labs exploring complex problems and responses to them Systems thinkers and public administration scholars engaging with behavioural science. Conclusion Behavioural public policy is at a crossroads. As Hallsworth argues, the field must evolve to remain relevant and effective in addressing global challenges. However, practitioners are already pushing the boundaries, integrating behavioural science with systemic change approaches in ways that are not yet fully documented. This session will surface those experiences, creating a collective knowledge base to inform both academic research and applied practice. Through this interactive workshop, we aim to capture and synthesise real-world strategies, helping to build a more sophisticated, context-sensitive, and system-aware approach to behavioural public policy. The insights generated will contribute to a broader research agenda and provide practical takeaways for participants seeking to improve their own practice. They will also inform a new Masters of Systems and Behaviour Change at Monash University.

We gratefully acknowledge contributions from past and present members of the Australian, Victorian and Queensland Government Behavioural Science teams, as well as the Estonion Public Sector Innovation Unit, in developing this session. We'd also like to thank colleagues from Behaviourworks, as well as BIT's Michael Hallsworth, who kindly provided input.

References: 1. Hallsworth, M. A manifesto for applying behavioural science. Nat. Hum. Behav. 7, 310–322 (2023). 2. Inspiring Change: How to Influence Behaviour for a Better World. (Monash University Publishing, Melbourne, 2024). 3. Valle, E. D. D., Jang, C. & Wendel, S. Behavioral Systems: Combining Behavioral Science and Systems Analysis. https://busara.global/wp-content/uploads/2024/10/GW-8-RAOF.pdf (2024). 4. European Commission. Joint Research Centre. Unlocking the Full Potential of Behavioural Insights for Policy: From Influencing the Individual to Shaping the System. (Publications Office, LU, 2025).

ABSTRACT. Session Goals Interrogate the critiques and opportunities related to applying behavioural science within complicated and complex, adaptive systems. Identify practical strategies and innovations that practitioners are already using to integrate behavioural science and systems approaches in their engagement with policy, administration and governance. Explore tensions, challenges and opportunites that arise when behavioural public policy engages with systemic approaches. Document and synthesise these insights to contribute to a broader research agenda on behavioural science and systems change. Key Discussion Themes Relevance of critiques and opportunities: How do practitioners interpret and respond to concerns? Are they valid? Are they overstated? What is missing? Engagement with systems thinking and systemic change: Are behavioural science practitioners integrating systems approaches? If so, how? Methodological and disciplinary tensions: What happens when behavioural insights intersect with disciplines such as complex systems, and system change in governance, and public administration? What are the conflicts or synergies? Practical workarounds and innovations: What real-world strategies have practitioners developed to navigate complexity, overcome barriers, and integrate behavioural science and systems approaches effectively? Skills and capacity-building: Do practitioners have the necessary skills, capabilities and tools to engage with complexity? How well do they collaborate with cognate public policy ‘trades’ like design, innovation, and across disciplinary divides? If not, what training or institutional support is needed? Session Format This session is an interactive workshop rather than a traditional panel or presentation. Participants will be engaged in structured activities designed to elicit and synthesise insights from diverse experiences. Framing Talk (10-15 min): Overview of challenges and opportunities of relating behavioural and systems approaches How can we do behavioural public policy that is not ‘stupid’? Live Polling & Real-Time Feedback (5 min): Which critiques resonate most? Where have you encountered successful applications of behavioural science within complex policy environments? Small Group Breakout Discussions (25 min): Groups discuss and document: What strategies and methodologies they use to navigate complexity. Key barriers and tensions in their work. Practical case studies of behavioural science in complex systems. Plenary Synthesis & Insights Harvesting (15-20 min): Common themes and divergences are mapped during report-back. Discussion of gaps, challenges, and opportunities for further collaboration and research. Next Steps & Research Agenda (5-10 min): Exploring a collaborative resource to document and share practitioner insights. Identifying opportunities for further methodological development. Who Should Attend? This session is designed for: Behavioural science practitioners working in government, industry, and consulting. Researchers interested in applied behavioural public policy. Nudge units, design thinkers, public sector innovators and policy labs exploring complex problems and responses to them Systems thinkers and public administration scholars engaging with behavioural science. Conclusion Behavioural public policy is at a crossroads. As Hallsworth argues, the field must evolve to remain relevant and effective in addressing global challenges. However, practitioners are already pushing the boundaries, integrating behavioural science with systemic change approaches in ways that are not yet fully documented. This session will surface those experiences, creating a collective knowledge base to inform both academic research and applied practice. Through this interactive workshop, we aim to capture and synthesise real-world strategies, helping to build a more sophisticated, context-sensitive, and system-aware approach to behavioural public policy. The insights generated will contribute to a broader research agenda and provide practical takeaways for participants seeking to improve their own practice. They will also inform a new Masters of Systems and Behaviour Change at Monash University. We gratefully acknowledge contributions from past and present members of the Australian, Victorian and Queensland Government Behavioural Science teams in developing this session, and colleagues from Behaviourworks, as well as Michael Hallsworth who kindly provided input. References: 1. Hallsworth, M. A manifesto for applying behavioural science. Nat. Hum. Behav. 7, 310–322 (2023). 2. Inspiring Change: How to Influence Behaviour for a Better World. (Monash University Publishing, Melbourne, 2024). 3. Valle, E. D. D., Jang, C. & Wendel, S. Behavioral Systems: Combining Behavioral Science and Systems Analysis.https://busara.global/wp-content/uploads/2024/10/GW-8-RAOF.pdf (2024). 4. European Commission. Joint Research Centre. Unlocking the Full Potential of Behavioural Insights for Policy: From Influencing the Individual to Shaping the System. (Publications Office, LU, 2025).

ABSTRACT. Co facilitating

ABSTRACT. Co-facilitating

ABSTRACT. Co-facilitating

ABSTRACT. 1 Introduction Cooperation in public goods provision remains a fundamental challenge in economic and social research. While numerous studies have explored incentives and group dynamics, the roles of gender, environmental beliefs, and emotional intelligence in shaping cooperative behavior are relatively underexplored. This study aims to investigate the interplay between these factors, employing a public goods experiment framework combined with validated psychological scales.

2 Research Objectives The primary objective of this research is to examine how gender differences, environmental beliefs, and emotional intelligence interact to influence cooperative behavior in public goods games. Specifically, we address the following research questions: 1. Do stronger environmental beliefs correlate with higher levels of cooperation? 2. How does emotional intelligence impact cooperative decision-making? 3. Are there gender-specific effects in the relationship between emotional intelligence, environmental beliefs, and cooperation?

3 Methodology To explore these questions, the study employs a mixed-methods approach, integrating experimental data collection, validated psychological scales, and rigorous statistical analysis. The methodology consists of three key components: • Environmental Beliefs: Measured using the New Environmental Paradigm (NEP) scale, which distinguishes between ecocentric and anthropocentric perspectives. • Emotional Intelligence: Assessed through the Trait Emotional Intelligence Questionnaire (TEIQue), which evaluates emotional awareness, regulation, and social skills. • Cooperation Measurement: Conducted through a public goods game, where participants decide how much to contribute to a collective pool benefiting the group. A sample of participants was recruited, with gender differences analyzed to determine variations in cooperation levels. Regression models were employed to assess the impact of environmental beliefs and emotional intelligence on cooperative behavior.

4 Hypotheses Based on existing literature and theoretical considerations, the following hypotheses were tested: • H1: Participants with stronger ecocentric environmental beliefs exhibit higher cooperative behav- ior. • H2: Emotional intelligence, particularly emotionality and sociability, positively influences cooperation. • H3: Gender differences moderate the impact of emotional intelligence and environmental beliefs on cooperation, with women exhibiting greater sensitivity to these factors.

5 Key Findings The results indicate a significant relationship between emotional intelligence and cooperative behavior, with individuals scoring higher in emotional awareness and regulation demonstrating greater contribu- tions in the public goods game. However, the impact of environmental beliefs on cooperation was mixed, with ecocentric beliefs correlating positively with contributions but anthropocentric views showing no significant effect. Gender differences emerged as a crucial factor. Women exhibited greater responsiveness to both emotional intelligence and environmental beliefs, suggesting a stronger alignment between their emo- tional and cooperative tendencies. Moreover, emotionality played a more pronounced role in fostering cooperation among female participants.

6 Discussion and Implications These findings underscore the importance of integrating psychological and social dimensions in studying cooperative behavior. The results suggest that interventions aiming to enhance cooperation should con- sider emotional intelligence training and environmental education, particularly targeting gender-specific influences. For policymakers and educators, these insights can inform strategies for promoting sustainable be- havior and collective action. Enhancing emotional intelligence skills may foster greater social cohesion and cooperation in addressing public goods challenges.

7 Conclusion This study contributes to the interdisciplinary understanding of cooperation by highlighting the roles of environmental beliefs, emotional intelligence, and gender. The results provide a foundation for future research exploring psychological and social mechanisms in public goods provision, ultimately supporting efforts to enhance cooperative and sustainable behaviors.

ABSTRACT. There is an inherent difficulty in motivating climate action (Rossa-Roccor et al. 2021). Climate action requires people to overcome the loss of present-day sacrifices to reap future benefits that are often uncertain (IPCC, 2022). This can make the implementation of environmental policies particularly challenging. Several papers have explored alternative framings when justifying environmental policies. For example using economic, social, or health rationales (Dasandi et al. 2022). Some experimental research suggests that health framing is more effective at garnering public support compared to traditional environmental framing (Hart and Feldman, 2018; Mossler et al. 2017). Changing the labelling of the issue from “climate change” or “global warming” to “air pollution” has been shown to have a positive effect on acceptability (Mossler et al. 2017). Other research shows that people are less likely to believe emissions have negative consequences when framed in the context of climate change (as opposed to public health) and consequently, they display lower levels of support (Hart and Feldman, 2018). However, there is evidence to the contrary with one study finding environmental messaging to be just as effective as equivalent health messaging in increasing public acceptability (Dasandi et al. 2022).

The present study adds to this body of literature in several distinct ways. Previous studies utilise a general measure of support for climate policies as the dependent variable. However, people's preferences differ quite drastically depending on the type of policy proposed (Drews et al. 2016). As such, we present participants with a defined policy vignette. In addition, previous studies have largely focused on the impact of one frame versus another. Yet, in practice, policymakers usually cite several benefits and justifications for the policy they are proposing. As such, the present study explores the effect of combining frames. Lastly, it has been theorised that health framing is preferable because it is more personally relevant, i.e. it is less psychologically distant. To the researchers' knowledge, this has not been empirically tested until now.

Using a randomised between-subjects experiment (n=698, UK-representative sample) we test the impact of environmental and public health justifications on public approval for a city centre driving charge. The dependent variable is the participants’ acceptability of the policy, which is measured on a Likert scale. Upon entering the study, participants randomly received either an environmental (E), health (H), environmental with health co-benefit (Eh), health with environmental co-benefit (He) justification for the policy or a neutral message (control group). The environmental messages refer to the link between car emissions and climate change whilst the health messages highlight the association between toxins produced by cars and their impact on public health. Environmental policy preferences are shaped by people's beliefs about climate change and their level of trust in political institutions (Drews et al. 2016). As such, our paper also explores whether treatment effects vary according to levels of political trust and climate change beliefs (moderation analysis). We conduct a mediation analysis to assess whether the effect of the messages on approval can be partially explained by psychological distance.

Our results show that compared to the control, the co-benefit messages (Eh and He) have a significant and positive effect on policy approval. However, the environmental (E) and health (H) messages are not reliable predictors. Furthermore, we find a positive association between psychological distance and approval. People are more approving of the policy when they perceive climate change/air pollution to be an issue that is certain, whose effects will be realised in the near future, and when they view the issue as local and directly affecting people like themselves. However, we do not find a difference in psychological distance perceptions between the treatment groups and there is little evidence that psychological distance explains the effect of the treatment messages on approval. The treatment messages do not have a differing effect on approval depending on whether people believe climate change is caused entirely by human activity or a mixture of human activity and natural processes. However, we find some evidence that environmental messages are particularly effective when people display higher levels of political trust.

Our exploratory analysis focuses on people’s beliefs about the motivations of policymakers. Roughly 39% of the sample believe policymakers are motivated by raising funds (as opposed to reducing emissions or toxins, as suggested by the messages). This belief is significantly and negatively associated with policy approval, demonstrating that those who hold this belief are more likely to disapprove of the policy. Furthermore, we show that receiving either environmental message (E or Eh) significantly increases the probability that someone believes policymakers are acting to raise funds. We also find that those with lower levels of trust in political institutions are more likely to believe policymakers are acting to raise funds as opposed to reducing emissions/toxins.

Dasandi, N., Graham, H., Hudson, D., Jankin, S., vanHeerde-Hudson, J., Watts, N., 2022. Positive, global, and health or environment framing bolsters public support for climate policies. Communications Earth and Environment, 3(1).

Drews, S., & van den Bergh, J. C. J. M., 2016. What explains public support for climate policies? A review of empirical and experimental studies. Climate Policy, 16(7), 855–876.

Fairbrother, M., Sevä, I.J., Kulin, J., 2019. Political trust and the relationship between climate change beliefs and support for fossil fuel taxes: Evidence from a survey of 23 European countries. Global Environmental Change, 59, 102003.

Hart, P.S., Feldman, L., 2018. Would it be better to not talk about climate change? The impact of climate change and air pollution frames on support for regulating power plant emissions. Journal of Environmental Psychology, 60,pp. 1-8.

IPCC, 2022. Climate Change 2022: Impacts, Adaptation, and Vulnerability. Cambridge University Press, Cambridge, UK and New York, NY, USA.

Mossler, M.V., Bostrom, A., Kelly, R.P., Crosman, K.M., Moy, P., 2017. How does framing affect policy support for emissions mitigation? Testing the effects of ocean acidification and other carbon emissions frames. Global Environmental Change, 45, pp.63-78.

Rossa-Roccor,V., Giang, A., Kershaw, P., 2021. Framing climate change as a human health issue: enough to tip the scale in climate policy? The Lancet Planetary Health, 5(8), pp. e553-e559.

ABSTRACT. Introduction Personality nuances have received little attention in research on pro-environmental attitudes, unlike Big Five or HEXACO domains, while often only self-reports and one sample are researched. Personality traits might play a vital role in pro-environmental attitudes and behaviours, as they are stable and prevail in many situations and resulting behaviours can accumulate over more extended periods of time, just like greenhouse gases in atmosphere (Karbalaei et al., 2014; Soutter et al., 2020). Meta-analytically, domains Openness (r = .22) and Honesty-Humility (r = .20) have shown the highest correlations with pro-environmental attitudes, followed by Agreeableness, Conscientiousness and Extraversion with weaker associations (Soutter et al., 2020). For facets, Openness facets Aesthetic Appreciation, Intellectual Curiosity have correlated the highest (Brick & Lewis, 2014; Markowitz et al., 2012; Soutter & Mõttus, 2021), under Agreeableness, Sympathy has shown the stronger correlation (Soutter & Mõttus, 2021). But Agreeableness facets did not correlate with pro-environmental attitudes in Markowitz et al. (2012), therefore conclusions are not final. As associations with personality domains and facets have enhanced our understanding of pro-environmental attitudes, personality nuances could do so even further. Especially as existing research has almost exclusively relied on self-report and one sample and due to that different measurement errors and systematic biases have inflated presented correlations to unknown degrees. Combining self-reports with informant-ratings and including Estonian, Russian and English-speaking samples, we estimated true correlations between personality traits and pro-environmental attitudes, as well as pro-environmental attitudes' true predictability from domains or nuances combined and intergeneration differences in the associations.

Methods The Estonian Biobank collected the Estonian and Russian samples within the Personality Study from November 2021 to April 2022. We only included participants whose significant other, child, grandchild, friend, parent, grandparent or other chosen acquaintance agreed to participate; all participants gave informed consent. The Estonian sample consisted of 20 926 participants (14259 females, average age 45.20 years, SD = 13.69), and the Russian sample of 762 participants (532 females, average age 43.44 years, SD = 12.96). English-speaking sample was collected through Prolific Academic platform from March to June 2020, consisting of 600 (403 females, average age 28.56 years, SD = 13.19). Participants' trait ratings were collected with 198 items from 100 Nuances of Personality, and pro-environmental attitudes were assessed with six items (three in English) regarding recycling, seriousness of climate change, making sacrifices to limit climate change, supporting green-minded politicians, thinking about the long-term impact of one’s choices. Pro-environmental attitudes aggregate score was calculated as a score from the principal component analysis (PCA) of the pro-environmental items. Building upon self- and other-reported personality and self-reported pro-environmental items, a novel method of true correlation (rtrue) and true overall predictability (rpred) was used to analyse data.

Results Rtrues were higher than self-reported correlations; nuances reported larger effect sizes than domains. In the Estonian sample, domains Openness (rtrue = .35), Agreeableness (rtrue = .27) and nuances “Believe that poor deserve our sympathy” (rtrue = .62), “Believe that everyone should have a say” (rtrue =.62), “Treat all ethnicities and religions equally” (rtrue =.52), “Support liberal political candidates” (rtrue = .50), “Enjoy hurting others” (rtrue=-.50), “Would like others to be afraid of me” (rtrue =-.46) and “Can't be bothered with other's needs” (rtrue =-.46) had the highest rtrues, while the highest self-reported correlation for domain Openness was r=.30 and for nuance “Believe in the importance of art” was r=.31. For the Russian-speaking sample “Believe that everyone should have a say” (rtrue = .84) and for English-speaking sample “Support liberal political candidates” (rtrue = .76) nuances had the highest correlation. Collection of nuances with at least moderate rtrues (rpred=.61) showed the highest pro-environmental aggregate predictive accuracy, while the three highest correlating nuances (rpred=.58) had almost as good as predictive accuracy in Estonian. Collection of domains had the lowest predictive accuracy (rpred=.50).

Conclusions Personality nuances “Believe that poor deserve our sympathy”, “Believe that everyone should have a say”, “Treat all ethnicities and religions equally” and "Support liberal political candidates" associate with pro-environmental attitudes aggregate the most. Nuances outperform domains, true correlations combining self- and other-reports outperform self-reports. Also, top three personality nuances out-predict combined domains and are just slightly behind all nuances and nuances with at least moderate correlation in Estonian. Integrating other-reports with self-reports enables to view personality nuance-level associations with pro-environmental attitudes from a new perspective, allowing to capture parts not available to participant oneself but also potentially lowering all error-based measuring. The novel method of true correlation and true prediction, also looking beyond domains, suggest that pro-environmental attitudes are more influenced by personality than previously thought.

References Brick, C., & Lewis, G. J. (2014). Unearthing the "Green" Personality: Core Traits Predict Environmentally Friendly Behavior. Http://Dx.Doi.Org/10.1177/0013916514554695, 48(5), 635–658. https://doi.org/10.1177/0013916514554695 Karbalaei, S., Abdollahi, A., Momtaz, V., & Abu Talib, M. (2014). Locus of control, Neuroticism, and spirituality as predictors of waste-prevention behaviors. Ecopsychology, 6(4), 252–259. https://doi.org/10.1089/ECO.2014.0038 Markowitz, E. M., Goldberg, L. R., Ashton, M. C., & Lee, K. (2012b). Profiling the "Pro-Environmental Individual": A Personality Perspective. Journal of Personality, 80(1), 81–111. https://doi.org/10.1111/J.1467-6494.2011.00721.X Soutter, A. R. B., Bates, T. C., & Mõttus, R. (2020a). Big Five and HEXACO Personality Traits, Proenvironmental Attitudes, and Behaviors: A Meta-Analysis: Https://Doi.Org/10.1177/1745691620903019, 15(4), 913–941. https://doi.org/10.1177/1745691620903019 Soutter, A. R. B., & Mõttus, R. (2021). Big Five facets' associations with pro-environmental attitudes and behaviors. Journal of Personality, 89(2), 203–215. https://doi.org/10.1111/JOPY.12576

ABSTRACT. Introduction

Climate change is a collective action problem that can only be solved if most people act together to reduce their carbon emissions, whether by taking direct action, or by supporting pro-climate policies.

However, collective illusions (or “pluralistic ignorance”) whereby people under-estimate other people’s willingness to act can threaten collective action, because most people are conditional cooperators who choose whether to cooperate based on what they think others will do (review in Martin et al. 2024).

One under-studied feature of the climate collective action problem is that different groups in society are being asked to cooperate in different ways: urban dwellers by reducing their car use, all consumers by buying low-emissions products, farmers by changing what they produce and how, and so on. This could lead to misperceptions between groups if they do not observe or identify others’ actions as cooperation.

In this study, we investigated collective illusions across groups by comparing how farmers and the rural and urban public in Ireland see their own, their peers’, and each others’ climate concern and willingness to act. We chose these groups because in Ireland, agriculture creates more carbon emissions than any other sector, and different pro-climate policies will likely impact rural and urban dwellers differently.

Literature review

We build on two (interconnected) strands of research: research on cooperation in climate collective action problems, and research on collective illusions and misperceptions of others’ climate views.

Regarding climate cooperation, Kjeldahl & Hendricks (2018) argue that people under-estimating others’ climate worry can discourage climate collective action. Martin et al.’s (2024) systematic scoping review finds that beliefs about others and conditional reciprocity are important factors of climate cooperation, but that existing climate cooperation studies (which often use lab experiments such as public goods games) do not account for the diversity of actions required by different groups to stop climate change.

There is also evidence that people hold collective illusions about climate change perceptions. Sparkman et al. (2022) find that Americans underestimate popular climate policy support by nearly half, and Andre et al.’s (2024) survey of almost 130,000 people in 125 countries finds that people systematically under-estimate others’ willingness to act, despite widespread climate concern and willingness. However, there is little research on misperceptions in or across groups (except by political ideology, Ballew et al. 2020).

Methodology

We conducted a pre-registered survey of 1,667 adults in Ireland, including 467 farmers and 1,200 members of the general public, split evenly between rural and urban residents. Farmers were recruited via multiple channels to increase diversity and representativeness (e.g., regarding region and farm type).

The survey asked participants how concerned they are about climate change, and how concerned they think farmers and the general public in Ireland are. It also asked how willing participants are to take different kinds of actions to help reduce climate change (for farmers, this also included sustainable farming practices), and how willing they think farmers and the general public are to take these actions.

Our analysis compared farmers and the rural and urban public’s answers to these questions, first descriptively (using survey weights), then using regression models (with demographic controls and secondary models checking differences by farmer characteristic). To identify collective illusions, our models then compared how concerned and willing each group actually is (based on their answers about themselves) to how worried and willing this group is perceived to be (when asked to rate this group).

Results

We found that farmers and the public hold climate collective illusions about their peers and each other.

Regarding climate concern, all three groups (farmers, the rural public, and the urban public) reported being highly concerned, and we found no significant differences between their levels of concern. We also found no significant differences between the three groups’ ratings of the Irish public’s concern, but the public rated Irish farmers’ concern significantly lower than farmers did.

We found collective illusions in concern: the public under-estimates farmers’ actual concern, although farmers accurately perceive the public’s actual concern. In addition, both the public and farmers under-estimate their own group’s actual concern, but farmers are more accurate.

Regarding willingness to act, all three groups reported a high willingness to take action in their everyday life (e.g., driving or flying less, improving home energy efficiency). Farmers were highly willing to change how they farm (e.g., plant trees, use energy-efficient farm technologies). We found no significant differences between all three groups’ willingness to change their everyday lives, but the public rated both Irish farmers and the Irish public as significantly less likely to change than farmers did.

We found collective illusions in willingness: the public under-estimates farmers’ actual willingness, and farmers under-estimate the public’s actual willingness, although to a lesser extent. Both the public and farmers under-estimate their own group’s actual willingness, but the public is more accurate.

Discussion

The results show that people under-estimate others, and the public especially under-estimates farmers. A potential explanation for between-group illusions is that people do not see or identify other groups’ cooperation because it looks different from their own, leading to biased beliefs and negative reciprocity.

Another potential explanation for both between- and within-group illusions could be, as argued in Kjeldahl & Hendricks (2018), that a loud minority creates misperceptions of what a group’s majority actually thinks. Our survey results show some bimodality in the distributions of concern and willingness to change, with a small minority reporting that they are not at all concerned or likely to change.

This study has implications for both researching and addressing the climate crisis. Future climate cooperation studies, such as those using economic games, could account for diverse possible climate inputs rather than having all subjects cooperate via the same action, to help identify whether people respond to cooperation from other groups when it looks different from their own.

Regarding policy implications, our results suggest that correcting collective illusions, for example by emphasising shared concerns and identities (Martin et al. 2024), could help foster cooperation, including by reducing potential feelings of marginalisation among groups whose climate actions look different.

ABSTRACT. Abstract—This research paper investigates the influence of the United States Real Gross Domestic Product (GDP) on the market capitalization of companies listed in the S&P 500 index, with a specific focus on individual sectors. Employing regression analysis, I examine data spanning the years 2007 to 2021, aiming to unravel the intricate relationship between economic growth and market performance. The study reveals intriguing insights, highlighting the presence of a consistent and positive correlation within the Communications and Real Estate sectors. By analyzing the interplay between US Real GDP and sector-specific market capitalization, I shed light on the potential impacts of macroeconomic factors on distinct segments of the market. My findings hold relevance for investors, analysts, and policymakers alike, as they offer valuable perspectives on how economic indicators influence the market behavior of specific sectors. This research contributes to a deeper understanding of financial dynamics and can aid in making well-informed decisions within the realm of finance and investment.

ABSTRACT. A preference for challenging tasks is a key aspect of an individual's intrinsic motivation, indicating the degree to which people are inclined to tackle difficult tasks and persist in their efforts (Finch & Obradović, 2017). Many valuable activities or jobs that contribute to personal development involve a certain level of difficulty and can often be somewhat tedious, necessitating effort and perseverance. Consequently, individuals with a stronger preference for challenging tasks are more likely to experience success in their personal development (Yeager et al., 2014).

Previous research has demonstrated that individuals from low-income families frequently exhibit a lower preference for challenging tasks, which can lead to a sense of contentment with the status quo. This reluctance to strive for change can perpetuate their poverty, creating what is known as the "poverty trap" (e.g., Banker et al., 2020). One possible explanation for how household income influences an individual's preference for challenging tasks is its impact on implicit self-esteem. Research by Wang et al. (2021) revealed that while individuals from registered low-income families in rural China had significantly higher explicit self-esteem compared to their non-poor counterparts, their implicit self-esteem was notably lower, illustrating a phenomenon referred to as "fragile high self-esteem."

In light of this, the present study aims to investigate whether low family income reduces students' preference for challenging tasks by diminishing their implicit self-esteem. Furthermore, it seeks to evaluate the effectiveness of a self-affirmation intervention technique to mitigate the adverse effects on these students' implicit self-esteem and enhance their preference for challenging tasks.

Two studies were conducted by recruiting students from a vocational secondary college in a former impoverished county in southern Gansu Province, China. This region has a challenging physical environment and underdeveloped transportation infrastructure, contributing to its relatively low economic development. As of 2013, the poverty incidence in this county was 40.4%, making it a representative example of a typical poor population in China.

In Study 1, 84 students were recruited to participate in a survey; among them, 32 were from registered low-income families (including 21 boys and 11 boys, and 96.8% of them had a monthly living allowance of less than 1000 Yuan), compared to 52 students from non-poor families (including 38 boys and 14 girls, and 42.2% of them had a monthly living allowance of more than 1000 Yuan). We used the Rosenberg self-esteem scale to measure their explicit self-esteem and conducted an IAT test to assess participants' implicit self-esteem. To measure the preference for challenging tasks, a task was adapted from the study by Banker et al. (2020), in which participants faced five rounds of choices consisting of an easy and challenging question from the Raven reasoning test. Participants were informed that “those with a higher accuracy rate may receive additional rewards.” In each round, participants earned one point for selecting the challenging question, while choosing the easy question awarded no points. The total score from the five rounds served as an indicator of preference for challenging tasks. The results indicated that students from low-income families exhibited a reduced inclination to engage in challenging tasks (M±SD = 1.25±1.19) compared to their peers from higher-income families (M±SD = 2.12±1.23). This diminished engagement was partly attributed to their lower implicit self-esteem rather than their explicit self-esteem.

In Study 2, another 120 students from registered low-income families in the same school were randomly assigned to three groups. All participants were between 16 and 20 years old; 95.4% received a monthly living allowance of under 1000 Yuan. The experimental group, including 23 boys and 13 girls, responded to a series of self-affirmation questions aimed at boosting their implicit self-esteem; the control group, including 24 boys and 11 girls, answered questions related to their food likes; while the blank group, including 21 boys and 11 girls, was assessed without any intervention. Before and after the intervention, all groups were evaluated on their preference for challenging tasks and implicit self-esteem. The method for measuring implicit self-esteem was identical to that used in Study 1. The measurement for challenging task preference was modified from the Duckworth & Gross (2014) study, where participants were required to engage in 6 rounds of 2-back memory tasks consisting of 30 trials, respectively. During the memory task, a series of interesting and engaging short videos (averaging 30 seconds each) were shown on the opposite side of the screen. Participants were allowed to stop the memory task and watch the video anytime. If they chose to interrupt, the memory task would end. Participants were informed beforehand that their choices would not have any negative consequences and that they could act freely. To emphasize the value of the memory task, they were told that "reliable research shows that completing the memory task enhances an individual's concentration and problem-solving abilities, positively affecting future work and life." The percentage of completed trials out of the total possible (6 rounds of 30 trials = 180) indicated the preference for challenging tasks. The findings once again indicated that a brief self-affirmation intervention has a significant enough impact to improve the implicit self-esteem of these economically disadvantaged students notably. This boost in self-esteem subsequently led to an increased preference for challenging tasks among this group of students (M±SD = 0.05±0.10 in the pretest and 0.19±0.20 in the posttest for the experimental group, while M±SD = 0.06±0.10 and 0.06±0.12 for the control group and M±SD = 0.06±0.09 and 0.06±0.11 for the blank group).

The current study indicated that (1) lower family income diminished middle school students' preferences for challenging tasks; (2) this effect is partly attributed to a decrease in their implicit self-esteem; (3) implementing a self-affirmation intervention can enhance implicit self-esteem and boost their preference for challenging tasks.

ABSTRACT. The presentation focuses on the impact of emotions and sentiment in mass and media messages on perceptions of macroeconomic phenomena and decisions made in times of crises, pandemics, political polarization and other socio-economic challenges. Particular attention was paid to the emotional overtones of media messages and its ability to shape inflation forecasts, public sentiment and economic behavior. The results of the experimental research conducted indicate that emotional asymmetry - greater sensitivity of audiences to negative information than to positive information - plays a key role in cognitive distortions. These distortions can destabilize the economy, affect the decisions of consumers and investors, and limit the effectiveness of public policies. Contemporary crises, such as the COVID-19 pandemic, global financial meltdowns and the rise of political polarization, show that emotions are a powerful factor shaping the reception of information in mass media. Media messages based on emotional narratives are effective in capturing the attention of audiences, but at the same time they amplify fear, uncertainty or excessive optimism. The framing effect and emotional asymmetry are particularly evident in this context, as people tend to react more strongly to negative information than positive. As a result, the negative tone of media coverage can lead to unwarranted pessimism, which destabilizes public sentiment, increases consumer caution and reduces consumer spending. In times of dictatorship and intense political polarization, the media are sometimes used as a tool for emotional manipulation. Media narratives based on fear or anger can reinforce social divisions, undermine trust in public institutions and influence consumer and investor attitudes. Such mechanisms not only destabilize the economy, but also reduce the effectiveness of monetary and fiscal policies, which become less effective under conditions of heightened uncertainty. Understanding how the emotional tone of messages influences audience perception is therefore crucial to designing effective communication and policy strategies. A study was conducted to analyze how the emotional tone of media messages shapes inflation perceptions and economic decisions. The experiment involved 254 participants divided into two groups. The first group was presented with news with a positive emotional tone that emphasized economic stability and optimistic inflation forecasts. The second group was presented with news with a negative tone, focusing on sharp price increases and potential risks to the economy. The main objective of the study was to determine whether differences in the emotional overtones of media messages affect participants' inflation forecasts, and what cognitive mechanisms might explain these differences. Analysis of the results using the Mann-Whitney U test revealed statistically significant differences between the two groups. Those exposed to positive messages forecast lower inflation rates and were more optimistic about the overall economic situation. This group also showed a greater propensity to make consumption decisions. In contrast, participants exposed to negative comments predicted higher inflation rates and declared greater caution in economic decision-making, suggesting a greater intensity of pessimism caused by the emotional tone of the messages. The survey results confirmed the existence of emotional asymmetry in the reception of media information. The negative tone of messages exerts a stronger influence on audience perception than messages with a positive tone, which is reflected in inflation forecasts and consumer decisions. The framing effect further underscores that the way macroeconomic data is presented significantly changes the way audiences interpret it. A message constructed in an alarmist tone arouses greater anxiety, even if the economic data do not warrant such a reaction. The study's conclusions point to important practical implications. The media and institutions responsible for communicating monetary and fiscal policy should take into account that the emotional tone of media messages can influence public behavior and the stability of the economy. In particular, consciously designing media narratives toward a more neutral or constructive tone can help mitigate the negative effects of emotional asymmetry. For example, during periods of economic crises, media messages with a subdued tone can reduce consumer panic and stabilize the market. Simultaneously, the study's findings suggest that during times of dynamic socio-economic transformations, such as the COVID-19 pandemic, emotionally charged media narratives can both support and undermine the effectiveness of political actions. Depending on the tone of the message, the emotions generated by the reception of information can influence collective reactions, including social trust, levels of consumption, and investment. Therefore, understanding the mechanisms by which emotions affect audience perception should be a priority for researchers and practitioners in behavioral economics, social communication, and economic policy. However, it is important to emphasize that interpreting the study's results requires a critical approach. First, participants in the experiment may have been aware of its artificial nature, which could have influenced their reactions. Second, the study focused on the short-term effects of the emotional tone of messages, while real processes of macroeconomic perception often develop over a longer period. Third, the experiment was limited to analyzing emotional tone, neglecting factors such as source credibility, socio-political context, or participants' personal experiences. Future research could incorporate these aspects to better understand the complexity of mechanisms shaping the perception of inflation and other macroeconomic phenomena. Additionally, the use of advanced technologies, such as natural language processing (NLP), could allow for more precise tracking of the impact of media narratives on public sentiment and macroeconomic forecasts. In summary, the study's results highlight the critical influence of emotional media narratives on inflation perception and economic decision-making. The conclusions underscore the need for greater responsibility in media communication and the conscious design of narratives that support rational consumer decisions and economic stability. In times of multidimensional crises, understanding the role of emotions in media communication becomes essential for effective management of the economy and society.

ABSTRACT. Insufficient sleep is a global concern, necessitating accurate sleep measurements to understand and prevent its adverse impacts on mental and physical health. While self-reporting is a common, cost-effective method for gathering sleep data, it is prone to biases like overreporting and error-prone due to imperfect memory. This study systematically compares self-reported measures of sleep duration with device-based sleep data collected via wearable activity trackers using the online platform Prolific. Participants at Prolific, upon registering, indicate whether or not they own a wearable activity tracker (that is capable of measuring sleep data). Combining online panels with device-measured sleep data may offer widespread access to low-cost, high-validity sleep data that can help advance our knowledge about which interventions help people to sleep better. However, online panels and selection on device ownership potentially introduce other biases that are explored in this paper.

Specifically, we conducted an online panel study in three waves with N = 885 respondents through Prolific. We divided the participants into two samples: i) an unrestricted sample (UR), ii) and a sample restricted to device owners (RDO). Both samples self-reported their sleep, with respondents in the RDO sample also uploading device-measured sleep data. The primary aim is to compare self-reported and device-measured sleep data while investigating four potential sources of variation. First, demographic disparities and dropout rates between the two samples are examined to understand potential selection effects due to device ownership. Second, the study assesses the impact of mandatory data uploading on dropout. Third, it explores if informing participants that their sleep is being measured affects their behavior. Lastly, the research investigates effects associated with wearing sleep-monitoring devices (e.g., related to the feedback these devices provide).

Our main results are as follows. Respondents in the RDO sample are generally older, wealthier, and more likely to work full-time, and self-report lower sleep duration compared to the UR sample in the first wave. Dropout in the study seems strongly related to being required to upload device data, as well as to some demographics and sleep duration itself. Self-reported and device-reported sleep are not systematically different in device owners, but some differences with self-reported sleep in the general public are observed.

Several lessons can be learned from our paper. The use of online panels targeting device owners seems to be a feasible strategy to collect objective data on sleep. However, this introduces selection effects that need to be taken into account when generalizing findings. Differences between self-reported sleep and device-measured sleep are small for device owners, presumably because these participants check their devices before self-reporting sleep. However, requiring participants to upload screenshots of their device-measured sleep significantly increases dropout rates. A middle ground between objectively verified sleep data and minimizing across-wave dropout rates seems to be to ask device-owners to self-report their device measured sleep data.

ABSTRACT. This study examines whether prosocietal individuals in organizations can leverage political tactics to gain power and curb socially harmful practices – despite resistance from peers who may prefer maximizing short-term gains. Drawing on three literatures, championing, CSR/ESG, and organizational politics, we introduce the idea of prosocietal organizational politics, or Benevolent Machiavellianism. We designed a laboratory experiment in which four-person groups (N=466) repeatedly decided how much funding to take from a shared charity pool, then competed for decision-making authority in a high-stakes final round. We found widespread antisocial behavior that prosocial players failed to eliminate. Analyses further show that participants who deviate markedly from their group’s withdrawal norm garner fewer votes, suggesting that moral alignment with peers matters more than traditional influence tactics. Notably, political skill had little effect on actual behavioral outcomes, yet strongly predicted participants’ subjective perceptions of ease and confidence during persuasion attempts. Additionally, Machiavellianism and power motives were positively associated with withdrawing more resources from charities in the critical final stage. The study extends championing, CSR, and organizational politics research by demonstrating how moral context shapes persuasion processes and illuminates the complexities facing employees who seek ethical reforms from within.

ABSTRACT. The visual perception of products, places, promotions, and related objects has become central to marketplace interactions (Krishna 2012). Consumers use maps frequently, especially when navigating to a new place. Firms use schematic representations to depict the steps to be taken by the consumer in their purchase decision journey. This research investigates whether these maps and schematic diagrams impact consumers’ expectations, evaluations, and subsequent choices.

Route Angularity Effect suggests that individuals overestimate the distance and travel time containing more turns and other spatial features like landmarks, barriers, intervening points, and intersections (Raghubir and Krishna 1996, Brunye et al. 2015).

The Visual Asymmetry Effect

Asymmetrical visual stimuli are more arousing than symmetrical stimuli (Krupinski and Locher 1988) as the processing of asymmetrical stimuli requires more extensive visual exploration. Since excitement is associated with higher levels of arousal (Russell 2003), asymmetrical (vs. symmetrical) logos may be perceived as more exciting (Luffarelli et al. 2019).

Route Alignment and Perception of Adventure

Adventure tourism is associated with outdoor pursuit, escapism from ordinary life, and the feeling of success (Humberstone 2009). Activities associated with adventures (e.g., roller coaster rides, river rafting, etc.) usually have curvilinear trajectories. Evidence from diverse literature suggests that crooked paths are associated with energy, chaos, and unpredictability. In pure sciences, Brown (1827) observed the haphazard motion exhibited by matter particles when suspended in a fluid, known as Brownian motion. Einstein (1905) noted that the more thermal energy provided to the fluid, the higher is the speed and irregularity in the movement of the particles (Kac 1947). In literature, Ralph Waldo Emerson, in his essay on self-reliance (1841), emphasized that drawing “a zigzag line” depicts an attitude to transform oneself, an attitude that is not mired in the past (Tomizuka 2020). He discusses the possibility of stepping out into a new action, charting a “zigzag line” because of incalculable and unpredictable environmental influences. The organizational theory of chaos suggests that external forces of uncertainty can cause a stir in the standard, leading to new patterns of interactions and behaviors (Keyes and Benavides 2018). According to macro and financial economics, chaos is a non-linear deterministic process that looks random (Hsieh 1991).

H1: Crooked (vs. Linear) paths are perceived to lead to more adventurous destinations.

Boundary condition: According to the Visual–Verbal cognitive style dimension (e.g., Paivio, 1971), individuals may be classified as visualizers and verbalizers. Visualizers rely and learn better from imagery and visual methods of instruction when performing cognitive tasks, while verbalizers rely and learn better on verbal-analytical strategies and verbal methods of instruction (Massa and Mayer 2006).

H2: The relationship between crookedness (vs. linearity) and adventurousness (vs. serenity) is stronger for individuals who tend to process visually (vs. verbally).

Mediator: Fit

We conceptualize fit as the increased motivational intensity that results when there is a match between the path orientation and the destination reached.

H3(a): Crooked (vs. Linear) path forms a fit with an adventurous (vs. peaceful) destination. H3(b): Crooked (vs. Linear) schematic diagrammatic representation forms a fit with an adventurous (vs. peaceful) destination. H4: Fit between the path/schematic diagram type (crooked vs. linear) and the type of destination (adventurous vs. peaceful) leads to a greater likelihood of visit to the destination.

We report six experiments conducted online on Prolific among US subjects.

Study 1 (A): Association of words. The Implicit Association Test (IAT) was conducted online using iatgen (Carpenter et al. 2019). 97 participants recruited from Prolific assigned a series of stimuli into one of the two categories as fast as possible (Details in Table 1). The D score mean (M=.50251, SD=.39) was significantly greater than zero (t(96)=12.50, p<.001, d=1.27).

Study 1(B): Association of paths. 98 Participants assigned a series of paths into one of the two categories (Serene vs. Exciting) as fast as possible. The D score mean (M=.498, SD=.38) was significantly greater than zero (t(97)=12.81, p< .001, d=1.29).

Study 2: Journey to a Tourist Attraction 234 Participants were randomly assigned to a 2 (Path: Linear vs. Crooked) between-subjects design. Crooked (vs. linear) condition participants perceived the tourist attraction to be more adventurous (MCrooked=6.77, MLinear=6.25, p=.014).

Study 3: Visualize-Verbalize 185 Participants answered a 6-item scale to assess visual versus verbal information processing (Mayer and Massa, 2003). Participants were randomly assigned to a 2 (Schematic diagram: Linear vs. Crooked) between-subjects design. Spotlight analysis at one standard deviation above the mean of participants’ information processing style (IPS) scores showed a significant difference (β=.193, t(248)=2.205, p<.028), indicating that visualizers perceived the resort with a crooked (vs. linear) booking schematic diagram to be significantly more adventurous. Johnson-Neyman technique or Floodlight analysis revealed that there was a significant positive effect of the IPS on adventurousness of the resort for IPS>1 .0359 (p<.05). This is in line with the spotlight analysis.

Study 4: Vacation 168 Participants were randomly assigned to a 2(Path: Linear vs. Crooked) x 2(Vacation Need: Peaceful vs. Adventurous) between-subjects design. The fit index was high in linear (vs. crooked) path condition when participants wanted a peaceful (vs. adventurous) stay (MLinear&Peaceful=19.77, MLinear&Adventurous=18.10, MCrooked&Peaceful=16.03, MCrooked&Adventurous=18.63; F(1,167)=7.717, p=.006). Participants’ likelihood to visit the resort was higher and they preferred a healthy (vs. tasty) welcome snack when the path and the vacation need formed a fit.

Study 5: Vacation Resort 226 participants were randomly assigned to a 2(Schematic diagram: Linear vs. Crooked) x 2(Vacation Need: Peaceful vs. Adventurous) between-subjects design. Linear (vs. crooked) schematic diagram forms a fit with the need for a peaceful (vs. adventurous) vacation (MLinear&Peaceful=20.33, MCrooked&Peaceful=18.40, MLinear&Adventurous =17.33, MCrooked&Adventurous =18.33; F(1,225)=6.062, p=.015). Moderated mediation (model 7) confirmed that fit between the schematic diagram and vacation mediated the likelihood of booking the resort.

This research suggests that managers can manage expectations about the end outcomes by judiciously showing the paths to the destinations. Marketers may provide appropriately scaled route maps to increase or decrease maps' crookedness to manage consumers’ expectations of the outcomes. This effect may be extended to other marketing domains like the use of linear and crooked elements in advertising, and managing buying behavior of utilitarian and hedonic products.

ABSTRACT. Although the Political Constitution of Mexico and the General Health Law recognize mental health as a fundamental human right, the practical implementation of these protections remains severely limited. Mexico ranks second globally in terms of discrimination against individuals with mental health conditions, reflecting both structural and sociocultural barriers that impede access to adequate, dignified, and inclusive care. A major challenge lies in the allocation of resources. In 2024, only 1.3% of the total health budget was allocated to mental health care, which is insufficient to meet the growing demand for services. This underinvestment results in an annual economic loss of approximately $206 million due to untreated mental disorders, underscoring the urgent need for increased infrastructure and funding in the mental health sector. In terms of disease burden, mental health disorders accounted for 16.3% of the total disease burden in 2019. Over the past decade, the number of psychologists practicing in Mexico has nearly doubled, from 74,000 in 2013 to over 131,000 in 2021. However, the distribution of mental health professionals remains uneven, with the majority concentrated in Mexico's most populous regions, such as Mexico City and the State of Mexico, which together account for nearly one-third of the country’s practicing psychologists. This research will explore the gap between legal frameworks and the reality of mental health care in Mexico, highlighting access barriers and the need for more effective public policies to ensure equitable services. Furthermore, strategies will be discussed to improve the distribution of specialists and strengthen the mental health care system, advocating for a comprehensive, rights-based approach.

ABSTRACT. Introduction Business English teaching, despite its well-established status (Ramirez, 2015) and current profitability (Maican, 2017), is still an underdeveloped area of ESP (English for Specific Purposes) teaching in terms of a lack of coherent, solid and modern methodology (Wieczorek, 2024, Geng, 2017) that would cater for the needs of adult, (pre)intermediate-level language users who need to use English for occupational purposes. The learning objectives of such mature students often revolve around specific lexis in a given field (like for instance, English for financial analysis or for law and administration), or more generally, English language usage in business environment (Frendo, 2005), entailing intercultural and speaking skills in the context of corporate communication. Both might be an affective challenge for students who need to (re)start learning English after a long break following their school graduation and who are very often quite reluctant to participate in the process due to language-related inhibitions and bad experience in the past. On the side of the teacher, those objectives require not only their specialist knowledge concerning an appropriate sub-field of Business English Teaching and language teaching in general, but also business-related knowledge which many typical foreign language teachers may lack and which, in turn, is likely to affect their well-being in a negative way and it may be perceived as a teaching challenge (Wieczorek, 2024). Study context and methodology The presented research is qualitative in nature and set in a non-Anglophone, post-communist, context where especially Business English teaching as a part of ESP is considered an underdeveloped area (Klimanova and Pikhart (2021). The current study results therefore fill the gap, as they refer to Business English teaching to adult, intermediate students from Poland, which is one of post-communist countries. The study area is between the service provider (here: the Business English teachers) and customers (here: Business English students) within the area of Business English teaching. The main objective of the described study was to find key barriers to successful Business English instruction for adult learners in view of teachers and students. The methods applied were a case study and an interview method which have already been used in prior studies concerning business relationships (e.g. Pérez and Cambra-Fierro, 2015; Wieczorek, 2020, 2024; Wieczorek and Mitręga, 2020) in a similar context as the relationship between a BE teacher and their students (here seen as customers) is also a business-like one. The principles of Seidel’s (1998) Qualitative Data Analysis methodology were applied to analyse the qualitative material obtained. The study informants were 22 teachers teaching Business English to adult, intermediate-level corporate workers at private courses and post-graduate studies and 40 students employed at diversified business sectors and attending Business English courses or post-graduate studies. Study findings On the basis of the analysed qualitative material interesting results with reference to the affective aspect of Business English teaching emerged, especially stressing the role of well-being and classroom atmosphere as the most important factor contributing to Business English instruction especially. The results were grouped into four clusters of factors, comprising the affective states of teachers and students, the creation of positive atmosphere through the implementation of entertaining, stress-minimising activities, the frequency of updating the course programme and content, and eclectic approaches to teaching. The results not only refer to students’ emotional barriers, problems and expectations, but also to their influence on teachers’ emotions which play a crucial role in the process of teaching and learning a foreign language especially with reference to teaching for and about business at the same time. The concept of Linguistic Entrepreneurship was revisited as it is seen as strongly connected to the described teaching and learning context. The phenomenon itself refers to treating language as a resource for enhancing one’s worth in occupational settings and as a moral imperative of a good citizen and an ideal neoliberal worker (de Costa et al., 2016). Study limitations However the study, as the majority of qualitative studies, is not free from limitations (e.g. small sample, restricted context), it gives a valuable insight in many aspects of Business English teaching and it may be treated as a call for more future research in the area and a starting point for quantitative research using bigger samples and further validation. References 1. De Costa, P., Park, J. and Wee, L. (2016). Language learning as linguistic entrepreneurship: Implications for language education. The Asia-Pacific Education Researcher 25(5–6). 695–702. https://doi.org/10.1007/s40299-016-0302-5. 2. Frendo, E. (2005). Teach Business English. Pearson Education. 3. Geng, Ch. (2017). On the Teaching Innovation of Business English Teaching: A study on Multimodal Communicative Competence of Ethnic Universities. Theory and Practice in Language Studies, Vol. 7, No. 4, 322-326, DOI: http://dx.doi.org/10.17507/tpls.0704.11 4. Klimova, B.; Pikhart, M. (2021). New Advances in Second Language Acquisition Methodology in Higher Education. Educ. Sci. 2021, 11, 128. https://doi.org/10.3390/ educsci11030128 5. Maican, M. (2017). Challenges of teaching Business English in higher education. Bulletin of the TransilvaniaUniversity of Braşov Series V: Economic Sciences • Vol. 10 (59) No. 2 - 2017 6. Pérez, L. and Cambra-Fierro, J. (2015). Learning to work in asymmetric relationships: insights from the computer software industry. Supply Chain Management: An International Journal, 20(1), 1–10 7. Ramirez, C. G. (2015). English for specific purposes: Brief history and definitions. Revista de Lenguas Modernas, 23, 379-386. 8. Seidel, J. (1998). Qualitative data Analysis. The Ethnograph, vol. 4 9. Wieczorek, A.L. (2020). The Lack of Business Experience Versus The Lack of Appropriate Linguistic Background in Business English Teaching Context. In: Michalik, U., Zakrajewski, P., Sznicer, I., Stwora, A. (eds) Exploring Business Language and Culture. Second Language Learning and Teaching. Springer, Cham. https://doi.org/10.1007/978-3-030-58551-8_4 10. Wieczorek, A.L. (2020). Eclecticism as a Key Factor in Successful Business English Instruction. XLinguae. Vol. 17, Isue 4, 284-299 11. Wieczorek, A.L., Mitręga, M. (2020). Managing Prosumption in University Education—A Case Study Approach. In: Michalik, U., Zakrajewski, P., Sznicer, I., Stwora, A. (eds) Exploring Business Language and Culture. Second Language Learning and Teaching. Springer, Cham. https://doi.org/10.1007/978-3-030-58551-8_3

ABSTRACT. The Czech Republic exhibits significantly low investment participation, with only 35% of its population regularly investing and 37% lacking investment experience altogether. This research investigates the behavioral, social, and educational factors influencing this trend, addressing gaps in existing literature on investment biases within Central and Eastern Europe. Given the long-term economic and personal financial consequences of low stock market participation, understanding these barriers is crucial for improving financial literacy, trust in capital markets, and investment behaviors among Czechs.

The study is part of a three-stage research project aimed at identifying barriers to stock market participation, testing behavioral nudges to encourage investing, and measuring the long-term financial impact of these interventions. The first stage surveys 1,000 Czech participants to assess their financial literacy, investment biases, and socio-demographic characteristics. Using cluster analysis, respondents are segmented by socio-economic status and behavioral tendencies to determine which barriers disproportionately impact specific population groups. This segmentation will guide policymakers in designing effective interventions tailored to address these biases and promote capital market participation.

Historical experiences, particularly the communist era between 1948 and 1989, have played a key role in shaping Czech risk aversion and distrust toward financial institutions. The forced nationalization of assets during this period likely contributed to a collective reluctance toward financial investments, a trend observed in other post-communist nations. The findings of this study therefore have broader implications beyond the Czech Republic, offering insights into similar barriers faced by other emerging economies.

The theoretical foundation of this research is based on existing literature on behavioral finance and financial literacy, particularly in post-communist societies. Studies comparing East and West Germany highlight persistent disparities in stock market participation decades after reunification, with East Germans investing less and demonstrating home bias toward post-Soviet companies. Similar patterns emerge across CEE nations, where financial literacy remains inadequate and stock market participation low.

Despite attempts to introduce financial literacy education in Czech schools during the 2010s, a structured nationwide curriculum is still lacking. The absence of systematic financial education contributes to investment hesitancy, as citizens struggle with fundamental financial concepts such as compound interest and risk diversification. Previous studies indicate that without clear guidance, individuals often exhibit suboptimal financial behaviors, such as excessive real estate investment at the expense of diversified portfolios. This aligns with the Czech homeownership rate of 78%, far exceeding the 66% observed in the US, where a higher percentage of investors allocate capital to equities.

Investment awareness is another critical barrier. Even financially literate individuals may avoid stock markets simply because they lack exposure to investment opportunities. Additionally, social norms discourage investment when an individual’s peer group exhibits low participation, further reinforcing a cycle of disengagement. Historical exclusion from financial markets, whether due to systemic barriers or cultural norms, also perpetuates generational reluctance to invest in stocks.

Czech investors demonstrate a strong preference for real estate, suggesting behavioral biases influencing portfolio allocation. The long-standing overinvestment in property markets drives housing prices upward, making the country one of the most unaffordable places to live relative to income levels. Czech homebuyers systematically deviate from rational financial behavior due to entrenched social norms. Similar trends are observed in Lithuania, where a lack of financial literacy and trust in financial institutions leads to greater household participation in real estate over stock markets.

Low trust in financial institutions is another major obstacle. Previous studies suggest that countries with repeated financial scandals experience diminished household stock market participation. A study in Bulgaria found that public skepticism toward corporate social responsibility initiatives stems from communist-era mistrust, a phenomenon likely extending to Czech capital markets as well. These findings suggest that overcoming trust deficits is essential for increasing stock market participation.

Comparing economic transitions between Russia and China provides insight into why investment behaviors diverge across post-communist countries. China’s gradual shift toward capitalism allowed for a more stable transition, whereas Russia’s abrupt liberalization led to financial instability and a loss of public confidence in capital markets. Understanding these differences is crucial for crafting investment policies suited to the Czech context.

Although research on financial literacy and behavioral biases exists, studies specifically addressing investment nudges in post-communist societies are scarce. However, behavioral interventions such as setting default investment options have proven highly effective. Research also highlights the role of inertia in promoting long-term financial habits. Additionally, targeted interventions addressing specific biases such as panic selling and home bias could be leveraged to encourage investment participation.

This study aims to identify which demographic groups are most susceptible to these barriers and to assess the effectiveness of different nudges in addressing them. While prior research has examined gender and age-related investment biases, few studies have explored the specific constraints preventing biased groups from stock market participation in the first place. Understanding these constraints is key to designing tailored policy interventions.

The research employs a mixed-methods approach combining quantitative and qualitative analysis. The survey data undergoes factor analysis and principal component analysis to validate hypothesized investment barriers. Clustering techniques segment respondents based on their financial behavior. Logistic regression models assess the likelihood of stock market participation relative to financial literacy and behavioral biases.

The expected contribution of this study extends beyond academia. By identifying key barriers to investing, the findings can inform policymakers, financial educators, and institutions on how to foster stock market participation. Insights from this research could guide the development of targeted financial education programs and trust-building initiatives within capital markets. Moreover, as other post-communist economies face similar challenges, these findings hold broader international relevance, particularly in Eastern European nations with shared socio-economic histories.

In conclusion, this study aims to bridge the knowledge gap regarding investment participation barriers in the Czech Republic by analyzing behavioral biases, financial literacy deficiencies, and historical influences. Through empirical research, the findings will inform targeted interventions to promote informed financial decision-making, enhance economic resilience, and foster a stronger investment culture within the Czech Republic and beyond. Future research will focus on implementing and testing these nudges in real-world financial environments to measure their long-term efficacy in increasing stock market participation.

ABSTRACT. In January 2023, Brazil issued a new instrument, RendA+ (or Retirement Income Extra) and in August 2023, it issued Educa+ (Education Financing Extra) to help retail investors better save for retirement and a child’s education. These bonds were designed with very specific cashflows for the respective goals, and which combined accumulation, decumulation, inflation and compounding to overcome the typical financial literacy challenges faced by average investors globally. These bonds, purchased voluntarily on a simple two-question app, have been successful and being considered by other countries globally. The only two questions to be answered are: (a) start date of targeted cashflows (e.g., one’s retirement date or when a child would enter college) and target real amount desired (e.g., R$ 5,000 per month) from such investments, with the idea being that ALL investors, of all levels of financial literacy and income levels can answer such questions. These instruments were issued with various start dates of payments to cater to different age cohorts to target their respective goal-cash flow needs. Brazil’s Tesouro Nacional (NT) has kept very detailed records of the age of investors in the respective instruments and distributional data (along with gender, region of purchase etc.). One of the intriguing elements from a quick survey of the data is that there is a major mismatch between what specific instrument the app on NT’s website would recommend to investors based on their target retirement date or child’s date of college entrance, and the actual age/date distribution of investors over the last 1.5-2 years of daily data provides. The unanswered research question we are pursuing in the current research is whether the app is useless in educating investors, or whether investors are displaying rational behavior, exhibiting some form of behavioral bias or are purely speculating given that these are very long-duration bonds (and hence offer oversized returns if say the investor goes long duration of 50 years and interest rates drop 1% in a short span of time) or some combination of the above. We will examine the data for all start dates of each of RendA+ (8 instruments with start dates of 2030, 2035….2065) and Educa+ (17 instruments with start dates of 2030, 2031….2046), and try to discern if the portfolio choices reflect “rational”, “behavioral” or speculative decisions.

ABSTRACT. Individuals need to make careful retirement saving decisions by choosing suitable investment strategies (Debets et al., 2022; Schröder et al., 2022, 2023). Insufficient preparation for retirement is a significant risk that can result in a considerably lower standard of living in retirement (Brüggen et al., 2019; Ebbinghaus, 2021). However, individuals are often not engaged in retirement planning (Eberhardt et al., 2022), possibly due to a lack of relevant knowledge, interest, and awareness (Benartzi & Thaler, 2007; Blakstad et al., 2018; Choi et al., 2002; Prast & van Soest, 2016; Thaler & Benartzi, 2004; van Rooij et al., 2012). Financial literacy, especially knowledge about the pension system, positively influences active engagement in retirement decision-making (Almenberg & Säve-Söderbergh, 2011; Bucher-Koenen & Lusardi, 2011; Clark et al., 2010; Debets et al., 2022). Therefore, effective communication about retirement is required, equipping individuals to make considered choices. Presenting financial information using visuals (Kothakota & Kiss, 2020; Schröder et al., 2022) and interactivity (Brüggen et al., 2019; Tang et al., 2011) could be useful in increasing comprehension (Abdalla et al., 2018; Chy & Buadi, 2023; Schröder et al., 2023). However, there’s a significant need to explore the impact of data visualization and interactive design in pension communication. This study aims to assess the effect of providing easy-to-understand and interactive pension information on individual’s knowledge, attention, and pension planning behaviour. We fulfil this aim by conducting an online survey-experiment (n = 2000) and an eye-tracking experiment (n = 40) in Estonia. The experiment is based on Schröder et al., (2022; 2023) and involves four treatment groups featuring combinations of a pension navigation tool, a table of possible pension fund choices, a narrative video, and a text presentation. As such, each group (apart from the control group, which receives no additional information) receives two sets of information combined from these options. After considering the information provided, participants were asked on a five-point scale (with one indicating least agreement and five indicating most agreement): 1) Subjective knowledge: “How informed do you feel about the Estonian pension system?” 2) Sufficiency rating “Is the amount sufficient for your retirement?” 3) Change probability. What is the probability that you will make changes regarding your pension in the next month?” In addition, we measured clicking on two provided hyperlinks, one for calculating a personalised monthly pension based on government data, and another for a comprehensive overview of the Estonian pension system from an official government site. The results indicate that visual and interactive pension communication raised subjective knowledge levels compared to text and table format presentations. Therefore, our study supports the conclusions of previous experiments comparing text-based and video-based comprehension of financial concepts (e.g., Kothakota & Kiss, 2020) while also adding to the literature on retirement engagement and interactive visuals (e.g., Brüggen et. al., 2019). We have the advantage of examining both interactivity and video format pension communication in one sample, comparing each communication format to each other and a control group baseline concurrently. We also show that interactive tools can reduce cognitive load and motivate people to engage in retirement planning. While previous studies have assumed this to be the case (Kothakota & Kiss, 2020; Liverence & Franconeri, 2015), we provide eye-tracking data on the cognitive mechanism. Additionally, participants exposed to the interactive tool and video were more likely to change their current pension planning behaviour. Eye-tracking data indicates that interactive, gradually presented information, reduces mental effort and enhances pension information comprehension. For this, we rely on evidence that more complex tasks produce higher “angry” emotion (Talen & den Uyl, 2022). Participants exposed to visual material and interactive tools expend less cognitive load and complete tasks faster than their counterparts reading identical text or analysing a table. Interestingly, we found that the treatments reduced searching for additional information. We argue that this is due to the treatment providing already subjectively comprehensive information - group participants showed increased self-assessed knowledge of the pension system, i.e. feeling they are already informed enough, which reduces motivation to search for more information. Furthermore, eye-tracking data shows that the hyperlinks got more attention in the interactive pension planning tool compared to the same information presented in a table. We infer that participants were aware of the possibility of searching for more information but elected not to click. We posit that this is an indication of the visual material being easier to comprehend Our findings contribute to filling an important research gap in effective retirement communication using visuals and interactive tools (Brüggen et. al., 2019; Kothakota & Kiss, 2020; Schröder et al., 2023). These results are important considerations for policymakers implementing pension communication strategies.

ABSTRACT. This research plan aims to explore a novel approach to enhancing retirement savings by leveraging parallels between charitable giving and saving for retirement. We investigate whether what is typically seen as a psychological barrier to retirement savings – the perception of one's future self as a stranger (Hershfield et al., 2011) – could potentially be transformed into an advantage. Building on research showing that spending on others increases happiness (Dunn, Aknin, & Norton, 2008), we examine whether framing pension contributions as donations to one's future self can transfer successful interventions from the charitable giving domain to promote saving for retirement. We hypothesize that this framing might generate the same "warm glow" effect associated with charitable giving, potentially making the act of saving more immediately rewarding.

Study 1 will be conducted on Prolific (April 2025), employing a between-subjects design on U.S. and/or UK subjects. Participants are presented with a scenario asking them to decide which proportion of a windfall income to allocate to immediate consumption and retirement savings. In one condition savings are framed as standard retirement savings (control) and in another as "donating to your future self" (treatment). This design allows us to test whether the charitable giving frame increases the proportion of money allocated to retirement savings compared to traditional framing.

Contingent on the results from Study 1, we plan to conduct Studies 2 and 3 in Estonia and Latvia, respectively, implementing the experiments in local languages. This cross-cultural extension will help validate the effectiveness of the intervention across different institutional contexts, cultural settings, and language differences.

The next phase involves collaborating with pension funds and policymakers to implement field experiments, testing the intervention's effectiveness in real-world settings. For instance, reframing the call-for-action message in e-mail reminders (Reinson et al., 2025) or other communication channels.

Throughout the studies, we investigate potential heterogeneous treatment effects based on individual differences such as financial literacy, trust in financial institutions, socioeconomic status and prior giving behavior. Building on previous literature, also we examine how personality traits moderate the intervention's effectiveness. For instance, people with high levels of empathy or agreeableness have been found to save less (Gerhard et al., 2018) but give more to charity (Ottoni-Wilhelm and Vesterlund, 2023). Our framework suggests that emphasizing empathy toward the future self might be particularly effective for this population.

Additionally, we explore how different decision-making styles interact with our intervention. Prior research shows that affect-driven decision-makers tend to avoid financial planning (Park and Sela, 2018), while affect increases charitable giving. Our intervention might help bridge this gap by making retirement savings more emotionally engaging through the charitable giving frame.

This research contributes to both the retirement savings and charitable giving literature by identifying novel mechanisms to increase savings behavior and highlighting the potential for cross-pollination between these domains. Our findings have important implications for policymakers and financial institutions seeking to improve retirement savings outcomes.

References: Dunn, E. W., Aknin, L. B., & Norton, M. I. (2014). Prosocial Spending and Happiness: Using Money to Benefit Others Pays Off. Current Directions in Psychological Science, 23(1), 41–47. https://doi.org/10.1177/0963721413512503 Gerhard, P., Gladstone, J. J., & Hoffmann, A. O. I. (2018). Psychological characteristics and household savings behavior: The importance of accounting for latent heterogeneity. Journal of Economic Behavior & Organization, 148, 66–82. https://doi.org/10.1016/j.jebo.2018.02.013 Hershfield, H. E. (2011). Future self-continuity: How conceptions of the future self transform intertemporal choice. Annals of the New York Academy of Sciences, 1235(1), 30–43. https://doi.org/10.1111/j.1749-6632.2011.06201.x Ottoni-Wilhelm, M., & Vesterlund, L. (2023). Motives to give in economics and psychology: A step toward unification. Journal of Economic Behavior & Organization, 216, 354–365. https://doi.org/10.1016/j.jebo.2023.09.018 Park, J. J., & Sela, A. (2018). Not My Type: Why Affective Decision Makers Are Reluctant to Make Financial Decisions. Journal of Consumer Research, 45(2), 298–319. https://doi.org/10.1093/jcr/ucx122 Reinson, H., Post, T., Mazar, N., Reeck, C., Saulitis, A., Shah, A., Syropoulos, S., Uusberg, A. (2025). Boosting Retirement Savings Through Email Reminders: Evidence from a Nationwide Megastudy. [Manuscript in preparation]

ABSTRACT. Research on financial decision-making and pension savings has traditionally focused on consumers' need for savings and financial literacy to prevent a pension gap or financial distress (Greenberg & Hershfield, 2019). It has also examined how consumers’ lifestyles influence their financial behavior and overall financial well-being (Brüggen et al., 2017). However, little is known about individuals who adopt radical saving and investment strategies with the explicit goal of achieving financial independence. This gap in the literature cannot be attributed to a lack of practical relevance, as a growing number of people are embracing a lifestyle known as Financial Independence, Retire Early (FIRE). This movement encourages aggressive saving and investing to accumulate sufficient assets for retirement at an unusually early age (Kerr, 2020).

In an initial study, we examine individuals within the FIRE movement to determine whether achieving financial independence—typically implying the accumulation of high net worth—is a feasible strategy for the general population. Specifically, we investigate whether living comfortably off one's assets is only attainable for individuals with specialized education and higher starting wealth. Furthermore, we aim to identify the mechanisms that enable FIRE adherents to achieve financial independence. As FIRE is a cultural and community-driven movement, its members may experience different financial outcomes than their peers due to various social, psychological, and financial mechanisms (Khan & Pandey, 2023). These include social encouragement and accountability, reinforced through active participation in online forums, social media groups, and financial tracking tools that promote high savings rates and minimal spending. Additionally, the spread of financial knowledge within the FIRE community may contribute to higher financial literacy among its members compared to the general population.

The first study employs representative survey data (n = 2,000) collected in Estonia to examine the demographic, financial, and behavioral characteristics of individuals pursuing FIRE compared to those who do not. Within this sample, individuals aiming for financial independence are identified by their agreement with the statement: "I am trying to accumulate enough assets for retirement so that I can live comfortably off passive income alone."

The findings indicate that 16% of the sample is actively planning to achieve financial independence, with this share being particularly high among younger individuals. Thirty percent of participants aged 18–25 expect to become financially independent in the future. Overall, FIRE adherents tend to be younger (M = 37.31 vs. 43.57), more likely to be self-employed (21% vs. 10%), and more inclined to pursue higher education than their peers. They report significantly higher household and personal incomes, demonstrate greater financial planning skills, and exhibit both higher objective financial literacy and greater confidence in their pension plans. Additionally, individuals planning for FIRE show higher risk tolerance (M = 5.70 vs. 4.61 on a 10-point scale) and greater willingness to delay gratification (time preference: M = 5.72 vs. 4.60).

FIRE followers also display distinct asset allocation strategies and more diverse investment portfolios compared to non-FIRE individuals. They show higher participation in voluntary pension schemes (Pillar III: 36% vs. 26%), stock and bond investments (42% vs. 16%), and real estate ownership for investment purpose (28% vs. 9%). Additionally, they have significantly higher engagement in alternative investments, such as cryptocurrencies (11% vs. 3%), private equity and peer-to-peer lending (8% vs. 2%), while being less likely to hold high-interest consumer loans (12% vs. 19%). Even among the youngest FIRE planners, stock and bond ownership rates are more than twice as high as those of their non-FIRE counterparts.

The results highlight significant financial differences between FIRE followers in Estonia and their peers. Individuals striving for financial independence demonstrate higher financial literacy and superior financial planning skills. Young FIRE adherents, in particular, exhibit many of the characteristics and behaviors necessary to accumulate sufficient wealth over their lifetime. However, their unusually high exposure to private equity and real estate holdings at a young age suggests two possible explanations. On the one hand, it reflects exceptional financial management skills and an entrepreneurial mindset. On the other, it may indicate privilege stemming from generational wealth, raising important questions about accessibility and inequality within the FIRE movement.

By highlighting the significant differences between FIRE followers and the general population, our findings have important implications for the literature on FIRE, early retirement, frugality, and financial independence. Previous research on early retirement (e.g., Quinn, 2002; Disney et al., 2015) has typically focused on individuals retiring only a few years before the official retirement age. In contrast, the FIRE followers in our sample illustrate the feasibility and relevance of much earlier retirement, often decades ahead of conventional timelines. The emphasis on frugal living (Lastovicka et al., 1999; Carrigan & Szmigin, 2004) aligns closely with the behaviors observed in FIRE adherents, who prioritize savings over consumption to enhance long-term financial security. Many within the movement actively seek savings rates of 50–75%, demonstrating a commitment to aggressive wealth accumulation. Moreover, the financial behaviors of FIRE followers in our sample are consistent with established wealth accumulation models (Hubbard et al., 1994; Gourinchas & Parker, 2002) and provide new insights into how individuals actually build assets over the course of their lifetimes. Our quantitative findings on attitudes, beliefs, and asset holdings associated with the pursuit of financial independence contribute to the growing body of research on the FIRE movement by addressing key questions about its demographic profile and behavioral patterns (Khan & Pandey, 2023; Taylor & Davies, 2021).

ABSTRACT. Previous research introduced the so-called three-point shooting paradox, which suggested that professional basketball players perform better under tight defensive pressure than when shooting freely. This phenomenon was initially attributed to cognitive mechanisms described by Kahneman’s two-system model, where System 1 (intuitive and automatic thinking) allows players to perform efficiently under defensive constraints, while System 2 (deliberate and reflective thinking) disrupts performance when players have more time to consider their actions. The argument was that defensive pressure forces reliance on automaticity, reducing overthinking and minimizing self-imposed cognitive disruptions. When players operate under System 1, their responses are quick, efficient, and rooted in extensive practice, allowing them to act without being slowed by conscious decision-making. Conversely, when given the extra time in an open-shot scenario, System 2 may become engaged, prompting analytical thought processes that can introduce doubt, hesitation, or a disruption of ingrained motor patterns. This transition from effortless, practiced motion to conscious deliberation has been associated with performance breakdowns in various high-skill activities, including basketball shooting. However, our large-scale replication study challenges this notion by analyzing 781,663 three-point shot attempts across 11 NBA seasons. Unlike the original study, which was based on 382 shots assessed subjectively by expert coaches, our analysis utilizes newly available NBA data that objectively categorize shots based on defender proximity: very tight (0–2 feet), tight (2–4 feet), open (4–6 feet), and wide open (6+ feet). This dataset allows for a more precise examination of the relationship between defensive pressure and shooting accuracy. Our findings strongly contradict the three-point shooting paradox. The results clearly show that as defensive pressure increases, shooting accuracy decreases. Across all seasons examined, three-point percentages were lowest in very tight (25.1%) and tight (29.1%) defensive conditions, while improving significantly in open (34.5%) and wide-open (38.5%) situations. In summary, our study refutes the existence of the three-point shooting paradox and reinforces the intuitive conclusion that defensive pressure is detrimental to shooting accuracy. This highlights the importance of large-scale, objective data analysis in behavioral research, ensuring that conclusions are not drawn from limited or potentially biased samples.

ABSTRACT. We examine if the salience of 'us-versus-them' affects cricket players' performance using One Day International (ODI) matches. Before 1979, ODIs were played in white jerseys regardless of team representation. After 2000, each team wore a unique coloured jersey. Between 1979-2000, ODIs featured both white and coloured jerseys, making cricket uniquely suitable to study our research question. We analyze players from this period and show they performed differently when opponents wore the same white jersey versus a different coloured jersey. Bowlers became more aggressive, and batters more cautious when opponents wore different colours.

Team sports are almost always played in coloured jerseys, making team representation clear. ODI cricket was unique in allowing both white and coloured jerseys between 1979-2000.[1] We hypothesize that when opponents wear different colours, 'us-versus-them' becomes salient, influencing performance. We examine ODI players from 1979-2000 and demonstrate that this salience significantly affects their performance.

Jersey colour decisions were made centrally by cricket associations, independent of player selection, eliminating self-selection bias.[2] Players did not choose matches based on jersey colour but simply represented their country. This exogeneity allows us to treat our study period as a natural experiment, addressing the rationality of performance differences based on jersey colour.

We focus on 140 players who played at least 10% of their ODIs in both conditions. Our dataset, containing over 17,000 observations (11,214 for batters and 6,308 for bowlers), was compiled from Wisden, Cricinfo, ESPN, YouTube, news channels, and the International Cricket Council. We assess differences in key performance indicators for batters (runs, strike-rate, batting-average, centuries, fifties) and bowlers (bowling-average, wickets, economy-rate) in white versus coloured jerseys.

We find that bowlers perform better and batters perform worse in coloured jersey matches. These results persist after controlling for age, player fixed-effects, country and opponent dummies, and time-effects. Players perform differently when the opponent is visually distinct. With two batters versus eleven opponents on the field, the visibility of opponent colours affects perception. Bowlers have a better economy-rate and engage in more aggressive bowling when batters look different. Batters, in contrast, score fewer runs and have a lower strike rate in coloured jerseys, suggesting a cautious approach. A white jersey environment allows batters to play more freely without the mental burden of distinction.

This suggests a departure from rationality, as players should perform consistently regardless of opponent jersey colour.[3]

[1] The first coloured jersey ODI was played on November 27, 1979 and the last white jersey ODI was played on December 14, 2000.

[2] Sourav Ganguly, who debuted ODI cricket in 1992 recollects his helplessness in getting selected to the team in the years 1996-2000 (Ganguly, 2018) and says, “Yes, I was at the mercy of a selection committee for every game.” The point here is that even reasonably established players had little control on their own chances of selection, let alone self-selecting themselves according to whether tournaments were played in coloured or white jerseys.

[3] Note that each country has specific designers for designing the jersey. Also, each country has a specified Cricket Council which selects the players who play an ODI. So, while the selectors select the team, designers select the colour of the jersey. Selection committees have better knowledge about cricket relative to ones selecting jersey colours. For instance, designer Fiona Clarke designs jerseys for the Australian cricket team and the National Selection Panel selects players for Australian national team – these organizations work independently. Jersey designers do not necessarily need an understanding of cricket (and consequently have little to do with the selection processes).

ABSTRACT. Introduction

The Saudi cultural sector is undergoing a transformative period, marked by significant investments to preserve and promote its rich heritage while fostering a vibrant cultural ecosystem. Increasing public engagement with cultural activities is a cornerstone of these efforts, as it not only drives demand for cultural offerings but also incentivizes sustainable growth in cultural supply. The Ministry of Culture (MoC) has recently started to assess demand for cultural activities to identify potential barriers to cultural participation (MoC, 2023, 2024). However, the mechanisms by which participation in cultural activities influences future demand for new activities from the same (vs. other) cultural sub-sector remain insufficiently understood. While existing research highlights the role of familiarity and exposure in shaping cultural preferences (e.g., Kolhede & Gomez-Arias, 2017; Thorsby et al., 2024; Váradi & Józsa, 2023), there is limited empirical evidence quantifying these effects in the Saudi context, particularly across distinct sub-sectors such as music, museums, archaeological sites, and theater. Addressing this gap is critical for designing targeted interventions that maximize the impact of cultural campaigns and policies to promote cultural participation in the Kingdom. In this regard, the MoC conducted two natural experiments aimed to test the impact that participation in cultural activities have on people’s attractiveness towards other cultural activities. More specifically, the two experiments were designed to explore 1) whether prior participation in a cultural sub-sector would increase the attractiveness of new activities within the same sub-sector (i.e., attractiveness effect); and 2) whether prior participation in one sub-sector would also influence the attractiveness of activities in another sub-sector (i.e., spillover effect).

Methodology

To answer these questions, the two natural experiments used quotas to build three groups with different levels of previous participation in cultural activities from the selected sub-sectors (no previous participation in any sub-sector; participation in sub-sector 1 but not in sub-sector 2; and participation in sub-sector 2 but not in sub-sector 1), balanced for gender and aged 18-70 years old. Participants (Total N=1288) completed a structured survey in which they were randomly shown two vignettes of fictitious events from two different sub-sectors (Experiment 1: music and museums sub-sectors; Experiment 2: theater and heritage sub-sectors). After each vignette, participants were asked to evaluate the attractiveness of the fictious event. Ratings included specific interest, likelihood to attend, expected satisfaction, and likelihood to recommend. An aggregate measure of event attractiveness was computed by averaging these four metrics, rescaled to a uniform 7-point scale . Vignettes were modeled after MoC’s promotional posts, ensuring consistent stimuli presentation. The experiments were approved by ethics committee from the GfeW (German Association for Experimental Economic Research, under IRB Certificates No. pC6JsSDn and No. cwLhFLpX).

Results

In both experiments, independent samples t-tests (two-sided) and robustness checks using linear regression models were conducted to assess the effects of prior participation on the perceived attractiveness of new cultural activities, while controlling for general interest.

Attractiveness effect

Results from the t-tests supported the attractiveness effect. People are 21.3% to 24.5% more attracted to new cultural activities after they had participated in activities in the same sub-sector (all ps <0.01). In the first experiment, linear regressions showed the results remain robust after controlling for general interest in museums or music concerts: previous participation in cultural activities in one sub-sector directly impacts the attractiveness of new activities in the same sub-sector. On the contrary, the same linear regression models in Experiment 2 showed that the impact of previous participation on the attractiveness of new cultural activities works indirectly: participation in activities from one sub-sector increases interest in activities from this sub-sector in general, which, in turn, further increases the attrativeness of new activities in this sub-sector.

Spillover effect

Experiment 1 results from the t-tests supported the spillover effect showing that people are also more 4.6% to 8.6% attracted to new cultural activities after they had participated in activities in another sub-sector (all ps < 0.01). Linear regression models showed that the impact of previous participation on the attractiveness of new cultural activities in another sub-sector works indirectly: people who previously participated in activities from one sub-sector become more interested in the other sub-sector, in general, which in turn increases the attractiveness of new activities in this other sub-sector. In Experiment 2 the spillover effect was found only for one of the sub-sectors: people who previously visited archeological sites rated the fictitious theater play 4.2% more attractive than people who did visit archeological sites before (p < 0.01). The fact that previous attendance to theater plays did impact the attractiveness of the fictitious archeological site suggest that the spillover effects found might not generalize to all sub-sectors. Linear regression models showed that the impact of previous participation on the attractiveness of new cultural activities in another sub-sector works indirectly: people who previously visited archeological sites become more interested in theater plays in general, which, in turn, further increases the attractiveness of new activities in these sub-sector.

Conclusion

The findings from both experiments provide valuable insights into the behavioral dynamics of cultural participation showing that prior cultural participation fosters both direct and indirect growth in cultural demand. Stronger effects were consistently observed within the same sub-sector, while spillover effects across sub-sectors were less pronounced and not for all the selected sub-sectors. Although the results are based on natural experiments, in which randomization was not possible, in our experiment temporal precedence is clearly established, and the robustness checks conducted account for potential confounding variables and provide strong statistical evidence that there is indeed a causal relationship between cultural participation and increased attractiveness of other cultural activities. These experiments are useful tools to better understand cultural demand and to support the development of targeted strategies to enhance cultural participation. These strategies might focus on promoting cultural activities among consumers who attend activities in the same subsector and leveraging spillover effects to promote interest in activities across different subsector.

ABSTRACT. Introduction

Handicrafts are fundamental to Saudi Arabia's cultural heritage, representing centuries of artistry and craftsmanship. Declaring 2025 as The Year of Handicrafts by the Saudi Council of Ministers underscores the nation's commitment to preserving and promoting this vital cultural element. Through events and exhibitions led by the Saudi Ministry of Culture (MoC), Saudi artists are showcasing their creativity, connecting the past, present, and future, and emphasizing the global and national variability of handicrafts. This celebration also highlights the importance of the untold stories behind these artifacts, which are integral to the Saudi identity and a source of national pride (Arab News, 2025). Despite this national commitment to support the crafts sector, the rapid development of technology and automated production is significantly impacting the sector. While automated production might help to increase productivity and innovation in the sector, by enabling faster and more cost-effective production of handicrafts, it also challenges the future of the traditional human-made handicrafts – especially if people fail to recognize the unique value of these human-crafted items. In this vein, it is crucial to understand how people value human-made versus machine-made handicrafts in order to preserve the cultural and artistic significance of traditional craftsmanship. To answer this question, we conducted an experiment aimed to test how people value handicrafts and how much this perception changes when they are explicitly informed the handicrafts were made by a human artist or by a machine. Building on this previous research on the impact of automation on perceived value of creative work (Khokar, 2024; MoC, 2024), we hypothesize that knowing the origin of the crafts would make participants more in favor of buying the hand-made ones (human favoritism hypothesis), and less willing-to-buy the machine-made crafts (machine discrimination hypothesis).

Methodology

To test these hypotheses, the MoC Behavioral Insights Unit conducted a survey-based experiment, with 2 (information about the origin: with vs. without) x 2 (origin: human- vs. machine-made handicraft) between-subjects design. The experiment was conducted between July 4- August 8-2024, and targeted a sample of 1,240 respondents (310 per group), proportionally balanced by gender (50% female, 50% male) and spanning an age range of 18–70 years. Participants were randomly intercepted in public places across Riyadh, Saudi Arabia, and completed a structured survey designed to measure perceptions and willingness-to-buy the crafts. At the beginning of the experiment participants were given a small sample of a handicraft rug (either hand-made or machine-made) and were informed (or not) about its origin (either machine-made or human-made). After touching the sample with their own hands, participants were asked to state their evaluation of/willingness to buy the handicraft . Ethical considerations were strictly adhered to, ensuring anonymity of responses and obtaining general consent from participants prior to data collection. The experiment was approved by ethics committee from the GfeW (German Association for Experimental Economic Research, IRB Certificate No. 96czKtRt by GfeW).

Results Independent sample t-tests were used to test the impact of informing participants about the origin of the handicraft sample. In line with the human favoritism hypothesis, being informed that a human artist created the handicraft increased in 4.7% participants’ evaluation/willingness to buy the product, compared to not knowing its origin (t(618) = 3.78, p < 0.01). On the contrary, no support was found for the machine discrimination hypothesis: being informed that the handicraft was machine-made did not significantly change participants’ evaluation of the handicraft, compared not knowing its origin (t(623) = 1.21, p = .89) Overall, these findings underscore the importance of informing consumers about the handicrafts' origin in enhancing consumers’ evaluation of/willingness to buy hand-made products.

Conclusion

This research aimed to contribute to a deeper understanding of consumers’ preferences towards products at the intersection of traditional craftsmanship and technological innovation. Additionally, it might also contribute to inform cultural policies aimed at protecting and fostering the appreciation for traditional hand-crafted products in the modern world. In this regard, to ensure that hand-made product are properly valued, policymakers should consider the following measures:

• Labeling regulations: Introduce guidelines for identifying hand-made products, promoting the inclusion of detailed information about their origin, production processes, and the craftsmanship involved. This approach will emphasize the authenticity, quality, uniqueness, and cultural significance of artisanal goods. • Educational campaigns: Initiate public awareness programs to enhance understanding and appreciation of traditional craftsmanship by highlighting the value of human creations. • Marketing campaigns: Conduct marketing campaign that highlight distinctive qualities of hand-made products, emphasizing their authenticity, superior quality, uniqueness, and the cultural importance of handicrafts in preserving Saudi cultural heritage.

By aligning with Saudi Vision 2030's goals of cultural preservation and economic diversification, these measures are also a timely contribution to the nation’s cultural agenda.

ABSTRACT. Why Don’t Donors Deduct? Behavioral Barriers to Tax Incentives in Charitable Giving Michael Hilweg-Waldeck & Argun Aman Hild University of Mannheim & ZEW Mannheim Abstract

Governments worldwide provide tax deductions for charitable donations to incentivize giving. However, empirical evidence reveals a striking discrepancy in deduction behavior: while 80% of online donors claim tax deductions, only 0.5% of donors in field settings—where donation and deduction choices are observable—do so. This gap raises fundamental questions about the behavioral drivers of tax-deduction decisions. Leveraging a combination of empirical data, survey insights, and experimental evidence, we examine whether social image concerns and information deficits drive this pattern. In two large-scale experiments—one conducted online in an anonymous setting (n=483) and the other in a socially observable field setting (n=6,728)—we confirm that deduction rates remain high in anonymity (87%) but collapse in observable environments (0.5%). Surprisingly, neither information provision nor belief correction treatments significantly increase deduction rates in either setting. Ongoing work includes a second field experiment with improved timing of interventions and a lab study that isolates the causal effect of observability. Our findings challenge standard economic assumptions about tax incentives and highlight the importance of social context in financial decision-making. 1. Introduction

Charitable giving is often framed as an act of pure altruism, yet extensive research suggests that social image concerns and strategic motivations also play a role. To encourage donations, many governments provide tax deductions, effectively reducing the cost of giving. However, donors’ willingness to claim these deductions varies substantially across settings.

Empirical data from a major Austrian charity reveal a stark behavioral contrast: while 80% of donors in anonymous online settings deduct their donations, only 0.5% of donors in face-to-face interactions claim deductions. This discrepancy suggests that factors beyond financial incentives—such as social norms, image concerns, and beliefs about moral appropriateness—may influence deduction behavior.

To systematically investigate this issue, we conducted two controlled experiments that varied observability while keeping the tax benefits constant. Importantly, the procedural steps for claiming tax deductions are identical across settings, meaning that the stark differences in deduction behavior cannot be attributed to variations in administrative complexity or accessibility. Additionally, survey data from a representative Austrian sample (n=314) indicate that many donors misunderstand the deduction process and hold misperceptions about its moral acceptability. To address these concerns, we implemented two interventions: an information treatment explaining the deduction procedure and a belief correction treatment providing information about Austrians’ moral perceptions of tax deductions. 2. Experimental Design

We designed two experiments to isolate the role of observability in deduction decisions. The first, an online experiment with 483 participants, allowed individuals to earn money through a real-effort task and donate to real charities in an anonymous setting. The second, a field experiment conducted with 6,728 participants, took place within Austria’s largest door-to-door charity campaign, where deduction decisions were socially observable. To assess the impact of knowledge and beliefs, we introduced two treatments: one providing detailed information on how tax deductions work and another correcting misperceptions about the moral perception of deducting donations in Austria. The control group received no additional information. 3. Results 3.1 Empirical Patterns & Experimental Validation

Both observational and experimental data confirm a stark difference in deduction behavior between settings. In the online experiment, 87% of donors claimed deductions, whereas in the field experiment, only 0.5% did so. This supports the notion that observability plays a crucial role in deduction decisions. 3.2 Treatment Effects

Despite addressing both informational gaps and moral perceptions, neither intervention significantly altered deduction rates in either setting (p > 0.1 in all cases). These findings suggest that social image concerns, rather than misinformation or uncertainty, drive the reluctance to claim deductions in public settings. 4. Ongoing Work & Next Steps

To further explore these mechanisms, we are conducting two additional studies. A second field experiment refines the timing of information delivery, ensuring that the intervention occurs immediately before donation decisions rather than weeks in advance. As part of this intervention, we employ large-scale radio advertisements as a treatment medium to disseminate information broadly and assess its impact on deduction behavior. The data have been collected and are currently being processed by our charity partner. Additionally, we are conducting a controlled lab experiment that manipulates the observability of deduction decisions to establish causality more precisely. Data collection is ongoing, with results expected by February 2025. 5. Conclusion

Our findings reveal that the decision to claim tax deductions is highly context-dependent. While donors in anonymous settings freely claim deductions, social interaction and observability dramatically reduce deduction rates. The failure of information and belief correction treatments to alter behavior suggests that social norms and signaling concerns override financial incentives. Our lab experiment will provide a stronger causal test of observability’s role in deduction behavior and further disentangle the mechanisms underlying social image concerns. These insights have implications for the design of tax policies intended to encourage charitable giving, particularly in public donation settings.

ABSTRACT. Tax evasion in Brazil reaches significant levels. Data from the Brazilian Institute of Planning and Taxation (2023) indicate that the country loses approximately USD 70 billion annually due to unreported corporate revenues, representing nearly 12% of the total tax revenue collected by public entities (National Treasury, 2024). The concealment of income sources is so deeply ingrained in Brazilian society that the publication of Regulatory Instruction RFB No. 2219/2024—intended to incorporate new digital tools into the Federal Revenue Service’s monitoring of financial institutions to identify transactions potentially linked to financial crimes (Brazil, 2024)—sparked speculative fears that the Brazilian government would impose taxes on financial transactions exceeding USD 1,000. The apprehension over this potential new tax was so significant that there was a 15% reduction in financial transactions via PIX, an instant payment system included in the regulatory instruction (Queiroz, 2025), ultimately leading the government to withdraw the initiative (Abel & Barcellos, 2025). Given this context, this study aims to analyze how the Tax Education Program (PEF-M), promoted by the Finance Department of a Brazilian municipality, impacts the financial citizenship of children and adolescents. The program encourages discussions on public finance in local schools, using interactive methods to highlight the importance of taxation for fiscal justice and the financing of public services. This initiative is proposed to enhance young individuals' fiscal awareness by demystifying tax administration, demonstrating the role of taxes in economic and social development, and ultimately discouraging tax evasion behaviors in the future.

Tax Education and Financial Citizenship Tax education encompasses awareness, understanding, and familiarity with tax laws, policies, and regulations. It also extends to the ability to apply tax legislation to specific situations, enabling individuals to make informed decisions regarding tax matters (Matshona et al., 2024). Research suggests that, to achieve a more efficient tax system, initiatives should be implemented to enhance taxpayers' fiscal knowledge and reduce the complexity of the tax system. Such actions appear to foster a positive perception of tax fairness among taxpayers, playing a crucial role in shaping their behavior (Bhalla et al., 2022; Latif et al., 2023). Despite evidence that tax education has positive effects on children's awareness (Geyik et al., 2024), this target group has been largely overlooked in the literature. Greenham and colleagues (2024) recommend that tax-related values be instilled in future taxpayers before they leave school, given that tax education fosters an understanding of the connection between taxes and public services, raising children's awareness of the purpose behind tax payments (Geyik et al., 2024). In this context, the PEF-M aims to stimulate civic engagement and foster a harmonious relationship between citizens and the municipal government by raising awareness of the socioeconomic role of taxation (Fortaleza, 2011). The challenge, therefore, is to position tax education as a tool for promoting financial citizenship, which can be understood as the ability to comprehend, engage with, and take responsibility for government decisions related to state financial governance (Riles, 2018).

Method This study used as data 982 essays written by students aged 14 to 17 enrolled in basic education schools. These essays were produced as part of the 2023 edition of the Sefin Award for Municipal Public Finance, a school competition aimed at promoting tax education, with the thematic approach: "taxation as a tool for promoting environmental sustainability" (Sefin, 2023). To ensure sample randomness, 277 essays were randomly selected for the study’s empirical analysis, avoiding selection bias and maintaining sample representativeness (Taherdoost, 2017). The use of writing as a basis for reproducing social meanings is supported by Tonkiss (2004). Through these essays, content analysis was applied—a process that considers both participants and context, classifying texts into categorical groups to identify similarities and differences, patterns, and associations, both explicit and latent (Kleinheksel et al., 2020).

Results Based on the objectives of the PEF-M, four categories were identified: State Financing, Tax Universalization, Social Control, and Effectiveness of Rights. The essays were analyzed and categorized according to their content. Since the texts were produced after the tax education activities in each school, the students' written accounts made it possible to verify that the program has achieved its intended outcomes. The classes, lectures, and seminars have contributed to the development of financial citizenship among children and adolescents. Several excerpts highlight that the PEF-M initiatives are reinforcing its four fundamental pillars: "Taxes are a means of raising revenue through which the State finances its activities and services provided to the population" (State Financing); "Fiscal policy must achieve economic, human, and environmental harmony, establishing a societal code of conduct to ensure a balanced environment in the collective interest" (Tax Universalization); "With taxes, there is no shortage of money, but the real issue is how these resources are being allocated" (Social Controle); and "Taxation, as a tool for promoting environmental sustainability, plays a fundamental role in the pursuit of a more balanced and ecologically conscious future" (Effectiveness of Rights). It is also noteworthy that the initiatives were not sufficient to completely eliminate negative perceptions regarding the mismanagement of public funds among children: "But part of this money is misused in the wrong way, and another part goes into the pockets of corrupt politicians."

Conclusions This study aimed to assess whether the activities of the PEF-M had a positive impact on promoting financial citizenship among children and adolescents. Through the written accounts of participating students, it was observed that the program has successfully achieved its objectives. The young participants internalized the importance of fiscal policy for state financing and the effectiveness of rights. Moreover, they recognized that social control is an essential tool to ensure the universalization of taxation and the responsible use of public funds by government officials. Thus, it is evident that tax education policies serve as a crucial instrument in fostering financial citizenship among children by raising awareness of the socioeconomic role of taxation.

ABSTRACT. This conceptual paper examines the interaction between small business (SME) taxpayers and their accountants, investigating a potential conflict of interest in the presence of tax aversion. Empirical data will be collected in the near future via a self-report survey to validate our hypotheses. We aim to inspire empirical research into tax aversion, a still underexplored behavioural phenomenon (Fochmann & Kleinstück, 2012), its drivers and consequences. Additionally, we address two other gaps in the literature by investigating behavioural drivers of accountants and their advice, and investigating client-accountant interactions and their conflicts of interest, responding to the research calls of Kang (2018) and Radulović and Savić (2024). Finally, we also contribute to the general field of corporate tax research. Taxation is a complex matter (Kirchler, 2007), business taxes even more so, and many SME owners feel their tax knowledge is limited (Ahmed & Braithwaite, 2005; Kamleitner et al., 2012). Therefore, they rely heavily on accountants (Everaert et al., 2007; Webley & Ashby, 2010). Their demand for accountants is primarily driven by the wish to file a correct return (Collins et al., 1992; Hite & McGill, 1992; Niemirowski & Wearing, 2003). In turn, accountants perceive a demand for aggressive advice from their clients (Tan, 1999) while this advice drives SME owners’ commitment to their accountant (Tan et al., 2016). Accountants seem to have good reason to believe they should help their clients save on taxes. The common saying that “good accountants pay for themselves” supports this view. While contemporary literature acknowledges that the standard rational taxpayer (Allingham & Sandmo, 1972) does not exist, we still largely rely on the image of an econ accountant (Thaler & Sunstein, 2008). However, it is unlikely that accountants are immune to the behavioural flukes that influence their human clients. Behavioural principles such as fairness perceptions (Farrar et al., 2019; Farrar et al., 2020) and emotions (Enachescu et al., 2019; Olsen et al., 2018; Privitera et al., 2021), as well as other personal characteristics, have been found to impact tax behaviour. Through executives, these characteristics also influence corporate tax behaviour (Dyreng et al., 2010; Hambrick & Mason, 1984; Hanlon et al., 2022). Might the same logic apply to SME owners’ most trusted advisor? Yet, little research examines how non-economic influences on accountants, through their advice, shape SME’s tax behaviour. An underexplored, intriguing and somewhat elusive behavioural tax phenomenon, is tax aversion (or tax aversion bias), which refers to the excessive disutility of taxes compared to economically equivalent payments (Blaufus & Möhlmann, 2014). It influences behaviour in atypical ways (Sussman & Olivola, 2011) and elicits a desire to avoid taxes, to a point beyond the rational economic motivation to avoid a monetary cost (McCaffery, 1993). This bias resembles loss aversion (Kahneman & Tversky, 1979) but extends further, increasing peoples’ disutility of paying taxes, even if avoidance or evasion is not pursued at all costs. For society, tax aversion might lower compliance rates and increase enforcement costs due to tax averse individuals taking on more risk (Fennell & Fennell, 2003). A specific manifestation of tax aversion is the “tax shelterer’s fallacy” McCaffery (1993), which leans towards the gambler’s fallacy. It leads people to focus excessively on tax benefits, disregarding the total costs associated with tax avoidance. Inspired by the concept of myopic loss aversion (Benartzi & Thaler, 1995), we term this behaviour “myopic tax sheltering”. Accountants’ role here is intriguing. On one hand, accountants recognising their clients' tax aversion, should help avoid irrational myopic tax sheltering decisions. However, this also presents an opportunity to exploit their client’s tax aversion bias to their advantage. Remember that accountants often believe clients favour aggressive advice (Tan et al., 2016). This might explain accountants’ puzzling, anecdotally observed advice to clients, still undocumented in literature, to “increase (business) costs”. However sensible for costs that genuinely benefit the business, costs purely created to generate tax benefits just exemplify myopic tax sheltering. This raises the question whether this advice reflects a principal-agent conflict, where accountants exploit information asymmetry for their benefit. If accountants shift their tax-averse clients’ focus solely to tax savings, they can create the simplistic impression of valuable tax-saving advice, which may justify their fees. However, if accountants themselves fall victim to the tax shelterer’s fallacy, due to their own tax aversion, this advice seems to be given in good faith. Could even professionals such as accountants be susceptible to this bias? After all they are taxpayers themselves. Their beliefs about the fairness of the tax system, and their emotions, may influence their advice. We face some major challenges in our research though. Tax aversion, its measurement, drivers and consequences are yet uncharted (Fennell & Fennell, 2003). It is even unclear whether behaviours attributed to tax aversion stem from genuine dislike of taxes, or are simply a consequence of tax complexity, miscalculations, or imprecise heuristics (Fochmann & Kleinstück, 2012). However, if we observe a systematic bias leading people to overvalue the true cost of taxes, it must have underlying causes that influence tax behaviour. If tax aversion occurs, it likely mediates the relationship between known drivers and tax behaviour itself. To initially explore of drivers of tax aversion, we focus on two known drivers of tax behaviour, fairness perceptions and emotions, alongside personal characteristics like age, gender, education, and professional experience. We hypothesize that accountants' tax aversion is positively related to the degree of myopic tax sheltering advice they provide and expect this relationship to be stronger when accountants perceive their clients to have higher tax-saving expectations. We predict that these client expectations moderate the impact of behavioural tax drivers, whether or not mediated by tax aversion, on myopic tax sheltering advice. Tax-averse accountants may believe they are offering valuable service by providing myopic tax sheltering advice, but what about the other accountants? Are accountants suffering from tax aversion-induced fallacies themselves, or are they merely “Phishing” their SME clients “for Phools” (Akerlof & Shiller, 2015)?

ABSTRACT. Tax evasion and shadow economy activities vary significantly across Europe, yet our understanding of how these variations affect survey responses about illegal economic behaviors remains limited, especially in diverse tax morale settings. This study addresses this knowledge gap by investigating the impact of tax compliance cultures on social desirability bias in surveys about tax evasion. Focusing on Denmark, Italy, and Latvia – countries with distinct combinations of tax morale and shadow economy sizes – the research employs an innovative design that combines list experiments with vignettes.

By engaging 6,600 participants across these three countries, the study examines how social norms and tax morale influence self-reported tax evasion behaviors. Participants were randomly assigned to control and treatment groups, with the treatment group receiving an additional sensitive item about tax evasion in their list. A placebo test using a non-sensitive item (house purchase) was also conducted to validate the method. Additionally, the study incorporated belief updating about tax morale norms to investigate their impact on tax evasion disclosure and tax morality.

While the list experimental results do not yield conclusive evidence of social desirability bias, the inconsistencies in list experiment responses for sensitive questions suggest that social desirability bias may still subtly influence how people report illegal behaviors. The findings also underscore the importance of considering how information about social norms, as employed in the vignettes, can influence individual attitudes towards tax compliance. Notably, when participants’ beliefs about the prevalence of tax evasion acceptance were updated to reflect a lower acceptance rate than previously assumed, it correlated with a decrease in tax morale. This study thus contributes valuable insights to our understanding of tax morale formation and highlights the potential unintended consequences of norm-based interventions in tax policy.

ABSTRACT. Introduction The commuter tax allowance (‘Entfernungspauschale‘ or ‘Pendlerpauschale‘) in Germany allows taxpayers to deduct commuting costs from their taxable income —€0.30/km for the first 20 km and €0.38/km beyond—reducing taxable income if it exceeds the €1,230 flat-rate deduction for professional expenses. It is a fiscal policy tool that lowers Germany's total tax revenue by €5–6 billion annually (Spiegel Online 2021, Bundesrechnungshof 2022). The commuter tax allowance has been the subject of criticism for several decades (see UBA 2021, Agora Verkehrswende 2022). One source of criticism is that the regulation applies to all modes of transport and thus has no steering effect on the use of sustainable modes of transport such as public transport. Further, higher commuting subsidies incentivize an increase in the length of the commute (Paetzold 2019). Additionally, the allowance tends to disproportionately benefit high-income groups, as tax savings increase with income (Rinscheid & Busemeyer 2023), while low-income taxpayers, often receive little to no benefit and are less likely to file tax returns (Hauck & Wallossek 2024). Despite the public debate, yet little attention has been given to how individuals perceive the regulation of the commuter tax allowance. Therefore, we conducted a survey with 3,055 participants in Germany to assess their beliefs about the current regulation in terms of personal financial interests, distributional aspects and environmental impact, as well as their knowledge of the commuter tax allowance and their support for a reform. We investigate the relevance of the environmental dimension and whether individuals are more likely to support reforms that target distributional aspects than those that increase personal financial benefits. In addition, participants were provided with information about either the distributional or environmental dimension of the current regulation. We examine how this information influences participants' beliefs about the policy and their support for several reform options. Rinscheid & Busemeyer (2023) find that even minimal information about these aspects has a strong impact on support for a reform in this context. Following them, we extend their findings by assessing individuals' prior knowledge and underlying beliefs to better understand the mechanisms driving these treatment effects. Understanding how information provision influences public acceptance is important for designing effective reforms. In addition, policymakers are unlikely to pursue reforms without widespread acceptance, making it relevant to understand public attitudes before introducing new policies (Grelle & Hofmann 2024). Moreover, we seek to contribute to the large literature showing that information provision can shape public preferences for policy reforms, such as pension reforms (Schuetz et al. 2023), education reforms (Lergetporer et al. 2020), and eurozone policy reforms (Dolls & Wehrhöfer 2021). We aim to contribute to the understanding of how beliefs about distributional aspects, environmental concerns, and personal interests and their interaction with information provision shape support for policy reforms within the German tax system.

Experimental Design Participants were assigned to either a between-subjects or a within-subjects design. In the within-subjects group, participants first indicated their initial beliefs on a 7-point Likert scale about the commuter tax allowance prior to the experiment, focusing on three dimensions: distributional aspects (the term used in the experiment was “social fairness” [“soziale Gerechtigkeit”]), environmental impact, and own financial interests. All participants were then randomly assigned either to a control group or one out of two treatment groups. The control group received basic information on the commuter tax allowance, while the two treatment groups received additional information about either its disproportionate benefit to high-income earners or its potential to encourage long commutes without incentivizing the use of sustainable modes of transport. After receiving this information, participants either updated their beliefs (within-subjects group) or stated their beliefs (between-subjects group). Finally, they indicated their support for five reform options using a 7-point Likert scale. These are: (1) removal of the commuter tax allowance without replacement, (2) an income-independent mobility allowance of 10 cents per kilometer for all commuters, (3) a ‘scandinavian model‘ restricting car journey deductions to cases where public transport is not viable, (4) capping the allowance to the cost of public transport (i.e. the Deutschlandticket (€49 per month) for short distances), and (5) increasing the allowance to 50 cents per kilometer. Lastly, participants’ beliefs regarding the three dimensions were also assessed for each reform option. If participants answered the assessment of the distributional aspects, environmental impact and own financial interests regarding the reform options with ‘don’t know/no answer‘, we asked them to name the reasons that led them to choose this option.

Data Our preregistered survey experiment draws on a sample of approximately 3,055 German household members from the forsa.omninet panel, which is representative of the German population aged 18 and older with access to the internet.

Results Initial belief elicitation in the within-subject group reveals that 40 % of participants consider the current commuter tax allowance as socially fair, while 26 % disagree. A share of 8 % beliefs that the current regulation benefits the climate and the environment, while 28 % state that it benefits them financially. Independent of information provision, support for a reform toward the ‘scandinavian model‘ is the highest at 39 % and the lowest at 13 % for a removal. Support for the various reform options varies according to party preference: participants with a (center)-right conservative party preference have a statistically significant higher acceptance for an increase than those with a left-wing party preference. We find the opposite result for the support for the ‘scandinavian model‘. Notably, support for an income-independent mobility allowance is consistent across different party preferences. Providing information on tax refund differences based on income reduces perceptions of fairness by 22 percentage points, with varying effects depending on prior knowledge of commuter tax allowances. It also increases support for income-independent mobility allowances by about 10 percentage points. Environmental information influences beliefs on fairness but doesn't significantly affect support for proposed reforms, with variations based on prior knowledge.

ABSTRACT. Artificial Intelligence (AI) agents represent a novel class of social actors within social and economic systems. To ensure the effective integration of human-AI societies, it is crucial to understand how people perceive and form impressions of AI agents. A particularly useful framework for understanding human-AI relationships is offered by social cognition research (Fiske, 1993, 2018). In this framework, the two fundamental dimensions of social cognition are warmth and competence. Warmth refers to one’s intentions, whether they are benevolent or malevolent, while competence refers to one’s ability to act on their intentions. The perception of an interaction partner’s low/high warmth and low/high competence is related to emotional and behavioral reactions towards them. Although both warmth and competence judgments are fundamental for social perception, evidence indicates the primacy of warmth judgments; warmth is typically judged before competence and carries more weight in affective and behavioral reactions (Fiske, 2018). Non-human actors, such as AI agents, elicit social cognitive judgments in a similar manner as humans, and prior work has identified warmth and competence as meaningful dimensions of AI agents’ perception (Harris-Watson et al., 2023; McKee, Bai & Fiske, 2023). In the present research, we compared human-human and human-AI trust-related interactions in terms of the fundamental dimensions of social cognition as the trustee’s characteristics, using a set of vignettes describing everyday trust-related situations. In the first study (N = 198), we compared the importance of warmth and competence characteristics of a human or AI trustee when making decisions in different types of trust-related situations. In the second study (N = 600), we compared warmth- and competence-related attributions when explaining the successful or unsuccessful result of a trust-related situation involving one of three trustee types: an AI agent, a human expert, or a human friend. In both studies, we used the same set of six vignettes, with two vignettes representing each of the three types of situations: (1) situations in which people would rather trust a friend, (2) situations in which people would rather trust a human expert, and (3) situations in which people would rather trust an AI agent. In Study 1, participants were asked to imagine that they found themselves in each of the six situations presented in the vignettes in which they were considering asking the trustee for support or advice. After reading each vignette, participants were asked to rate the importance of trustee characteristics when making a decision of whether to trust them or not. The results of Study 1 revealed that when deciding of how much trust to place, the two fundamental dimensions of social cognition were more influential in scenarios involving human trustees than in those involving AI trustees. Additionally, the competence of the trustee emerged as more critical than their warmth. However, the importance of the two dimensions depended on the type of the situation being considered. In situations in which people would rather trust a friend, the importance of warmth was higher, and the importance of competence was lower than in situations in which people would rather trust an expert or an AI agent. In Study 2, participants were first asked to imagine that they found themselves in each of the six situations presented in the vignettes. After reading each scenario, participants were informed about the trustee's performance (success or failure) and asked to rate the trustee on warmth and competence items. Below is a sample vignette with trustee type and performance manipulations, as well as warmth and competence measures: „You have a legal problem that could lead to serious consequences and you don’t know how to handle it. You decide to consult a/an [AI agent/expert/friend]. You follow their advice and end up really [happy/unhappy] about the consequences. In your opinion, the [AI agent/expert/friend] is: a) friendly; b) competent.” Results of Study 2 showed that when explaining trust-related situations in retrospect, i.e., already knowing their consequences, people attributed less warmth and less competence to AI agent trustees than to human trustees (experts or friends). Also, trustee performance had a smaller effect on the warmth and competence scores of AI agent trustees than in the case of human trustees. Just like experts, AI agents were rated as equally warm and competent, while friends were rated as more warm than competent. Our outcomes should be considered in the context of the functions of stereotypes, which help people make sense of the social world. In our present research, AI agents were viewed as less competent than humans from high competence groups (friends or experts), less warm than friends (high warmth stereotype), and equally or less warm than experts (low warmth stereotype). AI technologies are perceived as social actors, and people interact with them as such. Therefore, we believe it would be valuable to further explore stereotypes that can be associated with AI agents as a social group. The results might also be considered in the context of cognitive biases. Recent research on the perception of AI showed that in a number of contexts, the label “AI” could have a negative connotation (e.g. Bellaiche et al., 2023; Grassini & Koivisto, 2024). Our findings suggest that, under identical conditions, AI agents were rated lower than humans in both warmth and competence, which likely confirms the presence of an anti-AI bias in impression formation. Furthermore, the diminished importance of warmth and competence when making trust-related decisions in human- AI interactions may be linked to a form of apprehension toward AI technologies. Future research should explore whether these more negative attitudes and reduced trust in AI agents, compared to humans, are rooted in unconscious biases against AI technologies. Understanding what shapes people’s perceptions of AI technologies is crucial for the smooth functioning of human-AI societies. The presented results contribute to interdisciplinary research with the goal to ensure beneficial interactions between humans and AI technologies.

References are provided in a separate file as adding them here exceeds the word limit.

ABSTRACT. This paper explores the efficacy of AI-driven chatbots in accurately inferring personality traits compared to traditional psychometric tests within a real-world professional hiring context. The rapid implementation of Artificial Intelligence (AI) in recruitment has transformed how organisations identify, assess, and select candidates. Companies such as Pymetrics, Unilever, and Sapia have integrated AI-driven solutions to optimize hiring processes, reduce biases, and enhance the candidate experience (Dai et al., 2022; George et al., 2021). However, the fast adoption of these technologies has resulted in a significant gap between practical applications and academic validation—a practitioner-academia gap (Will et al., 2022). Historically, psychometric tests have been the gold standard in assessing candidates' personality traits, with a track record of predictive validity and reliability (Schmidt & Hunter, 1998). In contrast, AI’s role in hiring is relatively new, and its effectiveness, particularly in personality assessment, is an emerging area of research (George et al., 2021; Will et al., 2022). This study aims to bridge this gap by rigorously comparing AI-driven chatbots with traditional psychometric tests in inferring candidates' personality traits. Using a quasi-experimental design with propensity score matching, we analysed data from 159 candidates and other professionals from Serbian and Montenegrin regions who completed both traditional psychometric assessments and AI-based personality evaluations based on the Big Five Personality model. A novel one-question-per-facet approach was employed in the chatbot assessments with a goal of enabling more granular analysis of the chatbot’s psychometric properties. Results confirm the replicability of the Five-Factor structure through AI-inferred personality scores, consistent with the results of Fan et al. (2023), Dai et al. (2022), and Grunenberg et al. (2024). These prior studies highlighted AI’s potential to infer personality traits with moderate alignment to traditional psychometric tests, particularly for Extraversion and Neuroticism. The adoption of a one-question-per-facet technique represents an advancement on Fan et al.’s framework, offering more granular insights into the psychometric properties of chatbot-inferred scores and addressing substantive validity challenges noted by Hickman et al. (2022). The findings on social desirability bias offer a significant contribution. Consistent with Fan et al. (2023), this research demonstrated that AI chatbots are resistant to social desirability effects, unlike traditional psychometric tests, which are prone to response distortion in high-stakes settings (Mikulić & Sekulić, 2014; Ziegler & Kemper, 2013). By validating AI methods in actual hiring scenarios, this study extends the ecological validity of prior findings, reinforcing the practical applicability of chatbots in reducing bias. This aligns with Dai et al.’s (2022) recommendation to validate AI methodologies in diverse and real- world contexts. We aimed to begin unpacking the “black box” of AI (Mills, 2021) in behavioural science and bridge the practitioner-academia gap in the field of AI Psychometrics (Will et al., 2022) by applying rigorous methodologies in a real-world professional selection scenario. It introduced three significant contributions to the existing literature: 1) using a practical, real- world hiring setting, 2) implementing a question-per-facet approach, and 3) exploring AI- driven psychometrics in an underrepresented language. This study’s methodological innovation: introducing the one-question-per-facet approach demonstrates how AI can address longstanding psychometric issues, such as improving substantive and structural validity, advancing methodological rigor, and offering a pathway for continuous improvement. By combining this methodological refinement with a high-stakes professional context, it bridges the practitioner-academia gap identified by Will et al. (2022). The real-world setting provided an opportunity to evaluate AI tools under actual hiring conditions, enhancing ecological validity and offering insights directly relevant to practical applications. Furthermore, conducting this research in an underrepresented linguistic and cultural context expands the applicability of AI-driven assessments, addressing an area that remains relatively unexplored. From a practical perspective, the findings of this study hold significant implications for organizations adopting AI in recruitment. The demonstrated resistance of AI chatbots to social desirability bias highlights their potential in high stakes hiring scenarios where response distortion can compromise the accuracy of traditional assessments. Additionally, the ability to apply AI-driven psychometrics in diverse cultural and linguistic contexts expands their utility for global organizations seeking scalable, unbiased hiring solutions. By refining methodologies like the question-per-facet approach, this research provides a framework for organizations to enhance the reliability and validity of AI assessments, ensuring they meet both practical hiring needs and ethical standards. In conclusion, this study highlights the potential of AI-inferred personality traits, particularly their resistance to response manipulation, while underscoring the need for further refinement and validation to match the rigor of psychometric measures. By advancing AI- driven psychometric methods and emphasizing methodological rigor in real-world applications, this work contributes to the growing body of research at the intersection of AI and psychometrics, paving the way for more practical and scientifically robust personality assessments.

ABSTRACT. Harnessing artificial intelligence to develop navigators for use by commuters has the potential to reduce the economic and environmental cost of traffic congestion by coordinating route-choice. However, the use of such navigators may entail private costs, limiting uptake. It is therefore important to understand the strategic and behavioural factors underlying an individual’s choice to employ such a navigator, as well as features of the navigator that may encourage or discourage its use. In a lab experiment, we study individual and group behaviour in a congestion game where drivers with heterogenous travel costs must take one of two routes, one of which is shorter but subject to congestion. In some treatments, drivers may delegate their choice to an AI navigator which can manage traffic flows to improve efficiency. The navigator has the potential to reduce aggregate travel costs by acting as a coordinating device to prevent congestion, and, if allowed to make use of individual travel costs (as in some treatments), can allocate the drivers efficiently. However, this replaces one social dilemma with another, as a driver can free-ride on others’ decisions to use the navigator, choosing the shorter route in the knowledge those using the navigator will be diverted if necessary to prevent that route becoming congested. Also, while a navigator that uses information about individual travel costs will be more efficient in allocating those who choose to use it than one that does not, it may be less frequently utilized by drivers with low travel costs who are more likely to be diverted by the navigator to the longer route. In a second treatment arm we test whether a “nudge” which makes subjects aware of a real environmental cost of congestion in the game can improve take-up of the AI navigators. This is implemented by reducing the size of a donation to an environmental charity each time there is congestion in a payoff-relevant game. This experiment provides valuable insights into the effectiveness of environmental nudges in encouraging the adoption of AI navigation systems. By highlighting the real environmental costs of congestion, the study explores how such nudges can influence individuals' decision-making, ultimately promoting the use of AI solutions that alleviate traffic congestion and mitigate its associated environmental impacts.

ABSTRACT. Many practitioners in the public and private sector are increasingly realizing the value of applying behavioral sciences for solving real-world problems. However, it can be daunting to navigate the plethora of different recommended frameworks, guidelines, and methods. There are subtle differences between recommendations stemming from epidemiology, economics, psychology, anthropology, design and other behavioral sciences. In many ways this diversity is a source of strength as different disciplines illuminate slightly different aspects of complex human behavior. But the proliferation of seemingly competing recommendations can also be discouraging for time-pressured practitioners and a fodder for bad-faith criticisms of behavioral sciences as immature for application.

To alleviate this problem, a group of Estonian researchers recently developed a cross-discplinary Applied Behavioral Science (ABS) framework (Karolin et al., 2023). The framework links three kinds of behavioral scientific practices to three kinds of goals practitioners need to accomplish when developing effective solutions such as regulations, services, and programs to tackle problems with a behavioral component.

The three goals are a) to understand the problem at hand, b) to craft solutions, and c) to assess their impact. The three practices are investigative practices that help to understand problems and assess impacts; modelling practices that help map the causes of problems and mechanisms of potential solutions; and finally, experimentation practices that help develop solutions through prototyping and evaluate them in effectiveness studies.

Combing the three aims with the three practices yields a 6-step cycle: 1) investigate experiences, 2) map the problem, 3) map solutions, 4) design prototypes, 5) design experiments, and 6) investigate outcomes. The steps form only a loose sequence. Each successive step builds on the previous ones, but can also generate insights that help refine an earlier step.

Step 1: Investigating experiences involves learning about the problem at hand and potential solutions from past experience captured in research literature as well as from people with relevant lived experience. The investigative practices most useful here include “desk research” such as synthesizing literature and “field research” such as qualitative interviewing. Combining desk and field research can help situate broad insights into a specific context.

Step 2: Mapping the problem involves constructing an actionable model of the (often complex) systems that underlie the problem. The problem map links the distal outcomes of the problem that usually motivate action (e.g., economic or social costs) to the more proximal behaviors of specific target groups and also identifies the main factors that shape these behaviors. Problem maps can be drawn on existing frameworks and theories as well as be unique to the problem at hand.

Step 3: Mapping of solutions involves identifying ways to interfere with the system depicted on the problem map. Candidate solutions can be gleaned from Step 1 investigations or brainstormed using the Step 2 problem map. Promising solutions can be found by considering their expected feasibility (i.e., costs and challenges) and impacts (i.e., effects and side-effects). The solutions map could be structured using the problem map or existing solution-oriented frameworks.

Step 4: Designing prototypes involves iteratively implementing scaled-down versions of promising solutions to identify parameters of a successful implementation in a given context. Prototypes are usually easier, cheaper, faster, and safer than full-scale solutions, e.g., tabletop exercises, mock-ups, and pilot studies.

Step 5. Designing experiments involves finding a way to compare the consequences of implementing a solution to the consequences of not implementing it. This could involve comparing people who experience a solution to those who do not, assuming sufficiently similar groups can be created through random assignment of large samples or careful matching of smaller ones. Experiments could also involve comparing the same people before and after experiencing a solution, assuming the solution does not produce sustained effects. Often, it is advisable to combine both approaches.

Step 6. Investigating outcomes involves assessing variables that represent desired outcomes, potential side effects, and, if possible, the mechanisms of change. Here, the emphasis is on using reliable and generalizable measures.

The framework was developed as part of a broader initiative to encourage evidence-based policymaking in the Estonian public sector, led by the Estonian Government Office’s Innovation Team. This initiative has used the framework to procure an online guidebook as well as design and conduct training and co-creation sessions that help officials apply the framework to their practical problems. This process has highlighted the importance of reducing disciplinary jargon and focus on core scientific practices.

This framework is not the first or last set of guidelines for applying behavioral science (Conrad et al., 1999; Fernandez et al., 2019; Hardeman et al., 2005; Moullin et al., 2015; Powell et al., 2021). Its uniqueness lies in its cross-disciplinary and simplicity, allowing it to be used across different contexts and by practitioners with different disciplinary backgrounds. As such, ABS is not intended to be an exhaustive blueprint but rather a roadmap that can use to guide their actions and procure help from specialized service providers as needed.

ABSTRACT. Premise

Nudge theory is only seemingly reaching its maturity stage via a series of meta-analyses employed to assess the tool’s competency (i.e. Broers et al., 2017; Caraban et al., 2019). Counterintuitive results, nevertheless, make the nudge’s effectiveness inconclusive at best. For example, Mertens et al. (2022), meta-analysis identifies nudges to be effective across a range of contextual applications yet follow-up examinations by Maier et al. (2022) refuted these findings on the basis that publication bias was not accounted for. These examples demonstrate a growing pattern in the literature that focuses on the tool’s effectiveness (or ineffectiveness), yet there are fewer discussions of unintended adverse effects caused by choice architects.

Sunstein’s (2017) acknowledgment of failed nudges led to his categorization between ineffective (with zero utility effect) and counterproductive nudges (with negative utility). Nevertheless, other than reactance, little is offered on the causality of the latter type and, by extension, on the ways to mitigate them. In addressing counterproductive nudges, as a result of unaccounted behavioural trajectories by choice architects, research on negative spillovers serves as the primary source. Dolan and Galizzi (2015) discussed a number of possible predictors of spillovers, such as moral licensing and ego depletion. Their conceptualisation of different spillover types is based on how a targeted behaviour 1 can lead to an unintended behaviour 2, with the main focus fixed on behaviour 2. The premise of such conceptualisations is that spillovers occur after an individual has entered into the intended behavioural trajectory. For example, and applicable to nudges, a social comparison nudge pushes an individual to start exercising (behaviour 1) and as a consequence consumes a cake (behaviour 2) at the end of the exercising cycle, demonstrating a negative spillover, possibly, attributable to moral licencing.

We argue that adverse effects can also occur prior to entering into behaviour 1 and as a consequence of a rejected nudge. This signifies a difference to adverse effects attributable to spillovers. For example, an individual might reject a social comparison nudge to exercise in the form of “Out of 1000 similar users you are ranked 990th for your calorie-burning performance”. The individual does not engage in exercising, yet due to this social comparison, her self-esteem drops (White et al., 2006), ultimately leading to cake consumption. Here, behaviour two was triggered because of the nudge type and not as a consequence of behaviour 1. We conceptualise this as a Nudge Misfire and argue that contrary to nudge spillovers that can be examined and mitigated through the psychological lenses proposed by Dolan and Galizzi’s (2015), misfires are primarily caused by the type of nudge employed (Themistocleous and Karapanos, 2025).

Aim and Contribution

This paper provides a critical literature review leading to the distinction between nudge spillovers and nudge misfires. We focus on occasions where nudges generate anomalous and inconsistent behaviours compared to the intentions of the choice architect that specifically position individuals in worse-off situations. We provide implications to choice architects through a new taxonomy of counterproductive nudges, paving a clearer way for empirical investigations of these issues.

We next propose a tool-box for both spillovers and misfires that assists with mitigating the latter two, specifically, through the nudge type chosen as well as its timing (i.e., Caraban et al 2019). Through this we expand the work of Themistocleous and Karapanos (2025) by blueprinting how certain nudge types can mitigate spillovers and misfires. For example, moral licensing, a prominent source of spillovers, is accentuated primarily in altruistic behaviours such as donations and other acts of philanthropy (Clot et al. 2022). Choice architects specifically in these domains need to consider that individuals might indulge a nudge for a donation yet moral licensing might hinder them from more meaningful acts of charity. Default nudges with a recommended amount might be less relevant to tackle such a spillover effect. We propose informative, system 2 nudges that aim to instigate a sense of responsibility to philanthropic acts rather than being perceived as acts of generosity, as the latter could justify other unsustainable hedonic behaviours.

References

Broers, V. J., De Breucker, C., Van den Broucke, S., & Luminet, O. (2017). A systematic review and meta-analysis of the effectiveness of nudging to increase fruit and vegetable choice. The European Journal of Public Health, 27(5), 912-920.

Caraban A., Karapanos E., Gonçalves D. and Campos R. (2019). 23 Ways to Nudge: A Review of Technology- Mediated Nudging in Human-Computer Interaction. CHI Conference on Human Factors in Computing Systems Proceedings (CHI 2019), May 4–9, 2019, Glasgow, Scotland UK.

Clot, S., Della Giusta, M., & Jewell, S. (2022). Once good, always good? Testing nudge’s spillovers on pro environmental behavior. Environment and Behavior, 54(3), 655-669.

Dolan, P., & Galizzi, M. M. (2015). Like ripples on a pond: Behavioral spillovers and their implications for research and policy. Journal of Economic Psychology, 47,1-16.

Maier, M., Bartoš, F., Stanley, T. D., Shanks, D. R., Harris, A. J., & Wagenmakers, E. J. (2022). No evidence for nudging after adjusting for publication bias. Proceedings of the National Academy of Sciences,119(31).

Mertens, S., Herberz, M., Hahnel, U. J., & Brosch, T. (2022). The effectiveness of nudging: A meta-analysis of choice architecture interventions across behavioral domains. Proceedings of the National Academy of Sciences,119(1).

Sunstein, C. R. (2017). Nudges that fail. Behavioural public policy, 1(1),4-25.

Themistocleous C. and Karapanos E. (2025), “Precision Nudging: The Future of Behavioural Change”, Christou P. (Ed.) AI in Social Research. CABI Publishing. [Forthcoming]

ABSTRACT. Sludge is one of the most important yet underappreciated problems in modern society. Examples of sludge include unnecessarily complex paperwork requirements, difficulties in cancelling memberships, long waiting periods, and unfriendly staff interactions. Sludge is often argued to have disproportionate effects on vulnerable groups, as they are more likely to encounter and to struggle with such friction. This paper uses a survey to investigate whether certain people are more vulnerable to sludge than others. Participants saw ten different types of sludge required to complete a “task” in a random order using the wording described in Table 1. Participants then indicated how vulnerable they were to each type of sludge using items such as “It would be difficult for me to carry out this task”, with answer options on a 5-point Likert scale from “Strongly disagree 1” to “Strongly agree 5”. Participants also answered several questions about psychological variables and demographics. Drawing on data from a nationally representative survey with 1,591 participants in Ireland, the paper shows that people with lower levels of mental and physical health are more vulnerability to sludge (see Table 2). Self-reported administrative literacy emerges as a protective factor, while the tendency to procrastinate and a lack of time and mental energy predict vulnerability to sludge. Administrative literacy, procrastination, and time and energy scarcity fully mediate the association between mental health and vulnerability to sludge. The data also show that certain forms of sludge are particularly problematic, such as outdated websites with broken links, unfriendly staff interactions, complex documents laden with jargon, and hard-to-navigate websites (see Table 1). Moderation analyses reveal that administrative literacy is especially critical when navigating jargon-heavy documents but is less relevant when dealing with requests to provide personal information such as religious beliefs and sexual orientation (see Figure 1). We discuss implications for behavioural public policy design as well as user experience optimisation.

ABSTRACT. Tax systems are often described as complex and policy makers often propose simplifying tax systems to improve compliance (de Clerq, 2019; Pope, 2005; Baer and Silvani, 1997; Skinner and Slemrod, 1985). However, the effect of tax complexity on the tax compliance behaviour of individuals remains unclear. On the one hand complexity in tax systems assure the equity of tax systems by providing provisions for heterogenous populations (Kaplow:1996, 1998). On the other hand, complexity results in compliance cost in terms of time and money spent on complying as well as administration costs for tax administrations (Hite and McGill, 1992; Christian, Gupta, and Lin 1993, Eichfelder et al. 2012, Krause, 2000; MacDonald 1993). We use an incentivized experimental design to identify a causal effect of tax complexity on the compliance behaviour of individuals using a representative German sample.

Differing definitions and measures of both tax complexity and tax compliance exist making it difficult to understand the causal mechanism behind tax complexity and its effect on tax compliance. Additionally, tax complexity is either thought of in terms of tax system characteristics or taxpayer characteristics and experimental studies usually address either one or the other. De Neve et al. (2019) and Bellemare et al. (2019), for example, look at tax system characteristics by conducting experiments using complex tax correspondence and complex tax forms respectively. Eriksen and Fallen (1996), on the other hand, look at taxpayer characteristics by looking at how tax knowledge effects of perceptions of fairness.

However, tax system characteristics and taxpayer characteristics interact to affect the compliance behaviour of individuals. Salmon and Schniderman (2019), for example, find that self-reported earnings increase when ambiguity averse individuals are faced with uncertain audit probabilities. In a similar vein, this study combines both taxpayer and tax system characteristics into our design. We apply a factorial design using vignettes describing different hypothetical tax systems varying in complexity in terms of the number of rules as well as the extent to which the meaning of these rules is certain or not. Participants are asked to complete a real effort task for which they earn an income. Thereafter, they are shown a description of a tax system and asked to declare their income in a tax evasion game with multiple rounds. Compliance is measured by the declared income. The experiment is made incentive compatible by making the clarity of the rule dependent on the amount of effort the participant is willing to spend on understanding it

ABSTRACT. Sustainable behavior is a key factor in addressing global environmental, economic, and social challenges. Higher education institutions (HEIs) are expected to play a pivotal role in fostering sustainability awareness and equipping students with the necessary knowledge and competencies to integrate sustainable practices into their professional and personal lives. However, the relationship between sustainability knowledge and behavior remains complex, with previous research yielding conflicting findings regarding its impact (Vicente-Molina et al., 2013). While HEIs increasingly integrate sustainability and sustainable finance into their curricula and extracurricular activities, a gap remains in understanding how different types of knowledge—subjective and objective—shape students’ sustainability attitudes and behaviors. This study contributes to the literature by examining the factors influencing sustainability knowledge acquisition among university students and exploring the role of attitudes as a mediator between knowledge and behavior. It distinguishes between subjective knowledge (both general sustainability knowledge and specific sustainable finance knowledge) and objective knowledge (formal sustainability knowledge acquired through academic courses). By analyzing the impact of demographic and academic variables such as age, country, field of study, and education level, this research provides deeper insight into how sustainability knowledge is formed and how it translates into sustainable behavior. The findings offer valuable insights for educators, policymakers, and institutions seeking to enhance sustainability education and foster meaningful behavioral change among future leaders. Using survey data from 1,511 university students across Spain, Italy, and Israel, we classify subjective knowledge into two categories: (1) Sustainability Knowledge (SK), which covers general sustainability concepts, and (2) Sustainable Finance Knowledge (SFK), which focuses specifically on sustainable finance, a specialized area within sustainability. Objective knowledge is assessed through Formal Sustainability Knowledge (FSK), referring to knowledge acquired through courses in sustainability-related subjects. Univariate analysis was conducted according to the demographic and academic variables. The findings reveal notable differences between countries. Spanish students exhibit higher subjective knowledge (SK & SFK) scores than their Israeli peers. However, when comparing Spanish and Italian students, no significant difference is found in Sustainability Knowledge (SK), while Spanish students score higher in Sustainable Finance Knowledge (SFK) than their Italian counterparts. These differences may reflect stronger national sustainability education policies or a greater cultural emphasis on sustainability awareness. Additionally, younger students tend to have higher general Sustainability Knowledge (SK), whereas older students score higher in the more specific field of Sustainable Finance Knowledge (SFK). This pattern suggests that general sustainability knowledge is often introduced at earlier educational stages, while specific sustainable finance knowledge develops later, potentially through professional exposure or advanced courses. The analysis also reveals significant differences in sustainability knowledge by education level. Ph.D. students demonstrate higher Sustainability Knowledge (SK) than bachelor's students, while Master’s students score higher in Sustainable Finance Knowledge (SFK) than Bachelor's students. However, Formal Sustainability Knowledge (FSK) does not significantly differ across education levels, suggesting that the availability of sustainability courses remains relatively uniform. This finding highlights the need to ensure accessibility to sustainability education at all academic stages to develop relevant sustainability competencies. Academic discipline also plays a crucial role in sustainability knowledge acquisition. Students in Management and Economics demonstrate higher levels of Sustainable Finance Knowledge (SFK) compared to all other disciplines except Law, reinforcing previous findings that business and economics students have greater exposure to sustainability finance topics (Avelar et al., 2019; Haertle et al., 2017). In contrast, Education students exhibit lower levels of both Sustainability Knowledge and Sustainable Finance Knowledge, suggesting that sustainability content is less integrated into education curricula. This finding underscores the need for broader interdisciplinary sustainability education to ensure that sustainability principles are incorporated across academic fields. In addition, structural equation modeling (SEM) was applied to examine how those different types of knowledge influence sustainability attitudes and behaviors. The results indicate that the country of the academic institution has a significant positive effect on all types of knowledge. Age affects only specific Sustainable Finance Knowledge (SFK), with older individuals exhibiting a significant positive effect. Regarding education level, the findings reveal a significant negative effect on Formal Sustainability Knowledge (FSK), indicating that higher levels of education are associated with lower levels of formal sustainability knowledge. Furthermore, each type of knowledge has a significant positive effect on attitude, with Formal Sustainability Knowledge (FSK) exerting the strongest influence, followed by specific Sustainable Finance Knowledge (SFK), while general Sustainability Knowledge (SK) has the weakest effect. This suggests that self-perceived familiarity alone may not be sufficient to shape strong sustainability attitudes, as subjective knowledge can sometimes lead to overconfidence or gaps in applied understanding (Schlegelmilch et al., 1996; Laroche et al., 2001). In addition, it aligns with studies emphasizing the importance of structured sustainability education in higher education (Michel, 2020; Barros-Guerro & Bello-Benavides, 2022). Finally, attitude has a significant positive effect on sustainable behavior, reinforcing the notion that fostering positive sustainability attitudes is key to promoting sustainable actions. The results consistent with prior research emphasizing the Knowledge-Attitude-Behavior (KAB) model (Michael et al., 2020; Tang, 2018). Students with stronger pro-sustainability attitudes are more likely to engage in responsible consumption, advocacy, and sustainability-oriented decision-making. In summary, this study has several practical implications. Given that different types of sustainability knowledge influence attitudes and behavior in varying ways, HEIs should integrate both general and specific sustainability knowledge into curricula across disciplines and education levels. The observed country-specific differences highlight the impact of national sustainability policies, emphasizing the need for stronger reforms to enhance sustainability awareness and competency development. Additionally, fostering positive sustainability attitudes through experiential learning and interdisciplinary approaches is essential for promoting sustainable actions among students, preparing them to apply sustainability principles in their professional and personal lives. The findings also support a behavioral model where attitudes mediate the relationship between knowledge and behavior, reinforcing the importance of educational strategies that not only provide knowledge but also shape attitudes to drive meaningful behavioral change.

ABSTRACT. Public service motivation (PSM) has long been a subject of interest in public administration research, with the choice of a public sector career often viewed as a vocation driven by intrinsic motivations. However, recent evidence suggests that extrinsic factors such as job security and salary may play an increasingly important role in career choices. This may hold true in Romania in particular, given its communist heritage, and the chronical underperformance of its public administrative sector. Our research examines the factorial structure of established PSM scales in a student population with diverse academic backgrounds, testing for invariance across academic disciplines and over time. We aim to understand whether PSM constructs hold consistent meaning and measurement properties across different fields of study and whether these properties remain stable over time. This research contributes to the ongoing discussion about the nature of PSM in contemporary society and its implications for attracting and retaining talent in the public sector. Our sample consists of about 600 students at the University of Bucharest, enrolled in Business Administration, Marketing, and Public Administration bachelor programs in the 1st, 2nd and 3rd year of studies. The Ethical Committee of the University of Bucharest approved the research, decision no. 3/9.01.2025. First, we conduct an exploratory factor analysis (EFA) to identify the underlying factor structure of various PSM scales. We examine the original 24-item scale developed by Perry (1996), its subsequent 14-item reduced version, and other more recent abbreviated scales proposed in the literature. The original 24-item PSM scale posits a four-dimensional structure comprising attraction to policymaking, commitment to the public interest, compassion and self-sacrifice. Our EFA seeks to determine the most robust and parsimonious factor structure applicable to our student sample. Following the EFA stage, we conduct confirmatory factor analysis (CFA) to assess the fit of the identified factor structure to the observed data. In addition, we explore whether other student characteristics such as age, gender or early family influences shape their perception on public service matters. A central focus of this paper is the examination of measurement invariance across academic fields. We employ a multi-group CFA approach to test whether the factor loadings, intercepts and residual variances of the PSM scale are equivalent across students majoring in business administration, marketing and public administration. This analysis addresses the question of whether the PSM construct operates similarly across these different academic groups. We hypothesize that while a general PSM factor may exist, the relative importance of its specific dimensions may vary significantly depending on the student’s chosen field of study. This research is critical for understanding the motivational profiles of potential public sector employees and tailoring recruitment strategies accordingly. Furthermore, we investigate the time invariance of the PSM scale. By administering the PSM instrument to students from different years of study, we assess the stability of the factor structure and the consistency of individual PSM scores over time. This longitudinal perspective allows us to explore whether students’ perception on PSM changes during their academic career. Understanding temporal dynamics of PSM can prove crucial for designing academic curricula that cultivates public service motivation in future professionals. The findings of this study have important implications for both theory and practice. Theoretically, our research contributes to a deeper understanding of the multidimensional nature of PSM and its variability across different contexts. Practically, the study offers valuable insights for public sector organizations seeking to attract and retain motivated employees. By understanding the motivational drivers of students from different academic backgrounds, public administration programs and government agencies can develop targeted recruitment and development initiatives. Moreover, our findings can inform the design of educational programs aimed at fostering PSM among students interested in public service careers. Our work is not without limitations. While targeting student population is the aim of this research, our study includes only three academic backgrounds. This may limit the generalizability of the findings to other academic fields. Future research directions are discussed, including exploring the influence of other demographic variables and examining PSM among current public sector employees. Despite these limitations, this study contributes to the literature on public service motivation by providing a comprehensive analysis of its factorial structure and invariance across a diverse student population in Romania.

ABSTRACT. Abstract Psychology is a field that requires both intellectual engagement and emotional resilience, making the academic experience and future career plans of psychology students particularly dependent on their personal dispositions, professional identification, and social belonging. Previous research suggests that personality traits, empathy, and social identity play a crucial role in shaping students’ satisfaction with their studies and career aspirations. The Big Five personality traits (BFI; John & Srivastava, 1999), particularly extraversion and emotional stability (inverse of emotional instability/neuroticism), have been also found to predict study satisfaction and career motivation (Furnham, 2012). Meanwhile, empathy (Interpersonal Reactivity Index, IRI; Davis, 1983) plays an essential role in students’ engagement with psychology, especially in clinical and counseling settings. Furthermore, professional identification with psychology as a field, operationalized through collective self-esteem (CSE; Luhtanen & Crocker, 1992), may contribute to higher satisfaction and stronger career commitment. This study investigates how Big Five personality traits, empathy, and collective self-esteem predict satisfaction with psychology studies and future career plans in psychology students. Understanding these relationships may provide insight into how students develop academic and professional identity and what factors shape their long-term career plans in psychology.

1. Theoretical background

1.1. Role of Big5 personality dimensions in developing a professional career Personality is a crucial determinant of academic and career-related attitudes and behaviors. The Big Five Inventory (BFI; John & Srivastava, 1999) provides a well-established framework for assessing Big5 personality traits in an academic and professional context. •Extraversion is associated with higher social engagement, energy, and academic enthusiasm, which may contribute to greater satisfaction and stronger career orientation. •Emotional instability (a facet of neuroticism) is linked to higher stress, anxiety, and lower well-being, making it a likely negative predictor of both study satisfaction and career confidence. •Agreeableness, conscientiousness, and openness are often associated with academic performance, but their role in study satisfaction and career planning is less direct.

1.2. Empathy and Academic Engagement of psychology students Empathy plays a key role in psychology students’ academic motivation and career commitment, as their future profession often involves helping and understanding others (Mehrabian & Epstein, 1972). Measured by the Interpersonal Reactivity Index (IRI; Davis, 1983), empathy consists of different facets, two of which are particularly relevant: •Empathic concern reflects compassion and emotional sensitivity toward others, fostering academic engagement and professional identity in psychology students . •Personal distress, in contrast, refers to self-oriented discomfort in emotionally charged situations, which may lead to higher emotional burden and lower confidence in a psychology-related career.

1.3. Social Identity and Professional Identification Social identity theory (Tajfel & Turner, 1986) posits that group identification enhances motivation and well-being. In academic and professional settings, identifying with one’s discipline contributes to higher engagement, resilience, and long-term career commitment. The Collective Self-Esteem Scale (CSE; Luhtanen & Crocker, 1992) assesses individuals’ feelings of belonging and worth within their professional group. Specifically, two dimensions of CSE are particularly relevant: •Private collective self-esteem, which reflects an individual’s personal evaluation of their group membership (e.g., feeling proud to be a psychology student/professional). •Membership collective self-esteem, which captures the perceived sense of inclusion and acceptance within the psychology field. Given that professional identity formation is a crucial step in career decision-making, it is expected that stronger identification with the psychology profession will positively predict both study satisfaction and career aspirations.

2. Methodology

2.1. Participants and Measures A sample of 174 undergraduate psychology students (91% female, mean age = 32.1 years) was recruited from a Hungarian university. Participants completed an online survey, including: •Big Five personality traits (BFI; John & Srivastava, 1999) •Empathy (IRI; Davis, 1983) •Professional identification (Collective Self-Esteem Scale; Luhtanen & Crocker, 1992) •Satisfaction with psychology studies (Likert-scale self-report) •Career plans in psychology (intended career commitment measured via Likert-scale responses)

2.2. Statistical Analysis Multiple regression analyses were conducted to examine the predictive role of personality, empathy, and professional identification on satisfaction with studies and career plans.

3. Results

3.1. Predictors of Satisfaction with Psychology Studies The regression model predicting satisfaction with psychology studies was significant (R² = 21%, p < .001), with the following key predictors: •Private collective self-esteem (β = .41, p < .001): Stronger professional identification was associated with higher satisfaction. •Emotional stability (BFI, β = –.20, p < .001): Students with lower emotional stability reported higher satisfaction with studies, suggesting that emotional sensibility influences academic experiences positively. •Extraversion (BFI, β = .14, p = .067): Higher extraversion was linked to greater study satisfaction, likely due to increased social and intellectual engagement. •Empathic concern (IRI, β = –.17, p = .02): Higher compassion and emotional sensitivity predicted lower satisfaction with studies, indicating that emotionally invested students find psychology studies more challenging. •Personal distress (IRI, β = .21, p = .014): Greater self-oriented distress was associated with higher satisfaction, suggesting that strong emotional experiences may lead to more engagement in studies.

3.2. Predictors of Career Plans in Psychology The regression model predicting career plans as a psychologist was also significant but weak (R² = 15%, p < .001), with the following key predictors: •Identification (CSE membership: β = .27, p < .001): Stronger identification as a future psychologist increased career commitment. •Importance of contribution (CSE memrship: β = .15, p = .076): a greater sense of belonging in the psychology field had a marginally significant positive effect career intentions. •Extraversion (BFI, β = .15, p = .054) also had a marginally significant positive effect on career plans.

5. Discussion and Implications These findings provide empirical support for the influence of personality, empathy, and social identity on psychology students’ satisfaction and career plans. Strengthening professional identity and emotional resilience may enhance students’ academic experiences and long-term career commitment in psychology.