ABSTRACT. There are 2.8 billion credit cards in the world (Shift, 2021). While this payment method has offered convenience, it had also brought financial distress to people who cannot manage their debt payment (Hodson, Dwyer, & Neilson, 2014). As a partial solution, policy regulations in several countries have included a mandatory minimum payment warning in credit card statements. However, several studies using administrative data or laboratory-based experiments have shown that highlighting the minimum payment may cause lower full balance payments (e.g., Navarro-Martinez et al., 2011; Salisbury, 2014; Wang & Keys, 2016). For this reason, the minimum payment policy has been described as a perverse nudge due to the anchoring bias (The Economist, 2008; Wang & Keys, 2014; Stewart, 2009). This paper introduces a novel statement balance warning to answer whether and why a minimum and/or a statement balance warning can change credit card payment behavior. In addition, research from psychology distinguishes that when there is a hierarchy of goals, one is prioritized, and non-prioritized goals are ignored (Unsworth, Yeo, & Beck, 2014). For that, this paper uses a randomized preregistered field experiment with credit card debtors who received email payment reminders before their credit card payment was due.

Debtors (N=179,706) were randomly assigned to four experimental conditions. The Minimum Payment Warning condition included the sentence: "If you at least pay the minimum ([$]) before the due date, you will pay additional interest charges but avoid late fees." The Statement Balance Warning condition included the sentence: "If you pay the statement balance ([$]) before the due date, you will avoid additional interest charges." The Both Warnings condition included the statement balance and minimum payment warnings. Finally, the Control condition was a simple reminder that included neither the statement balance nor minimum payment warnings.

Results. All warning messages decreased the likelihood of paying less than the minimum compared to a simple payment reminder without warning messages. On average, debtors decreased their less-than-the-minimum payments by 7.9% more than debtors who received a simple payment reminder (an absolute difference of 0.77 percentage points; ps<0.01). However, this masks a significant change in payment distribution depending upon the specific warning message. Debtors who received the statement balance warning were 1% more likely to pay in full than debtors who received the simple payment reminder (a difference of 0.64 percentage points; ps<0.01). In contrast, debtors who received the minimum payment warning increased their minimum payment by 5.9% compared to the simple reminder (a difference of 0.52 percentage points; p<0.01), and there was no sizably change in the likelihood of paying in full despite the precise estimate. This means that debtors who received the minimum-payment warning shifted their payment mostly toward the minimum amount (63% of the shift), and debtors who received a message including the statement-balance warning shifted their low payments almost entirely toward paying in full (87%). Even though debtors who received both warning messages behave very similarly to those who received only the statement balance warning, their effect on the interests charged on the following billing cycle after the experiment is not the same. Debtors who received a message including both warning or the only minimum warning saw their delinquent interest (i.e., the interest generated by not paying at least the minimum) being reduced by 3.9% and 3.4% (ps<0.05), respectively, more than those who received the message with only the statement balance warning, and 8.0 and 7.5% (ps<0.01), respectively, compared to debtors who received the simple reminder. Also, consequent to the changes in payment distribution, there is a difference in the revolving interest charged (i.e., the interest generated by paying less than the statement balance). Debtors who received both warnings or only the statement balance warning saw a reduction of 6.0 and 3.8% (ps<0.06), respectively, more than debtors who received only the minimum warning message, and 9.0 and 6.9% than those who received the simple reminder (ps<0.01). In order to explore heterogeneous effects, the field experiment was combined with a recently developed machine learning technique, causal random forests (Wager & Athey, 2018). Although a large fraction of debtors increases the portion they paid of their statement balance because of receiving a message with both warnings, a small fraction reduces this outcome. Different factors explain a positive and large treatment effect in this regard: debtors with previous payments closer to the statement balance, a small distance between the minimum and statement balance amounts, and a considerable variation on previous percent payments. These factors are consistent with debtors prioritizing one warning depending on which becomes more relevant and reachable to them.

In conclusion, results show that a statement balance warning not only increases payments in full, but also decreases delinquency and revolving interest. This paper contributes to the literature on nudges and financial decision-making by testing in the field the role of the minimum payment warning and, more importantly, its interaction with a new statement balance warning.

ABSTRACT. Consumer debt is a concern in most countries, increasing financial vulnerability and reducing financial resilience. Consumer culture is promoting immediate gratification and services such as Buy Now Pay Later (BNPL) are tempting into taking on debt without careful consideration. The framing of the consumer credit and BNPL conditions can have a significant effect on the intention to take the offer. To contribute to that, we tested in a randomized controlled trial the effects of including the APR and the total cost in such online offers to buy now and pay later already on the landing page. We also controlled for debt literacy and compared the affects across socio-economic indicators.

Our randomized control trial included a representative (age, gender, location) sample of 1204 participants in Estonia. The experiment was carried out online using a vignette with a hypothetical situation, where a person was having a dilemma whether to buy on credit or not. The participant was prompted to put themselves in the described persons’ position and decide whether to buy the item on credit or reject the offer. The conditions for the purchase were based directly on a current offer from a major online seller. The experiment also included a short survey to extract demographic indicators and measure the participants’ debt literacy. We included three conditions that the participants could be randomly assigned to:

• A) control group who was exposed to the actual offer from an online seller, displaying only the monthly payment sum to the participant (n=403); • B) keeping everything else the same, the total cost of the purchase was added into the offer (n=401); • C) keeping everything else the same, the annual percentage rate together with additional fees was added into the offer (n=400).

We analyzed the data in SPSS using between group Chi-square tests. The results of the experiment indicated a substantial decrease of intention to buy on credit for both groups B and C, with respectively a 8.5 times and a 4.25 times decrease in the share of participants who would choose to purchase the indicated item on credit. Although the direction of the effects was rather expected, the size of them is surprisingly large. The results show that the way credit information is currently presented to the consumer in online shops is highly conducive in making a credit decision.

The study also included a 20-question survey to determine the factors that consumers use to choose between credit options and to assess debt literacy and other potentially relevant background factors for deciding to buy on credit. The key findings of the survey indicate that 24% of the participants could not solve a simple interest calculation, 33% gave a wrong answer for a percentage-based credit question and 57% gave an incorrect answer for the definition of APR. This lack of debt literacy is worrisome considering nearly 50% of the participants had an unpaid loan. As over 50% of participants could not give a right answer for the definition of APR, the reduction in intention to buy on credit for group C was likely due to poor knowledge about APR. However, even though a person might not completely grasp the concept of APR, a high number listed could still deter intention to buy on credit. When comparing the credit intentions of those who responded correctly to any of the debt literacy questions to those with incorrect answers we did not observe any significant differences.

When analyzing attitudes regarding consumer loans, we found that participants reported being responsible borrowers. On a scale from 1 (completely agree) to 5 (completely disagree), the average score for considering taking on credit to pay previous debts was 4.3, thinking that life is easier with loans was 4.2 and finding that loans can be risky (2.3) and should be only taken as a last resort was rated with an average score of 2.3. However, when analyzing whether attitudes of those choosing to take on credit in our experiment differed from those who chose not to take it, we found that participants either completely trustful or moderately trustful of consumer credit businesses are 30% more likely to take on credit. Likewise, participants who indicated a positive consumer credit decision were nearly twice as likely to think that there is nothing wrong with taking on new debt to pay off previous debts, and over two times more likely to think that that there are less problems in life when they have access to credit.

When analyzing socio-economic factors, we did not detect any connection between buying on credit and demographic variables (such as gender, age group, region, living in city or rural areas), except for nationality: Russians were 40% more likely to buy on credit than Estonians. Furthermore, participants who intended to take on consumer credit were two times more likely to consider the comfortability of taking out a loan and 16% less likely to consider the price of the loan the key criteria. This highlights how certain consumers are more prone to take unconsidered impulsive credit decisions than others. In addition, having previous experience with a (paid or unpaid) consumer loan and more frequent use of credit cards were detected as factors for whether the person would take on a consumer loan.

As about 6% of the Estonian population have an active court mandated debt collection, we urge the necessity of easy to understand, concise information presentation in the e-shop environment, as evidently credit decisions are very malleable in this regard. The results could also be used in a wider context detailing the importance of choice architecture in credit decisions and understanding how attitudes towards financial institutions and consumer credit influence buying on credit.

ABSTRACT. Personalized investor newsletters are an attractive tool for influencing investor behaviour. They can be used to send tailored messages that take into account recent investor behaviour as well as the market situation. However, it is unclear which aspects of investor behaviour should be taken into account and which messaging strategies should be used for different investor profiles.

In this study, we investigated the effects of messaging strategies on the amount of funds invested in a field experiment with 4782 clients at an Estonian retail bank. The participants were sorted into two groups, with the treatment group (n=1518) receiving personalized newsletters about their portfolio performance each month for three months (09.-11.2020), and the control group (n=3264) not receiving any such e-mails.

Our research had four goals: •To see whether the investment of the intervention group increased more than the control group in terms of amount invested. •To determine which type of messages cause a change in the amount a person would invest. •To assess whether other financial behaviour (such as savings or cash balance) changed due to the intervention. •To examine the underlying psychological effects of the mechanism behind the newsletters’ influence and to give recommendations on how to design an effective newsletter for nudging investor behaviour.

The newsletter contained a message about the performance of the client’s existing portfolio (increase/decrease) and a comparison with their own past behaviour, either:

a) their consistency in investing new funds (investing every month / skipping some months) b) the relative size of their latest investment (more than personal average / less than personal average).

Therefore, the newsletter could have a positive or negative tone based on market performance and one´s own behaviour. The negative frame of the newsletter was induced by negative portfolio performance, inconsistent investing schedule, and below average investment size. Likewise, a positive portfolio performance, consistent investing schedule, and above average investment size made the newsletter sound more positive.

Depending on which services the client used, the message also contained a call to either activate a micro-investment service or a standing order for investments, or make an additional investment (if they had not already joined the automated investment scheme).

The investment behaviour of the intervention group was compared to both the behaviour of a case-controlled control group and to their own behaviour during a base-rate measurement conducted two months prior.

The results of a general lineal model analysis showed that the newsletter did not have an overall significant effect on increasing investments. No connection between messages and the sum invested was found. In addition, we found that the newsletter did not influence the account balance or savings but it did have an effect on the securities account. The average increase in the value of securities between the treatment and control group was about 20%, or 228 Euros (p=0.01).

We did not find that any specific type of message would have increased overall investments statistically significantly. However, we found significant increases in investments in certain investor groups. Specifically, we found that investments were increased by newsletters sent to investors who had invested more than 20% of their annual average the previous month. Likewise, investments were increased by newsletters sent to investors whose net investment was zero or negative the previous month but who had invested during the previous year. We also found that the treatment increased securities account balance for those clients who had invested consistently for at least 10 months in a row but had not invested anything the previous month.

Our results suggest that newsletters could serve as an encouragement for those who already invest significant amounts each month, and as a reminder for those who used to invest but have not done so for a month. Therefore, our results show the importance of investor behaviour analysis, as a targeted newsletter can increase investments only in certain segments.

We recommend employing call-to-action messages with a clickable button that the client can instantly use to invest a chosen sum in one click. We also find that the newsletter presentation schedule could benefit from better timing, so that it arrives the next day after a clients’ payday, and that the letter could be more effective if it considered the individual investor behaviour. Therefore, the study has important practical implications for financial institution newsletter design.

ABSTRACT. In many countries, the rate of mortgage switching and refinancing is notably subdued when viewed against a backdrop of widespread opportunities for substantial financial savings. Classical theory would lead us to anticipate a greater take-up of available refinancing option than is observed, a deficit that behavioural economic theory can help us to understand and address. Improving borrower engagement with and take-up of advantageous alternative offers is important for the health and resilience of household balance sheets, and the effective functioning of an essential product market. We carried out a field experiment in partnership with a major retail bank in the Irish mortgage market to test whether subtle and unobtrusive changes to a mandatory financial disclosure related to mortgage refinancing could prompt borrowers into greater take-up of these opportunities. Our treatment arms were designed to target potential behavioural barriers to optimal engagement. In the best performing version, we were able to lift the refinancing rate from 8.6 to 15 per cent, an increase of 74 per cent in the probability of refinancing.

ABSTRACT. Greed is as intriguing as it is controversial. Across various cultures and historical periods greed has been observed as an important motive underlying people’s behaviors and almost equally many times has greed been condemned as being immoral, sinful or outright evil (cf. Bloch, 1984; Haynes, 2021; Tickle, 2004). Most people do not want to be called greedy (e.g., Gilliland & Anderson, 2011), because this label has a very negative connotation. Empirical research has found that greedy people tend to be more dishonest, corrupt, and egoistic (Li et al., 2021; Seuntjens, Zeelenberg, Van de Ven, et al., 2015; Seuntjens et al., 2019). It is intriguing to think about how it is possible that a motive that appears to be so universally condemned has, at the same time, been able to persist over the centuries. Might it be that there are also benefits to greed or to being greedy, despite its condemnation and suppression? In other words, is there anything that greed is good for? In a representative sample of the Dutch population (N = 2,367, 51.3% female, Mage = 54.06, SD = 17.90), we examined two main questions. First, we examined whether greedy people are more or less successful than less greedy people (measured by the Dispositional Greed Scale, Seuntjens et al. 2015). We adopted the approach that Eriksson et al. (2020) recently used to study the benefits of selfishness. More specifically, we studied whether greedy people generate more personal income and more household income (indicators of economic success), whether they have more sexual partners, longer relationships, and more offspring (indicators of evolutionary success), and whether they are more satisfied in life (indicators of psychological success). We found that greedy individuals had more economic success, but mixed evolutionary success (more sexual relationships, but fewer long lasting relationships and fewer children) and less psychological success, than their less greedy counterparts. Second, we aimed to disentangle greed from self-interest in relation to these indicators. Here we found that greed differed from self-interest in terms of economic success, and partly in terms of evolutionary success (i.e., the number of sexual partners), but that they are similar in terms of psychological success. An analysis of partial correlations suggests that greed is the most likely driver of detrimental effects on evolutionary success. This research provides insights into the factors that affect the strength of greed and may open up avenues for intervention. Directions for further research are discussed.

ABSTRACT. Trust is a major element in people’s interpersonal relations and has been the subject of extensive research. Studies have highlighted its role in economic growth, trading goods, financial assets and investments between countries, the likelihood of entrepreneurship and the tendency to invest in stocks. In politics, trust promotes government efficiency, and is a major factor in people’s view of a country’s institutions as incorruptible. In education, trust affects teachers’ morale and productivity. In the fields of psychology and sociology, trust helps reduce social dilemmas. In biology, trust is a central feature in human beings’ moral systems.

One of the most common methods to measure trust experimentally is the Trust Game. This game involves two anonymous players (“the trustor” and “the trustee”) who are paired together. The trustor receives an initial endowment of money and on his/her turn, decides how much (s)he wants to transfer to the trustee to invest in him/her. More specifically, the trustor determines an amount of money between zero and the initial endowment that (s)he transfers to the trustee, knowing that each amount given will be tripled (“the multiplier”) by the experimenter before it is transferred to the trustee. The trustee then decides how much money between zero and the tripled amount (s)he wishes to return to the trustor.

In this study, we test experimentally whether the structure of people’s personal funds can affect trust. Specifically, we investigate the possible effect of several changes in the Trust Game on the level of trust: (a) transforming certain endowments into simple probabilistic endowments; (b) framing simple probabilistic endowments as complex probabilistic endowments and (c) framing complex probabilistic endowments with salient numerical anchors. We chose these changes because people often need to make decisions when they are distracted and their personal funds are limited, difficult to access or uncertain. The dimensions of uncertainty, complexity, and salient numbers are important to examine in the context of prosocial behavior. To date, previous studies on the Trust Game have not paid much attention to these factors.

In our study, 258 participants, incentivized with real money, played the role of a trustor in a Trust Game. They were asked to indicate the amount they were willing to transfer from their endowment to the trustee. Some of these endowments were certain, while other were presented as a simple lottery or a complex lottery whose outcome was derived from the outcome of another lottery. We found that uncertainty about the endowment did not affect trust regardless of the degree of the uncertainty. Furthermore, when the endowment was uncertain and complex, the presence of low numerical anchors affected the propensity to trust. However, the complex framing of the endowment did not.

This paper extends our knowledge about how behavioral biases can affect prosocial behavior, particularly trust, by introducing different methods of obtaining personal funds. This is another step towards developing a behavioral framework that can explain people’s propensity to trust, which is independent of the trustees' characteristics.

ABSTRACT. Consumers' power has been regarded as a potential force for a more equitable and environmentally healthy world. Indeed, many consumers claim that they prefer fairly traded and environmentally friendly products. This, however, is not always reflected in observed consumer choices---a phenomenon known as the attitude-behavior or intention-behavior gap. One reason for this gap might be that people are not aware of the consequences of their choices. In fact, they may even prefer to remain ignorant. Dana et al. (2007) provide experimental evidence that up to 46% of subjects do not inform themselves about the potentially negative consequences of their choices on others---even if they could do so by one mouse click. Their findings have been confirmed multiple times in the lab (See, e.g., Feiler 2014, Grossman 2014, Larson and Capra 2009 or Grossman and Van Der Weele 2017) and also have been observed in several situations in the field (See, e.g., Andreoni et al. 2017, Dellavigna et al. 2012, Ganguly and Tasoff 2017 or Knutsson et al.2013). Engaging in such willful ignorance is naïve. Consider a person who is against child labor but does not want to learn how the cocoa for their chocolate is harvested. The reason might be that if she learns that her favorite chocolate involves child labor, she has to buy pricier or less tasty chocolate. Not examining the consequences allows the person to believe that she is doing no harm by purchasing her favorite chocolate. At the same time, she might believe that she would have done the right thing, had she only known. This behavior is naïve in the sense that someone who cares enough about the wellbeing of children, informs herself if she has the suspicion that the cocoa of her chocolate was produced using child labor. Careful consideration of her choices would reveal to the person her naivety and she would no longer be able to maintain the image of a truly caring person while not informing herself.

Here, we present a simple experiment that builds on Dana et al. (2007) and tests whether a reflection of the choice to inform oneself affects this choice and leads to less harm. We find that reflection reduces ignorant behavior by 12.3 % (Result 1) and harmful behavior by 14.4 % (Result 2). In the experiment, subjects have to make a dictatorial decision between two choices. The decision affects the own payoff and the payoff of another person. The choice options are framed as 'high' and 'low'. Choosing 'high' one receives 6 ECU and choosing 'low' one receives 5 ECU. The consequences of the decision on the other player differ with the state of the world. In the first possible state, the other player receives 5 ECU if 'high' is chosen and 1 ECU if 'low' is chosen and in the second possible state of the world the payoffs are symmetrically switched. In the second possible state, the other player receives 1 ECU if 'high' is chosen and 5 ECU if 'low' is chosen. Subjects can either decide to learn the state of the world or to remain uninformed and thus behave willfully ignorant. Our design addresses the problem of 'bullying' subjects into the right behavior by asking subjects to state not only their intended strategy but also their aim before they actually implement their decision. The intervention has the form of a structured and neutrally worded questionnaire that is similar in style to that of an ex-post experimental questionnaire. By asking the subject about their fundamental aim, we give the subject freedom to define their goal themselves. They retain their `sovereignty'. Subjects can then state both aspects of their strategy, the decision to inform oneself and the decision whether or to select their (potentially harmful) favorite option. Subjects can specify conditional strategies and have room to explain how the strategy relates to their aim. Our results hint that once people perceive the step of informing oneself as part of the decision, they will less likely remain uninformed about the consequences of their actions, and thus harmful behavior could be reduced.

ABSTRACT. People’s relationships vary concerning characteristics such as intimacy, closeness, degree of interdependence, strength and type of emotions, expectations, and more (Berscheid, 1994; Berscheid & Regan, 2016). Some relationships are formal and based on contracts or written rules, whereas others are entirely unofficial, guided by feelings, and formed spontaneously (Fiske, 1992; Fiske & Haslam, 2005). One of the most widely cited and discussed typologies of social relationships was introduced by Fiske (1991, 2004; Fiske & Haslam, 2005), who proposed that people might represent relationships with others as communal sharing, equality matching, authority ranking, or market pricing. Specific relational context is associated with the norms that govern behavior insofar as how people respond to each other and regulate their actions depends on the type of relationships (Clark & Mills, 2012; Fiske, 1992). In other words, different kinds of relationships invoke different mental sets (mindsets), which in turn can influence the way people think, feel, and behave, with the most striking difference being present between communal sharing and market pricing modes (Fiske, 2004; McGraw & Tetlock, 2005). In the present work, we propose that one of the elements of our social life that can be sensitive to the mindset evoked by a specific relationship is trust. In particular, we explore how people’s willingness to trust others depends on their exposure to market-pricing vs. communal-sharing relationships. We conducted six experiments to investigate how the market mindset affects people’s trust and the psychological mechanisms behind this effect. Our results suggest that interpreting social reality through the prism of the market mindset—an archetype of rationality, independence, individualism, and self-efficiency (Fiske & Haslam, 2005; Persky, 1995; Zaki et al., 2021)—has a hampering effect on interpersonal trust. In Experiment 1 (N = 100), we included two relational mode conditions—market relations and communal relations—and found that people with the market mindset revealed lower general trust than those with the communal mindset. In the following experiments we also included a neutral condition to demonstrate that the difference between market and communal mindset conditions was caused by the detrimental effect of the market mindset rather than the beneficial effect of the communal mindset. In Experiment 2 (N = 150), we provided additional support for our prediction that priming people with the market mindset causes them to demonstrate lower general trust towards others than priming them with communal relationships. However, the results concerning the comparison of the two experimental conditions with the neutral condition were inconclusive. In Experiment 3 (N = 162), we separated the original items measuring general trust into two sets. We also decided to analyze all six items separately without forming one broad trust indicator. We demonstrated that participants in the market mindset condition declared less trust than those in the communal and neutral conditions, while the two latter conditions did not differ in the level of trust. However, we also demonstrated that the experimental manipulation of mindset affected the perceived helpfulness of other people but did not affect their declared caution or cynicism. In further experiments, we investigated potential psychological mechanisms behind this effect. In Experiment 4 (N = 367), we demonstrated that the detrimental effect of the market mindset on trust toward specific people presented on the photos might result from the fact that the market cognition is based on proportionality, which means that it is directed towards calculating and acting in accord with ratios and rates (Fiske, 2004; Rai & Fiske, 2011; Zaleskiewicz et al., 2020). In experiment 5 (N = 366), using the same measure of trust, we showed that the effect we examine might also be driven by the fact that the market mindset reduces people’s state empathy (Wang et al., 2021). Finally, in a preregistered Experiment 6 (N = 534), we showed that these two psychological mechanisms are relatively independent of each other. Furthermore, in this experiment, we used another operationalization of the dependent variable to ascertain that the effects we found were not due to the specific procedures but were robust across various methodological approaches. Here, we again presented participants with a series of human faces, but this time we employed a Smiling Faces Task (Kesebir et al., 2021). The rationale behind such a change was that the neutral faces we used in Experiments 4 and 5 do not reveal much information about trustworthiness or benevolence. Therefore, it might be relatively easy to affect trust towards such people by evoking the market mindset. Smiling faces openly reveal prosocial intentions and positive emotions, which can trigger trust and cooperation (Scharlemann et al., 2001). The fact that the salience of the market mindset could also hinder trust towards a nice, benevolent person provides a strong support for our theoretical model. In the experiments conducted within this project, we employed different manipulation methods, dependent variable measurements, and manipulation checks to avoid incidental results and to ascertain that the effects we found were not due to the specific procedures we used but are robust across various methodological approaches. Furthermore, by systematically modifying our experiments, that is, by changing manipulations, manipulation checks, and operationalizations of DVs, we were able to demonstrate high reliability (repeatability) of the effects, maximize internal and external validity, eliminate alternative explanations, and test for the mediators explaining psychological mechanisms behind the effects.

ABSTRACT. Abstract for thematic session: Consumer Willingness to Change Behavior to Mitigate Climate Change

Introduction One-quarter to one-third of all food is lost or wasted. This makes food waste a sustainability problem of considerable scale in both developed and emerging economies. Globalization and urbanization enhance the causes and thus contribute to increasing consumer-related food waste, in particular in cities. Consumer food waste stems from multiple factors, as the individual consumer, his or her social context, and the micro- and macro-environment all influence why, where, and how people dispose of still edible food. Most consumers do not want to waste food, but do so anyway, because different food-related goals are in conflict with each other. Waste occurs as a side effect of consumers’ modern lifestyle and in consumers’ and citizens’ search of a good life in relation to food, characterized by amongst others convenience, taste, health, and social interactions around meals. A primary cause of food waste in households is a lack of food capabilities. Yet, the antecedents of food capabilities and its relationship with causes of food waste across different situational contexts remain poorly understood. This paper aims to contribute to that by focusing on a consumer-centered concept yet under-researched in relation to food waste: self-efficacy. We apply self-efficacy theory in combination with convenience orientation and good provider identity in a mixed-methods study.

Method An experimental consumer online survey was conducted in Uruguay consisting of qualitative open questions, assessments of food waste incidents portrayed in typical situations in vignette form, and psychographic variables. Uruguay is a country representative of the trend of growing middle classes and globalization, as it is particularly characterized by urbanization, Western diets, and working women. 1039 respondents participated, recruited through a Facebook advertisement in exchange for a chance to win a supermarket voucher. To fairly represent the population, a random sample of 540 participants was drawn following quotas that ensured 20% low, 60% middle, and 20% high socioeconomic status.

Results Quantitative findings reveal (see Figure 1) that the sequential structure of self-efficacy explains how capabilities contribute to food waste avoidance. A good provider identity, however, increases food waste, and also turns out to further convenience orientation. Food capabilities, in turn, reduce convenience orientation. Qualitative insights (see Table 1) provide an exploration of consumer´s own accounts of their ability to influence, as well as the challenges in avoiding food waste in specific situations in their home. The findings contribute to a theory-based understanding of how consumers manage the complex issue of food waste. Findings imply that public policy and marketing should aim to strengthen consumers’ self-efficacy beliefs in order to increase the likelihood that consumers master and learn food capabilities. The social role of food and of being a good provider is important for consumers, but appears to cause food waste in its current understanding of the role – in addition, it is linked to convenience orientation, and convenience food use is connected to the increasing obesity problem. It thus seems important to work on redefining the good provider identity, in particular in social contexts, to reverse its effect on waste, but increase its impact on wellbeing and a good life.

FIGURE 1. Model of Relation between Self-Efficacy and its Effect on Food Waste, and the Role of Convenience Orientation and Good Provider Identity

Table 1. Specific Food Waste Incidents – Open Question on How Respondents Would Avoid the Incident

ABSTRACT. Climate-friendly diets are urgently called for, but emerging very slowly at best (Vermeir et al., 2020). Existing diets are defended by norms and other institutions that are difficult to change (Nyborg et al., 2016). However, the increasing moralization of the consumption of meat and especially red meat entails a social pressure to at least reduce the consumption (Cheah, Sadat Shimul, Liang, & Phau, 2020). If this counter-pressure reaches a “social tipping point”, it could lead to dramatic change towards a climate-friendly diet (Nyborg et al., 2016). Here, we investigate this process in Norway in 2019-2021.

The broad-based changes called for to obtain a stable climate are experienced as costly by most consumers, making it a social dilemma (Van Lange, Joireman, Parks, & Van Dijk, 2013). Because it threatens vital interests of various stakeholders, the process is contested at each step along the way. We tested the propositions summarized in Figure 1 by means of longitudinal survey data. The arrows represent hypotheses about direction and sign of influence.

Figure 1: The emergence and specification of climate change norms challenging red meat consumption norms and behavior

Method

Data come from a repeated web-based survey in March-April 2019, 2020 and 2021, sampling 4057, 4031 and 4897 particip¬ants from Kantar’s standing panel in Norway, of which a random half answered questions about norms regarding meat consumption. Kantar used an interactive procedure, stratifying participant invitations over time, to ensure a representative sample of the adult population.

We measured red meat consumption with the question: ‘How often do you have dinner with beef or sheep/lamb?’ Measures of the antecedent constructs in Figure 1 were responses to a set of propositions on a 5-point response scale (‘Does not match at all’ = 1 to ‘Matches very well’ = 5).

We estimated the structure of relationships between the antecedents in our model and red meat con¬sump¬tion (i.e., direct and indirect effects on red meat con¬sump¬tion) while also controlling for demographic variables using the 2019 sample and Maximum Likelihood (ML) structural equation modeling (SEM). Next, we tested the stability of identified statistical relationships using multi-group SEM. Finally, we estimated the dominating direction of influence between red meat consumption and its apparent immediate antecedents by means of cross-lagged (CL) panel analysis.

Results

The SEM model of the structure of relationships between antecedents and red meat con-sump¬tion fits the data from 2019 well. The strongest immediate antecedent of red meat consumption is perceived supportive social norms. Internalized (personal) norms about eating climate-friendly food also play a significant, and negative, role, as expected. However, the positive effect of the former is more than twice as strong as the negative effect of the latter. In addition, climate change denial has a positive impact on red-meat consumption, suggesting that the normative pressure to cut down on red meat consumption is contested.

Further back in the effect hierarchy, it is found that behavior-specific norms are derived from broader climate norms. In addition, personal norms about climate change are influenced positively by social attention towards climate change in one’s surroundings and negatively by climate change denial or skepticism.

There are also significant effects of gender (men eat more red meat then women) and age (younger Norwegians eat more red meat than older), and people in more rural areas perceive stronger social norms for red meat consumption and are also more likely to hold climate skeptical beliefs, leading to a positive indirect effect on meat consumption.

Most structural regression weights could be fixed to be unchanged over time, except four in the 2019-2020 period and two in the 2019-2021 period, without loss of fit. Hence, the structural model is stable over the studied time period.

As is generally found in CL panel analyses, the auto-regressive effects (i.e., stabilities) are considerably stronger than the CL effects. The most abstract construct – climate change denial – is the most stable (with a standardized regression weight close to .90). Perceived social norms about red meat consumption are also very stable (standardized regression weight > .80). Compared to these, personal norms about climate friendly food are less stable and red meat consumption is the least stable among the included constructs.

Positive CL effects on red meat consumption are found from perceived supporting social norms and, weaker, from climate change denial. These effects are stable over both time periods. In the 2019-2020 period, the expected negative effect of personal norms about climate friendly food is found as well, but this CL effect is non-significant in the second period.

The CL effects from behavior to the psychological antecedents are generally non-significant, confirming the assumed causal direction. The only exception is the significant and positive, but weak effect on perceived supporting social norms. Hence, there are CL effects in both directions between perceived supporting social norms and red meat consumption, but stronger from social norms to behavior than the other way round.

Conclusion

As expected, red meat consumption is strongly supported by social norms in Norway, but increasingly challenged by the societal discourse about negative climate impacts and the resulting moralizing of the diet. As consumers internalize norms about a climate-friendly diet, they reduce their red-meat consumption. Until now that process has been (too) slow, but that may change if it reaches a social “tipping point” (Nyborg et al., 2016).

References

Cheah, I., Sadat Shimul, A., Liang, J., & Phau, I. (2020). Drivers and barriers toward reducing meat consumption. Appetite, 149, 104636. doi:https://doi.org/10.1016/j.appet.2020.104636 Nyborg, K., Anderies, J. M., Dannenberg, A., Lindahl, T., Schill, C., Schlüter, M., . . . Carpenter, S. (2016). Social norms as solutions. Science, 354(6308), 42-43. Van Lange, P. A. M., Joireman, J., Parks, C. D., & Van Dijk, E. (2013). The psychology of social dilemmas: A review. Organizational Behavior and Human Decision Processes, 120(2), 125-141. doi:http://dx.doi.org/10.1016/j.obhdp.2012.11.003 Vermeir, I., Weijters, B., De Houwer, J., Geuens, M., Slabbinck, H., Spruyt, A., . . . Verbeke, W. (2020). Environmentally Sustainable Food Consumption: A Review and Research Agenda From a Goal-Directed Perspective. Frontiers in Psychology, 11(1603). doi:10.3389/fpsyg.2020.01603

ABSTRACT. Introduction Sustainability labels are acknowledged tools to drive individual consumption decisions in a more sustainable direction (Thøgersen, 2005), but are also accused of creating consumer confusion due to too much, too complex, too similar, and too ambiguous information (Chen & Chang, 2013). Meta sustainability labelling – “a label about labels” (Ge & Brewster, 2016) – has been discussed as a potential future instrument to drive more sustainable consumer choices. A meta sustainability label scheme “condenses existing product-labels and other communication measures into an overarching Sustainability message in order to enable household consumers better to align purchase decisions with Sustainable Development goals”(Dendler, 2014, p. 81). It could have the potential to integrate all dimensions of sustainability and signal the overall sustainability performance of a labeled product to the consumers. We distinguish different ways of communicating a meta sustainability label graphically on the product at the point of purchase: Either as a supplement to existing sustainability labels or by replacing them (Torma & Thøgersen, 2021). However, until now, the debate on the impact of a meta sustainability label to drive more sustainable consumer choices lacks empirical evidence.

Method We empirically tested the added value of meta sustainability labelling to drive sustainable consumer choices, over and above established single-issue labelling, such as the Fairtrade or the organic label. This was done in a comparative study of US and German consumers’ choices of bags of coffee (beans or ground coffee) varying in typical product attributes and price, in addition to sustainability labeling conditions. Survey participants in the USA (N = 518) and Germany (N = 520) were randomly assigned to these four different labelling conditions: (1) existing labelling, (2) a new, common label design, (3) existing labelling plus a meta label, and (4) a new common label design plus a meta label. The existing labelling consisted of well-known Fairtrade and organic labels. In the second condition, the same labels were presented in a new design that signals their shared belonging to a family of qualified sustainability labels. In the third condition, a meta label was added to the labels in the first condition. And in the fourth condition, the meta label was added to the new label design in the second condition.

Results Overall, in both countries, sustainability labeling appears to be of high relative importance for consumer choices of an everyday (coffee) product, compared to the other varied product characteristics. This is consistent with prior research finding that sustainability is important for consumer choices and that sustainability labeling supports sustainable consumer choices, both regarding coffee (e.g., Thøgersen & Nielsen, 2016) and other products (Ní Choisdealbha & Lunn, 2020). In the USA, it was found that the meta sustainability labeling outperformed the traditional labeling (confirming Hypothesis 1). However, whereas changing qualified sustainability labels into a common design line led to a significant increase in the utility of sustainability labeling (confirming Hypothesis 1a), adding a meta label had no effects (rejecting Hypothesis 1b). In Germany, at the aggregate level, the different labeling conditions did not lead to significantly different utilities of labeling. However, we found that, in Germany, the effect of the different label conditions differs between segments of consumers defined in terms of labeling confusion, knowledge, and attitudes (Hypothesis 2). Hence, overall, it appears that a graphically communicated meta sustainability labeling scheme outperforms traditional labeling among consumers in general in the USA, and among some segments of German consumers, in particular those who are most confused by the existing sustainability labeling landscape. In addition, German consumers who evaluate sustainability labeling more favorably or feel more knowledgeable about sustainability labeling also appreciate some of the suggested meta labeling improvements. These moderator effects were not found in the US (rejecting Hypothesis 2 in the US).

Discussion Meta sustainability labeling is a potential tool to facilitate quality information about sustainability in an overarching, interconnected, and easily comprehensible way. However, it is crucial that consumers value the meta sustainability labeling and that it facilitates more sustainable consumer choices in practice. This study has delivered the first empirical evidence on the perceived value and usefulness of a meta sustainability label to drive sustainable consumer choices. The new, common labels were designed based on the graphical branding of UN’s Sustainable Development Goals, which makes this a practically relevant case, illustrating meta sustainability labelling that could become implemented globally. There are also already companies that certify compliance with the SDGs, which is a prerequisite for using it for labeling and documents the practical relevance of using exactly this graphical expression for sustainability meta labeling. 

Figure 1. The design single issue labels and the meta label build upon the SDGs branding (could bot be inserted in this abstract format)

Figure 2. Theoretical model (could bot be inserted in this abstract format)

References:

Chen, Y.-S., & Chang, C.-H. (2013). Greenwash and Green Trust: The Mediation Effects of Green Consumer Confusion and Green Perceived Risk. Journal of Business Ethics, 114, 489-500. Dendler, L. (2014). Sustainability Meta Labelling: an effective measure to facilitate more sustainable consumption and production? Journal of Cleaner Production, 63, 74-83. Ge, L., & Brewster, C. A. (2016). Informational institutions in the agrifood sector: meta-information and meta-governance of environmental sustainability. Current Opinion in Environmental Sustainability, 18, 73-81. Ní Choisdealbha, Á., & Lunn, P. D. (2020). Green and simple: Disclosures on eco-labels interact with situational constraints in consumer choice. Journal of Consumer Policy, 43, 699-722. Thøgersen, J. (2005). Consumer behaviour and the environment: Which role for information. In S. Krarup & C. S. Russell (Eds.), Environment, information and consumer behaviour (pp. 51-63). Cheltenham, UK: Edward Elgar. Thøgersen, J., & Nielsen, K. S. (2016). A better carbon footprint label. Journal of Cleaner Production, 125, 86-94. Torma, G., & Thøgersen, J. (2021). A Systematic Literature Review on Meta Sustainability Labeling -What Do We (not) Know? Journal of Cleaner Production, 293, 126194.

ABSTRACT. Introduction Food contributes to approx. 20-35% of GHG emissions (Foley et al., 2011), and animal-based foods/diets cause considerably higher GHG emissions (Bruno et al., 2019). Shifting human consumption to predominantly plant-rich diets is highly impactful (Willett et al., 2019). Thus, consumers’ individual choices are crucial in the transition to more sustainable food systems. However, ‘sustainable food’ or ‘sustainable diets’ remains a niche (Reipurth et al., 2019). In addition, the polarised discussion on ‘plant’ versus ‘meat’ can lead to more pronounced perceived barriers to change (Cheah, et al., 2020). Therefore, it is important to explore how consumer perceptions and behaviours across the population differ, and how consumer segments are characterised in their attitudes, beliefs and perceptions of the issue. In this paper, we investigate consumer perceptions and behaviours towards plant-based foods and diets, across consumer lifestyle groups. Of particular interest is the perceived conflict concerning plant-based eating and meat production.

Method A representative sample of Danish 1126 consumers, in May 2021, was subjected to a two-step cluster analysis procedure based on six selected psychographic variables related to the issue at hand (perception of conflict about plant vs. meat, number of people one knows who have reduced their meat consumption, importance of other persons who reduce meat consumption, believes about the benefit of reducing meat consumption, beliefs about animal production, environmental concern). We conducted a hierarchical cluster analysis to identify the appropriate number, followed by a K-Means cluster analysis to establish the clusters. Then, we characterized clusters with ANOVA on a range of sociodemographic variables (rural vs. city, education level, age, gender) and relevant behaviours and attitudes (their current dietary style, cooking capabilities, Danish nationalism).

Results Results show that consumers generally have a fair amount of trust in plant-based brands (M=4.72, SD=1.46), yet low brand awareness (i.e., low free brand recall capabilities). Moreover, few people currently consume plant-based replacement products (e.g., plant-based milk or meat) on a regular basis (i.e., 52.15% report never having used them), and most consumers still consider themselves meat eaters (over 75%, whereas only 1.9% or 1.3% consider themselves vegetarian or vegan, respectively). Nevertheless, many people report eating vegetarian/vegan meals and dinners: 27.1% and 16.5% of respondents indicate eating vegetarian/vegan lunches or dinners, respectively, more than half the days of the week. The clustering procedure resulted in eight different consumer segments (see table 1). Most respondents saw benefits in reducing meat consumption, but respondents in two segments (nr 3 and 5) particularly disagreed. Several clusters agreed that meat production can be organised sustainably (5, 6 and 7), but respondents in two clusters disagreed (1 and 3). Consumers in two of the clusters knew relatively many who have already reduced meat consumption (1 and 6), though what these others think of them was particularly important for cluster 8. Cluster 5 perceived the most conflict, cluster 3 the lowest; meanwhile, cluster 2, 3 and 5 had lower environmental concern, cluster 1, 6 and 7 high environmental concern.

Discussion These findings show that there are marked segments at the extremes – e.g. cluster 1 eats most plant-based and is more likely young, female and city-based, while segment 5 is more likely male, older, rurally located and lower educated. However, there are a range of segments with different patterns in-between, as for example cluster 7, which has a high share of meat eaters but also high environmental concern and believe that animal production can be organised sustainably, though lack peers who have reduced meat consumption. We conclude that these patterns of the segments in-between the extreme niches on both sides of the spectrum allow to derive recommendation for targeted approaches to support different kind of sustainable food transitions patterns. These recommendations could be valuable for value chain actors, to help design marketing approaches adequately for these different groups, which in accordance to their stance may prefer different communicational framing or different products. Future surveys within this project will allow to observe if the differences in perception, behaviour and perceived barriers differently shift in the identified consumer segments, and which segments change their stance to the issue most throughout the project’s timeline.   -- Table 1: please see attached abstract file for the table

References: Bruno, M., et al. (2019). The carbon footprint of Danish diets. Climatic Change, 156, 489-507. Cheah, I., et al. (2020). Drivers and barriers toward reducing meat consumption. Appetite, 149, 1-9. Foley, J. A., et al. (2011). Solutions for a cultivated planet. Nature, 478(7369), 337-342. Reipurth, M. F., et al. (2019). Barriers and facilitators towards adopting a more plant-based diet in a sample of Danish consumers. Food quality and preference, 73, 288-292. Willett, W., et al. (2019). Food in the Anthropocene: the EAT–Lancet Commission on healthy diets from sustainable food systems. The Lancet, 393, 447-492.

-Note that this abstract is part of a special session submission by Prof. John Thøgersen, under the session title 'Changing Consumer Behavior to Mitigate Climate Change'.

ABSTRACT. Title: Bundling Violence: Trading off combinations of violent acts Authors: Magda Osman1, Isabelle Mareschal2, Emily Hannon2 Affiliations:1 Centre for Science and Policy, University of Cambridge, 10 Trumpington Street, Cambridge, CB21QA; 2Biological and Experimental Psychology, School of Biological and Chemical Sciences, Queen Mary University of London, Mile End Road, London, E14NS Background: Violence risk assessments are used in a variety of settings (e.g. forensic, psychiatric, public) to determine current and future risk of violence which is then used by practitioners and policy makers to determine future demands on services. In fact, risk assessments feed into analyses that make up reports, such as those published by the U.K.’s Government agency HM Prison and Probation service and the annual National Offender Management Service (NOMS) for which statistics are presented on levels of violence in prisons (e.g. NOMS Annual Report and Accounts, 2016-2017). In the case of violence in prisons, violence risk assessments can reveal the extent to which, between prisons, and across time, violence is on the rise. Critically, violent outbreaks occur in which different combinations of violence acts within each outbreak are recorded. In order to determine future demands on prison services, what approach should be taken to assess if an outbreak in one prison is less than, equal to, or more violent overall than an outbreak in another prison? This is a particularly challenging question to answer because the aggregate score (total violence score) will significantly vary depending on how the severity of violent acts is taken into account. In fact it varies to such an extent that the inferences drawn can be entirely opposite across risk assessments. For instance, some risk assessments (e.g., those used by Center for Disease Control and Prevention U.S. [CDC]) treat all violent acts as equally violent (ignoring differences between the severity) so comparing between outbreaks merely involves summing the violent acts in each and determining which has the greater of violent acts. Some risk assessments (e.g. those used by the World Health Organisation [WHO]), rank the violent acts, which means taking account of the variation in severity of violent acts (e.g. High, Medium, low), and then compares, in this example, prison outbreaks, by the total number of the most severe violent acts committed. A third approach to risk assessment (e.g. those used by the Global Peace Index [GPI]) involves the application of a weighted sum. This means that the variation in severity of violent acts is incorporated into the analysis, along with assigning different weights to the class of violent acts falling into each level of severity. Purpose: The purpose of this study is to examine how non-experts typically rank combinations of violent acts, to gauge, via a non-expert audience, which of the different conceptualisations of quantifying violence is the most common, by which we mean which is the most psychologically intuitive to non-experts. Experiments: In the present study we presented non-experts with putative violent bundles (combinations of violent acts), having taken place in a fictional prison setting, in order to determine the most common way in which people intuitively aggregate violent acts and make trade-offs. The study comprised 4 experiments. Experiment 1, 2, 3 and 4 involved UK samples (Experiment 1: UK N = 100; Experiment 2: UK N = 104; Experiment 3: UK N = 101; Experiment 4: UK N = 105), each were born in the UK, were UK residents, and their first language is English. All four experiments included two types of tasks: 1) Forced-choice outbreak task and 2) Missing Value outbreak task. The tasks presented participants with descriptions of two fictitious prisons, and they were to adopt the role of Governor of one of them. In the Forced choice task, participants were presented with bundles of violent acts from a prison outbreak that had occurred (e.g. Prison 1 Outbreak [High violent acts = 8, Low violent acts 2], Prison 2 Outbreak [High violent acts = 4, Low violent acts = 7], and were asked to report whether the overall violence of the outbreak in Prison 2 was equal to, lower, or higher than that of Prison 1. The missing value outbreak task required that participants enter a value into one of the bundles (e.g. Prison 1 Outbreak [High violent acts = X, Low violent acts 2] in order to make the total violence in the prison outbreak, lower than, equal to, or more violent than that of a comparator bundle. Through these tasks it is possible to determine whether participants generate responses in line with taking severity of violent acts into account when performing their trade-offs, or whether they are doing so without sensitivity to the severity of violent acts. Main findings and conclusions: Overall, looking across the findings from all four experiments, the proportion of participants responding in line with an approach that indicates aggregation of violent acts is sensitive to level of violence was higher in the Forced-choice outbreak task (60%) compared to the Missing values outbreak task (35%), but nevertheless, for both types of task, this pattern of response was by far the most common compared with a response pattern that indicated aggregation by insensitivity to the severity of violent acts (Forced choice task, 16%; Missing values task, 2%).

Thus, based on our findings across four experiments, this study is the first to be able to show that when assessed, people commonly make comparisons between bundles of violent acts in which they aggregate the violent acts (high and Low) while taking into account the level of violence of the acts, rather than focusing on simply aggregating the total number of violent acts irrespective of level of severity.

ABSTRACT. Violation of Covid-19 regulations are only one example of seemingly small violations that accumulate fast, with potentially dire social consequences. The high cost of close monitoring and severe sanctions render full enforcements impractical, usually leading policy makers to prioritize either the probability of inspection or punishment severity. The current paper examines the effectiveness of two compensatory policies: High probability of Inspection with Low Severity of fine (HILS) vs. Low probability of Inspection with High Severity of fine (LIHS).

Different lines of research offer competing predictions regarding the relative importance of each component: For example, using common one-shot, descriptive settings, findings from experimental economics support the superiority of severe punishments, whereas findings from behavioral ethics highlight the role of internal rather than external enforcement. However, these settings are estranged from real-life environments in which learning about the external enforcement policy naturally occurs via repeated experience. In contrast, research on repeated decisions from experience indicates that choice behavior is more sensitive to the frequency of experienced outcomes than to their magnitude. Importantly, this line of research shows that when people learn from experience, they tend to behave as though they underweight rare outcomes. In the context of violation behaviors, these findings imply that rare large fines may not have the intended outcome, and frequent small fines should be more effective at decreasing violations in the long run.

In a lab pilot and 3 online studies (Ntotal = 816), we tested these competing predictions by assessing which enforcement strategy is more effective in settings where people can repeatedly engage in small violations. We utilized the dots task, a repeated perceptual task that creates a conflict between accuracy and profit maximization. In the lab pilot participants engaged in both the HILS and the LIHS policies (within-subject) and in the three main studies, the enforcement policy was manipulated between participants, but only after playing a no-enforcement game to measure baseline violation rates.

Controlling for expected value, we found that a policy combining a high probability of inspection with a low severity of fines (HILS) was more effective than an economically equivalent policy that combined a low probability of inspection with a high severity of fines (LIHS). The advantage of prioritizing inspection frequency over punishment severity (HILS over LIHS) was greater for participants who, at baseline (in the absence of enforcement), had a higher violation rate. Consistent with studies of decisions from experience, frequent enforcement with small fines was more effective than rare severe fines even when we announced the severity of the fine in advance to boost deterrence. In addition, in line with the phenomenon of underweighting of rare events, the effect was stronger when the probability of inspection was rarer (as in most real-life inspection probabilities) and was eliminated under moderate inspection probabilities.

Interestingly, our findings are consistent with empirical studies showing that the severity of criminal sanctions is not correlated with the level of crime in society, whereas enforcement prevalence has consistently been found to be related to crime rates. Our controlled experimental setting demonstrates a causal link between enforcement prevalence and violation rates for a general (noncriminal) population and for small-scale violations and punishments. Theoretical implications will be discussed.

From a practical standpoint, our findings suggest that when the inspection rate is low, policymakers should prioritize increasing the frequency of inspections over the severity of punishments. Since, in the real world, inspection rates are commonly very low, our findings suggest that increasing inspection rates even by as little as a few percentage points could be highly effective in reducing violations. For example, in our studies, an inspection rate of 6% reduced violation rates by 12%, while an inspection rate of 10% with smaller fines reduced violation rates by more than 38%. Although increasing the magnitude of fines was often considered as less costly, recent advances in technology and the increasing usage of artificial intelligence algorithms enable more effective monitoring at significantly lower costs. Moreover, large fines could result in a perception of unfairness and consequently reduce the probability of detection, which, according to our results, is the key factor. In a similar vein, empirical crime researchers have argued that programs focusing on increasing punishment severity entail greater costs in their implementation. Thus, while many regulators in the current COVID-19 pandemic have publicly called to increase the magnitude of fines, our findings strongly suggest that “gentle rule enforcement,” which includes smaller punishments with a higher probability, would be more effective in reducing violation rates, especially for high offenders, the target population of any enforcement policy.

ABSTRACT. Objectives and Motivation: Participation rates in gambling activities have increased in the last decades (Delfabbro & King, 2021). So has done the rate of problem gambling, a potentially addictive behavior affecting around 1-5 percent of the world-wide population (Calado & Griffiths, 2016), which has been further exacerbated by contemporary social issues. In Italy, a study conducted by the National Research Council (Cnr) in 2017 revealed that almost 43% of Italian population had gambled at least once in the previous 12 months, establishing a negative trend compared to the percentage found in 2014 (Cerrai et al., 2017). The study also showed an increasing number of problem gamblers, four times higher than 2007. Although a general drop was registered during the lockdown in 2020 (Lugo et al., 2021), the practice of gambling rose again during the following period. In the light of the above, it is not surprising that gambling practice and problem gambling have gained raising attention in the last years. As the awareness of such issues grows, so does the need to plan and design programs able to prevent problem gambling and support problem gamblers. Gaining a better understanding of players’ profiles is then pivotal, in order to design customized interventions able to capture and consider the specificities of players in terms of behavioral habits, psychological profile and socio-demographic background. Indeed, previous research suggested several different determinants and risk factors associated with problem gambling, among which family and peer relationships (Gainsbury et al. 2014), lower education and unemployment (Effertz et al., 2018), as well as a prevalence of male gender (Volberg, 1994). Furthermore, players might be driven by different experiences (e.g., coping with negative feelings, or as an opportunity to generate positive ones) and different motivations, such as winning money, intellectual challenge, and self-worth (Binde, 2013; Flack & Morris, 2015; Francis et al., 2015). Based on such premises, the objective of the present study is to segment players presenting different levels of problem gambling by identifying specific behavioral, psychosocial, and socio-demographic profiles. Specifically, the research aims to profile the riskiest players, so to avoid flattening their differences and to catch characteristics able to support the design of targeted promotion and prevention programs. Methods: Survey data were collected between December 2021 and January 2022 by Sisal, a company operating in the gambling industry, aiming to achieve a deeper knowledge of their players in order to better design responsible and safe gambling programs and to define mitigation actions to prevent the problem gambling phenomenon. The target population was Sisal customer base, active in the online setting, and the total sample included n = 11.968 participants. The PGSI (Problem Gambling Severity Index) was employed as a measure of problem gambling, whose psychometric properties were also confirmed in the Italian population (Barbaranelli et al., 2013). A k-means cluster analysis was carried out with IBM SPSS 27 including 9% of respondents, corresponding to the participants more at risk (i.e., subjects who scored 8 or above in the Problem Gambling Severity Index; n = 1118). A broad set of indicators was included. Among them, several psychological variables were considered: the motivational profile of gamblers, with the Modified Gambling Motivation Scale (Shinaprayoon et al., 2018); the risk-seeking propensity and the perceived severity of gambling, through ad hoc items; the emotional state experienced by players before and after gambling, using a semantic differential scale. Furthermore, behavioral variables were meant to evaluate the number of games played, the frequency of gambling, and the amount of money spent on gambling. Last, demographic details and risk factors (e.g., smoking and alcohol consumption) were included in the survey. Results: Three clusters were identified. Cluster 1 (n = 515, 46%) is composed by players with the lowest PGSI score compared to the other two groups, and whose main motivation to gamble is to improve their mood and to feel better. Indeed, it seems that gambling might have a relaxing effect rather than exciting. Cluster 2 (n = 332, 30%) includes players who are aware of the severity of extreme gambling activities and who experience the most negative feelings after gambling, since their main motivation to gamble is the prospective win, which is frequently disregarded. Participants in this cluster are more likely to have a lower socioeconomic status and higher difficulties in saving money. Cluster 3 (n = 271, 24%) includes participants who live the game as a means to test themselves, without problematizing their own behavior. These players scored higher on the PGSI score and are more likely to have higher income and savings. Conclusion: The present research showed that players with high level of risk do not represent a unique homogeneous group, whether they present profiles deeply differentiated in terms of motivational drivers and psychological aspects, behavioral propensities and preferences, as well as socio-demographic features. Therefore, the results of this study might be useful for designing targeted prevention programs able to effectively address players with specific needs and features, and responsive to diverse kind of interventions.

ABSTRACT. Incentive-based intervention programs aimed at promoting healthy eating behaviors usually focus on incentivizing repeating the desired behavior. Unfortunately, even when effective, these interventions are often short-lived and do not lead to a lasting behavioral change. We present a new type of intervention program focused on incentivizing exploration of new healthy alternatives rather than incentivizing repeated healthy behaviors. This intervention aims to help participants find long-lasting “personal treasures” —new foods that are both healthy and tasty for them. Our field study included a final sample of 48 students with low or medium daily consumption of fresh salads. Participants in the control group received a fixed payment for completing the program, while the participants in the incentivized exploration group received a lower fixed fee for completing the task and a bonus for each new salad they tried. Results show that participants in the incentivized exploration group reported eating more salads even 1 year after the program ended compared to the participants in the control group. Though preliminary, our results paint a promising picture for the successful application of incentivizing exploration interventions to promote healthy lifestyle.

ABSTRACT. Customer Privacy: Investigating the impact of data capturing techniques on consumer information disclosure

1. INTRODUCTION AND HYPOTHESES

The present paper examines how different question presentation techniques influence voluntary disclosure of private information by consumers focusing on the effects of three questionnaire design factors namely Question Sequence (Moon, 2000), Comparative Nature (Acquisti et al. 2012) and Dyadic Relationships (Zimmer et al., 2010), and their impact on information disclosure.

The literature notes that Question Sequence (QS) which is the presentation order of questions in a questionnaire, affects consumer responses yet literature seems ambivalent on how question sequence effects influence voluntary disclosure of private information. Moon (2000) supports that an ascending sequence of privacy-invasive questions (the least privacy-invasive questions at the start followed by questions with higher degree of privacy-invasiveness) maximises information disclosure suggesting that early, easy-to-answer questions warm up respondents for greater information disclosure. On the contrary, Acquisti et al. (2012) suggest that a descending question sequence (most intrusive questions at the start) produces better results due to respondents feeling less and less threatened by the lessening degree of privacy invasiveness. Focusing on the Ascending order (Moon, 2000; Zimmer et al., 2010), H1 indicates:

H1: An Ascending Sequence of privacy invasive questions will increase the Overall Actual Disclosure of respondents.

Comparative Nature (CN) is the innate tendency of individuals to compare themselves to other individuals and emanates from the basic instinct of imitation and mimicry. Acquisti et al. (2012; 2015) suggest that respondents feel less uncomfortable to disclose sensitive information when they are led to believe that others have provided the same or similar information. Synthesis of perceptions that the majority of others disclosed the asked information, leads to alleviation of the individual’s perceived risks from a potential disclosure demonstrating compliance with the masses’ choice consequently leading to divulgence (Acquisti et al. 2012). Based on the latter, H2 indicates:

H2: A High-level Comparative Nature of privacy invasive questions will increase the Overall Actual Disclosure of respondents.

Zimmer et al. (2010) specify that Dyadic Relationships (DR) refer to the ability to build a relationship between the two parties (dyadic) involved in a survey (the party conducting the survey gives out some piece of information with every question) and thereby foster voluntary disclosure of information due to the feeling of disclosure reciprocity. They differentiate between three types of dyadic relationships (reasoned, unreasoned, non-dyadic). In the reasoned dyadic the organisation provides respondents with direct information as to how the acquired data will be used by the organisation. Zimmer et al. (2010) note that reasoned dyadic relationship seems to increase intentions for divulgence but has not been tested towards actual divulgence. Bridging this gap H3 is formulated as:

H3: Fostering Reasoned Dyadic Relationships will increase the Overall Actual Disclosure of respondents.

None of the literature cited thus far has actually examined the combined effects of all three instrumental factors to identify interactive effects. Zimmer et al. (2010; p.403) recommend the examination of the “synergistic influence of dyadic relationships with other disclosure-influencing concepts” on voluntary disclosure of information. Similarly, Acquisti et al. (2012) propose triangulations between comparative statements and the sequential structure of questionnaires. The present paper addresses these views and the respective gap through H4:

H4: The combined utilisation of an Ascending Sequence of privacy invasive questions, high-level Comparative Nature and a Reasoned Dyadic Relationship will increase the Overall Actual Disclosure of respondents.

2. RESEARCH DESIGN AND FINDINGS

The main study focused on a 3x3x3 between-subjects quasi-experimental design utilising the three main instrumental factors (Comparative Nature, Questions Sequence, Dyadic relationships) and the three respective conditions of each. Ultimately this led to 27 different questionnaires being designed using Qualtrics and administered to 1276 respondents with 40 to 60 participants being included randomly in each condition. Each version utilised 18 privacy related questions being asked in 27 different ways. Participants were asked to answer the 18 questions ultimately leading to information disclosure. A percentage was generated for each condition based on how many questions participants answered and how many they preferred not to disclose. This ultimately led to Overall Actual Disclosure (OAD) which was treated as the dependent variable. A 3-way ANOVA was conducted with OAD as the dependent and CN, QS and DR as the independent variables.

The present paper represents the first attempt to examine both the individual and synergistic effects of instrumental factors that drive information disclosure. The Ascending order of invasiveness generated significantly higher Overall Actual Disclosure (OAD) than the other two conditions of the Questions Sequence variable. Similarly, High Comparative Nature yielded the highest OAD for the Comparative Nature variable while the Reasoned dyadic relationship was also found to be the most effective compared to the Unreasoned and Non-dyadic relationships. H1, H2 and H3 were supported and in alignment with the recommendations of Moon (2000) for H1, Acquisti et al. (2012) for H2 and Zimmer et al. (2010) for H3. Significant differences were also found in certain dual and triple combinations providing partial support of H4.

Capitalising on the results of the present study, companies can utilise short statements (resembling reasoned dyadic relationships) prior to each attempt to capture customer data explaining the exact use and need for that data. This is to complement extended documents of privacy policies. When asking for the disclosure of private information, results also demonstrate that less privacy invasive questions need to precede ones that seek to capture more sensitive information as the ascending privacy-invasiveness approach decreases the chance for individuals to ‘clam-up’.

4. REFERENCES

Acquisti A., Brandimarte L., and Loewenstein G. (2015). Privacy and human behaviour in the age of information. Science, 347(6221), 509-514.

Acquisti A., John L.K., and Loewenstein G. (2012). The impact of relative standards on the propensity of disclosure. Journal of Marketing Research, 48, 160-174.

Moon Y. (2000). Intimate exchanges: using computers to elicit self-disclosure from Consumers. Journal of Consumer Research, 26(4), 323–339.

Zimmer J.C., Arsal R., Al-Marzouq M,. Moore D., and Grover V. (2010). Knowing your customer: Using a reciprocal relationship to enhance voluntary information disclosure. Decision support systems, 48, 395-406.

ABSTRACT. EXTENDED ABSTRACT Due to the technological developments in the collection, storage, analysis, and usage of information, which altogether are referred to as “big data”, information privacy is becoming an increasingly bigger concern for consumers, policymakers, and scholars alike (Acquisti, Brandimarte, & Loewenstein 2015). In 2016, TRUSTe/National Cyber Security Alliance Consumer Privacy Index revealed that Americans were more worried about their data privacy than about losing their main source of income. However, despite consumers’ self-reported concern about their information privacy, and despite the regulations and provisions that increase consumers’ control over their personal data, it appears that consumers voluntarily provide personal, sensitive, and sometimes even incriminating information about themselves, especially on social media – a phenomenon termed “the privacy paradox” (Norberg, Horne, & Horne 2007). The questions of when and why consumers reveal private information about themselves (i.e., self-disclose) are important not only within consumer research but also across several other domains, including law (e.g., fraud and data leakages), criminology (e.g., crime confession), insurance (e.g., risk calculations), healthcare (e.g., patients’ information disclosure to doctors), psychology (e.g., trauma treatment). Therefore, understanding the conditions under which consumers voluntarily disclose sensitive, private, or unfavorable information about themselves is key for both theory and practice (Alter & Oppenheimer 2009). The current research argues that one important driver of consumers’ self-disclosure is psychological stress, and that the effect of stress on self-disclosure can be eliminated by activating self-presentation goals. Self-presentation denotes the process by which individuals attempt to control the impressions that others form of them (Schlenker 1980). Although the desire to present oneself positively is innate (Goffman 1959), self-presentation efforts are not constant, but vary across individuals and situations (Tyler 2012). When self-presentations are salient, individuals attempt to minimize the risk of portraying a suboptimal image (Leary & Kowalski 1990). Since self-disclosure entails a self-presentational component, the extent to which individuals are willing to self-disclose personal information varies according to the salience of their self-presentation goals (Tyler 2012). Thus, the more salient their self-presentation goals are, the more likely individuals are to conceal private information (Ellison, Heino, & Gibbs 2006). Drawing on this body of work, we propose that the activation of individuals’ self-presentation efforts reduces the effect of stress on self-disclosure by encouraging individuals to protect their self-image. Hence, we hypothesize that when consumers are under high (vs. low) stress, their self-disclosure will increase, but only when their self-presentations motives are not salient. Results from three studies using different manipulations of stress and different research settings (i.e., field, laboratory, and online) lend support to these hypotheses. In Study 1, using a naturalistic stress manipulation, we tested the basic prediction that stress amplifies individuals’ tendency to divulge sensitive information about themselves. We ran a quasi-experimental field study that capitalized on a public oral exam with a one factor design (stress: high vs. low). Students were either part of a group that had to take the oral exam or part of a neutral group since the oral exam occured at two different moments. This enabled us to contrast two comparable groups with different levels of naturally occurring stress: high stress (driven by the oral exam) vs. low stress (no oral exam). We measured self-disclosure by asking participants to answer a set of fifteen highly intrusive questions (John, Acquisti, and Loewenstein 2011). The questions were pretested on this particular graduate student population (e.g., “Have you every shoplifted?”, “Have you ever used a fake ID?”, “Have you ever smoked marijuana?”). Our dependent variable comprised the number of “Yes” responses. Results revealed the students who took the oral exam (high stress condition) answered a higher number of intrusive questions affirmatively than the students who did not take the oral exam (low stress condition) (F(1, 36) = 4.32, p = .045, d = .67). Therefore, Study 1 showed that a naturally occurring stressful situation significantly increased self-disclosure. Study 2 employed a more controlled laboratory environment and a different stress manipulation. Students were invited to play The Mental Game, which was based on The Montreal Imaging Stress Task, an established protocol for inducing moderate psychological stress (Dedovic et al. 2005). The Mental Game consisted of a series of computerized mental arithmetic tasks, displayed on the computer screen. The arithmetic tasks were the same across the high and low stress conditions. However, participants in the high stress condition had to do it under time pressure (i.e., 10 seconds/question) and “correct” or “incorrect” feedback. Akin to Study 1, in the second part of the study, self-disclosure was measured with fifteen highly intrusive questions adapted from John et al. (2011). Study 3 employed a 2 (stress: low vs. high) by 3 (prime: none vs. neutral vs. self-presentation) factorial experimental design. Akin to Study 2, the Mental Game was used to manipulate psychological stress. Self-presentation motives were primed using Tyler’s (2012) scrambled sentence task. Participants were randomly assigned to one of the following three conditions: the no-priming condition (akin to Study 2), the impression prime condition, or the neutral prime condition. Participants in the neutral and impression prime conditions were provided with ten sets of scrambled five words that were either neutral or accordingly, impression-related. Following the self-disclosure measure, participants assessed their privacy concerns (2 items: “To what extent are you concerned with your online privacy?” and “How important is online privacy to you?” from 1 = “Not at all” to 5 = “Very much”; α = .77, M = 3.87, SD = 0.87), and were debriefed. As predicted, results revealed that activating a self-presentation goal is a boundary condition for the positive effect of stress on disclosure such that priming the participants’ self-presentations motives eliminated the effect of stress on self-disclosure. Study 3 also ruled out the potential role of privacy concerns. Considered jointly, the present studies support our central hypothesis that stress enhances self-disclosure. In an ongoing study, we aim to shed light on the underlying processes. Overall, we augment the current understanding of stress and contribute to the disclosure literature by providing insights for organizations that value customer disclosure (i.e., insurance companies).

ABSTRACT. COVID-19 has made tourism demand highly unpredictable. Regular forecasting techniques using time series data is of little worth if the past does not resemble the future. Collective intelligence studies show that crowds can be more accurate than experts when making judgmental forecasts (Page 2007). As formulated by Larrick, Mannes, and Soll (2012) «Combining judgements takes the individual imperfection and smoothes the rough edges to isolate the collective’s view of the truth.»

The objectives of this study are twofold. First, we test if crowds of tourist industry experts who forecasts tourist demand outperform their own individual predictions. Second, we test if the collective forecast can be improved upon by applying the pivoting technique of Palley and Soll (2009). Palley and Soll (2009) methodological approach addresses the presumed misweighting between private and shared information.

Using crowd predictions is particularly relevant when knowledge is dispersed, since this represents as a situation where no has the complete information, but rather that the pieces of information are scattered around among many individuals. A challenge in crowd forecasting when using standard aggregation methods like the mean or median is that they tend to overweight and underweight different types of information. In particular, shared information tends to be overweighted while private information underweighted. Palley and Soll (2009) introduced a pivoting technique to improve crowd forecasts by mitigating this information issue.

In tourism, few if any has a complete picture of the different tourist segments’ demand. In a uncertain pandemic environment, sophisticated and data intensive methods can compensate to some degree for the shortcomings of standard forecasting models (e.g., see Fotiadis, Polyzos, & Huan (2021). However, tourist destinations and firms often lack resources and capabilities to employ such sophisticated data analyses to forecast tourism demand

A different and more feasible approach can be to build forecasts using tourist experts’ intuitive judgments. Chase and Simon (1973) claimed that intuition is recognition, i.e., that our intuition reflects patterns we recognize from our experience and knowledge. Kahneman and Klein (2009) accord with this notion of intuition and stated that the quality of intuitive judgment depends on: 1) the predictability of the environment in which the judgment is made, and of 2) the individual’s opportunity to learn the regularities of that environment

Our hypothesis is that a diverse group of tourist experts who observe and learn about the latest development in their business segments may increase the quality of forecasts of aggregated tourist demand indicators like commercial accommodation guest nights (i.e., hotels, hostels, and camping) and AirBnB sales, in addition to air passenger traffic in the destination’s airport.

Experts can be operationally defined as those who have been recognized within their profession as having the necessary skills and abilities to perform at the highest level” (p. 255, Shanteau, 1992). In this study, we initially treat everybody in the tourism industry who is member of DMO Region Stavanger as experts. Shared-information bias when aggregating individual responses can be alleviated by asking an additional question: “What do you think will be the average judgment of the others n-1 judges?». Experts with private information tend to expect other experts to judge differently than themselves. A pivoting technique exploit these gaps by estimating private information and giving it more weight (Palley & Soll, 2019).

In the study period, around 420 members of Region Stavanger (the regional DMO) received the survey. 18 respondents first round and 31 respondents second round. Each round respondents made a total of 18 predictions about tourism demand in the Stavanger region. Demand of commercial accommodation, AirBnB bookings, and air passenger travel. Nine of those 18 predictions were based on own beliefs, while the other nine represented the respondents’ beliefs about other experts’ predictions.

Preliminary treatment of the data indicates wisdom of tourist experts. Wisdom of crowds as defined by Page (2007) is present in all seven prediction exercises, but only marginally so in some of them. The presence of wisdom of tourist experts appears to be strongly linked to the conditions identified by Kahneman and Klein (2009), i.e., predictability of environment and learning opportunity. Pivot technique appears to be relatively effective in improving accuracy for accommodation and AirBnB demand, but counterproductive for air passenger demand. Conditions of expertise appear to matter, as the tourist experts appear to have little expertise about air passenger traffic. Additional results remain from the last sample and, in addition, we plan to boostrap results to better describe performance of crowd judgements.

ABSTRACT. This paper uses transactional banking data from more than 300,000 U.K. residents to analyse consumption responses to income. Managing discretionary spending is important for consumers, but the available evidence suggests it is challenging for majority of individuals. Moreover, policy makers are interested in how one-off payments to individuals are consumed. In the present research we examine the extent to which consumers display a payday effect in which their discretionary spending predictably spikes above average on or shortly after payday, and whether using a personal finance app is associated with a reduction in these spikes. To accomplish this, we use a longitudinal data set that includes consumers’ spending both before and after they start using an app that helps them track income and expenses. The data reveal that discretionary spending can be reliably predicted by the arrival of payday: the average payday effect before consumers start using the app is 56% to 83% across spending categories such as discretionary items, groceries, fuel, clothing and online shopping, which is substantially larger than previously thought. Conducting the payday analysis grouped by the demographic variables age, salary and gender further illustrated differences in the magnitude of the payday effect. Confirming prior evidence regarding hand-to-mouth consumption (e.g., Campbell & Mankiw, 1989), we find that those individuals with low income (less than £1,748 net income per month) display behaviour characteristics of the payday effect to a far larger extent than those in the higher income groups. One assumption is that the lack of availability of funds is not the only issue that drives the payday effect but rather the lack of financial literacy (e.g., Shah et al., 2012; Fernandes et al., 2014) to make planned decisions and smooth consumption over the course of the month. However, after users frequently checked their financial situation, measured through detailed login information, they significantly change their spending pattern. We find that users only spend 28% more on discretionary items one day after payday if they log in once per month and only 21% more if they log in at least six times in the respective months. Hence, this suggests that paying more attention to daily spending, i.e., logging in at least once per month significantly reduces the tendency to overspend after payday. More specifically, the deviation from average daily discretionary consumption drops by around 60% when comparing the two groups. Interestingly, once users who frequently used (at least bi-weekly for three months in a row) the app in the past stop using the service, their propensity to consume more around payday increases significantly. Three days after payday, individuals in the active group spend 20% more than they do on average during the other days of the month. Those users who stopped using the app then spend on average 68% more on the same day. This suggests that while using a service such as MDB seems to help individuals to smooth their consumption, they need to keep using it to retain this effect. This suggests that users do not seem to learn from frequently checking their accounts and that they start spending more around payday, independent of their prior behaviour when they were actively using the app. These findings add large-scale evidence of consumption decisions to the existing literature on attention and financial decision-making (e.g., Hirshleifer & Theo, 2003; Stango & Zinman, 2014) by extending the scope to general consumption patterns. It also adds to our understanding on the influence of technology on financial literacy (e.g., Hillis, 2017; Levi, 2016) by indicating the limited learning effect of using a financial aggregator. Lastly, it contributes to existing studies on the payday effect (e.g., Gelman et al., 2014 & Olafsson & Pagel, 2018) by suggesting that the usage of financial technology might lead to underestimation of the payday effect in real-life settings.

ABSTRACT. This research studies an under-researched dimension of lifecycle decision-making: the choice of when and how to retire. Following reform of Estonia’s state pension scheme, we conducted a survey experiment on over 2,000 Estonian Facebook users approaching early retirement age (50-64). Respondents were asked to choose between early retirement, flexible pension before retirement age or to continue working until reaching retirement age. In an experimental manipulation, we randomly assigned respondents to either a control group or to one of four checklists. Our motivation for testing a checklist comes from research suggesting that even financially literate people find the task of making choices around retirement overwhelming. The prior literature has tended to address this concern by nudging choice towards options that have high expected value for the average person. This approach may nudge certain subpopulations towards an option that is, for them, suboptimal. Our checklists are instead a “signpost”. They are designed to prompt retrieval of relevant information and help the individual identify the option that best fits their desires. The first key result is that our checklists reduce the likelihood of choosing early retirement. On average the likelihood of choosing to continue working full time was 51% higher in the treatment conditions relative to the control condition. This shift in preference would have substantial economic implications e.g. for income tax revenue. A second key result is that respondents who were treated with the checklists expressed greater confidence in their choice, suggesting that our checklists are effective as signposts.

ABSTRACT. Many individuals around the world are not adequately prepared for their retirement (Benartzi & Thaler, 2013; McGowan & Lunn, 2019; Shin, Kim, & Heath, 2019). For example, the UK Pension and Lifetime Savings association (2018) reports that about half of the UK population will not reach an adequate retirement income. Similarly, in the US, the national retirement risk index (NRRI) shows that half of American households risk not being able to maintain their living standard during retirement (Munnell, Chen, & Siliciano, 2021).

One important determinant of pension adequacy is the choice of a pension contribution rate (MacDonald, Bianchi & Drew, 2012; OECD, 2018), that is, which percentage of income is set aside for retirement. Common presentation formats of contribution rates consist of numerical, abstract information. For example, pension contribution rates and the associated expected retirement income may be presented in a numerical table. The way that such information is presented affects the comprehension and use of it (Kelton, Pennington, & Tuttle, 2010; Johnson et al., 2012). Thus, the specific presentation formats of pension contribution rates may hinder individuals to reach sufficient saving levels, as many individuals do not comprehend what numerical information means for them (Lusardi & Mitchell, 2014; Peters, Västfjäll, Slovic, Mertz, Mazzocco, & Dickert, 2006). As a result, individuals may not feel confident enough to select a contribution rate, or may select a contribution rate that is too low to achieve the desired or adequate retirement outcome.

Our research aligns with recent developments in the pension industry. Retirement living standards expressed in labels have been developed in the UK (Padley & Shepherd, 2019) and Australia (AFSA, 2018). The living standards are labelled as ‘minimum’, ‘moderate’, and ‘comfortable’ (UK) or ‘modest’ and ‘comfortable (Australia). These living standards create insights into what lifestyle an individual can expect based on their retirement income. These labels are supposed to aid individuals in making informed choices about their pension plan contributions so that their choices matches their personal preferences. Whereas pension providers may want to present these living standards to their customers, it is unclear whether and how these evaluative labels and corresponding consumption baskets affect saving behaviour.

We propose that the interpretation of numbers such as the contribution rate may be eased by adding evaluative labels describing the expected pension lifestyle (such as “moderate” or “comfortable”) and consumption baskets (specification of the consumption possibilities per expected pension lifestyle) to retirement income projections. For example, evaluative labels and consumption baskets describing expected pension lifestyles may improve retirement savings if they motivate individuals to reach the next upward lifestyle level, make retirement outcomes more vivid, or when they make the choice of a pension contribution rate more evaluable. On the contrary, evaluative labels or consumption baskets may hinder individuals to choose a higher contribution rate than those associated with a specific label, or if they demotivate individuals for whom the next upward lifestyle label seems unachievable to increase their contribution rate. We contribute to the knowledge on evaluative labels by studying how adding labels to pension contribution rate information affects savings behavior, and which psychological processes underlie these effects. In addition, we study which individuals are more responsive to labels than others by assessing three possible moderators that have been established predictors of pension decision making (financial literacy, trust and patience) and whether the effectiveness of labels depends on one’s current pension contribution rate decisions.

We add novel insights on the effectiveness of evaluative labelling in intertemporal choice. This expands the current knowledge on evaluative labelling, which focuses on short-term decision making (e.g. choice of food). We develop an understanding of the process through which labels may affect behavior, by studying how labels affect the interpretability of numerical information and expected retirement incomes, and how they may increase savings through increased mental imagery of retirement outcomes.

We assess in three experimental studies (Experiment 1, N = 482; Experiment 2; N = 361; Experiment 3, N = 498) whether adding evaluative labels describing the expected pension lifestyle and consumption baskets to retirement income projections can induce individuals to save more. Study 1 shows that evaluative labels and consumption baskets increase pension contributions, regardless of one’s level of financial literacy or trust in pension institutions. Consumption baskets lead to increased pension contributions by increased mental imagery, and the effect of consumption baskets on pension contributions holds when the effect of additional savings on current income is made salient. Study 2 demonstrates that the effectiveness of evaluative labels and consumption baskets depends on the current contribution rate. Evaluative labels and consumption baskets do not always lead to higher contribution rates, but for specific contribution rates (especially those at the higher end of the distribution) they do. Evaluative labels and consumption baskets do not demotivate or discourage individuals, as our results indicate that they do not lead to lower contribution rates than a numerical table without an evaluative structure. In addition, we find that the effectiveness of evaluative labels and consumption baskets does not depend on one’s level of patience. Finally, Study 3 demonstrates that the positive effects of evaluative labels are also present when individuals make decisions for themselves, and not only when they are making hypothetical decisions for others. In addition, Study 3 shows that evaluative labels can trigger saving in a general saving context, that is not pension specific.

Our results indicate that evaluative labels and consumption baskets can increase contribution rates. In addition, we find that consumption baskets increase the ease of imagining the lifestyle one can expect in retirement. The effectiveness of interventions using evaluative labels and / or consumption baskets is not moderated by one’s level of financial literacy, trust or patience. However, the effectiveness of labels depends on one’s current contribution rate. We conclude that labels may act as an impactful anchor to increase savings, and that the implementation of pension labels in pension communication requires great care.

ABSTRACT. Extended abstract

Regularly saving for retirement can be challenging and has been shown to depend on various internal and external factors (e.g. Nyhus, 2017; Benartzi & Thaler, 2007). Long-term financial well-being, however, requires not only deciding when and how much to save but also when and how much to withdraw. The decision to withdraw savings can be suboptimal, as it can offset previous saving efforts made by individuals as well as by the state. It is, therefore, crucial to understand the determinants of the withdrawal decision and explore the possibilities of influencing this choice.

In this research, we capitalised on a recent political reform in Estonia that allows people to withdraw their pension fund savings before reaching retirement age, providing a valuable setting for studying withdrawal decisions. In Study 1 we investigated the effectiveness of two interventions, informed by the literature on loss aversion and delay discounting, to reduce the number of people withdrawing their existing savings. The online experiment was organized a few months before the reform and involved 1213 participants, recruited from a representative panel of the Estonian population. Participants were presented with a hypothetical situation involving a pension withdrawal decision and asked to make a binary decision to withdraw or to keep saving. The control group saw a neutral interface imitating a pension fund website with the total sum of savings and the income tax deduction presented. The salient loss aversion group was presented with an additional text box emphasizing the opportunity cost of the income tax paid upon withdrawal. The future view group was shown a comparison between the current sum withdrawn and the estimated higher sum at retirement age, if not withdrawing now.

Controlling for demographic and personality traits, we found that the salient loss aversion intervention significantly decreased withdrawal, 31% withdrawing in the loss aversion group and 39% withdrawing in the control group. The effect was evident in all demographic groups but slightly stronger in the 35-44 age group, the segment most likely to withdraw. The other intervention, showcasing future returns did not significantly affect withdrawal rates (37%) compared to the control group. The hypothetical situation asked the participants to imagine being under financial stress, so the withdrawal rates were above what was estimated in normal economic conditions.

Study 2 took place a few months after the reform and shed light on the differences between those who withdrew and those who kept saving in their now-voluntary pension fund. Interestingly, the main self-reported reason for continuing saving was the high cost of withdrawal, i.e income tax paid. This finding helps explain why making the loss of tax money salient could have nudged a significant portion of people to keep saving for retirement in Study 1. It is important to note that the tax paid on withdrawal is not an additional fine but the regular income tax left unpaid when the monthly salary had been transferred to the retirement fund.

Across both studies, we also asked which personality traits predict either hypothetical or actual pension withdrawal. In Study 1, we observed that continued retirement saving was associated with a higher level of agreeableness and internal locus of control. In Study 2, we found that continued retirement saving was associated with lower levels of risk-seeking in an investment vignette as well as higher trust in institutions.

The findings have policy implications and can inspire private sector financial institutions to look for win-win solutions in their user interface or communication strategies. Many countries are in the process of reforming their retirement policies and will hopefully make behaviourally informed decisions when possible. Withdrawal decisions become especially critical during turbulent times, e.g. recession and job loss (Argento et al, 2015).

References Argento, R., Bryant, V. L., & Sabelhaus, J. (2015). Early withdrawals from retirement accounts during the Great Recession. Contemporary Economic Policy, 33(1), 1-16. Benartzi, S., & Thaler, R. (2007). Heuristics and biases in retirement savings behavior. Journal of Economic perspectives, 21 (3), 81-104. Nyhus, E. K. (2017). 13 Saving Behaviour: Economic and Psychological Approaches. Economic psychology, 206.

ABSTRACT. Most of the literature on Holocaust survivors concentrates on the survivors’ post trauma and their way of life. Specifically, it focuses on how survivors cope with the trauma in their daily life and post-traumatic stress disorder (PTSD). Other research examines the way trauma is passed from the survivors to their offspring. The literature on the Holocaust survivors’ offspring (HSO) suggests that some of the trauma is indeed transmitted by parents and reflected in the behavior of HSO. In order the examine the influence of the Holocaust trauma on HSO, some studies compare between HSO and non-HSO groups. Others focus on the parents’ status at the time of the Holocaust or the parents’ level of post-trauma and its influence on their offspring. Much less research is concerned with the subjective perception of HSO regarding the influence of the Holocaust on their life. Rather than external perceptions of the influence of the Holocaust on HSO, these studies examine self-perception of the influence of the Holocaust on HSO. Another method for examining the influence of the Holocaust on the life of HSO is the Subjective Holocaust Influence Level (SHIL), which measures HSO’s perception regarding the influence of the Holocaust on their life. In this paper, we expand the discussion on the subjective influence of the Holocaust on the HSOs’ daily financial thinking and attitude, using the SHIL measure. Specifically, we are interested in the HSOs financial thinking and attitude because shortage, and specifically financial shortage, was a dominant issue in the Holocaust survivors’ life and in their trauma. The research on HSOs’ financial thinking and attitude is another way to examine the influence of the Holocaust survivors’ trauma on their offspring, especially because of the dominance of the financial shortage in their life. To examine HSOs’ daily financial thinking and attitude, we conducted a survey to measure financial satisfaction, materialism, income confidence, willingness to save money, generosity, and miserliness. We expanded the measures for materialism and saving money by including several questions based on well-known tools. We also asked questions regarding the attitude to the Holocaust and their family’s connection to the Holocaust. The survey was completed by 253 Jewish-Israeli HSO, who we divided them into three groups based on their SHIL. We find that SHIL affects materialism, income confidence and generosity, but is not related to financial satisfaction or motivation to save money. The results of this research are another step towards finding a subjective measure which can be used to examine how the Holocaust influences the daily life of HSOs; in this case, their daily financial thinking and attitude. Future research might use SHIL as the foundation to create subjective measures for other populations to study the influence that other traumatic events have on the descendants of victims. We suggest that future research use SHIL to measure other differences between HSOs and widen the discussion to more areas in daily life. Future research could also frame questions differently to test if some terms (such as “bad times”) trigger different perception and behavior than others.

ABSTRACT. Intuitive decision making has a large and often negative impact in economic decisions, but its measurement and quantification remains challenging. Following research from psychology, behavioral economists have often attempted to causally manipulate the balance of intuition and deliberation by relying on experimental manipulations as cognitive load. However, these attempts have resulted in mixed success, with many null results and no clear general pattern. We explain the possible reasons behind these developments and offer avenues for improvement. First, we show that a very simple formal model of decision processes offers a straightforward test to determine whether cognitive load has been successfully induced, hence disentangling failed inductions and true null results. Specifically, our model claims that cognitive load in complex decision tasks must result in shorter response times. Second, we show that the intuitive arguments on the behavioral implications of cognitive load do not hold on closer, formal examination, unless strong assumptions are made that may or may not hold in typical economic experiments. We then report on seven economic experiments (joint N=628) using different cognitive load manipulations and confirm the implications of the model. Our research elaborates on the differences between psychological experiments and economic tasks and the difficulties associated with importing methods and how one field can inspire another field.

ABSTRACT. The purpose of our research is to enhance the understanding of whether – and possibly how – the initial design of screen displays for information systems, such as dashboards and decision-support systems, may influence decision-making processes. In addition to the type of processes (analytic vs perceptual) and the decision results, we focus on how the initial presentation formats influence the decision-making strategies. Our research extends the research on presentation formats and decision-making performance by focusing on the relationships between the external presentation of data and decision-makers’ mental representation and processing of the task. According to Kelton, Pennington and Tutle (2010), these aspects are the least understood and under-researched areas related to research on the effects of presentation formats. To the best of our knowledge, this claim is still valid. In addition, we respond to a call by Bacic and Fadlalla (2016) to design studies that reflect how decisions are made in organisations. Most studies on the relationship between presentation formats and decision outcome have focused on rather simple tasks, often limited to data acquisition tasks. However, real-life managerial tasks are usually not simple, and data acquisition only pertains to the first phase of a decision-making process (Simon, 1977). In addition, most recent research has focused on the users’ possibilities to choose among graphic presentations (e.g., Luo, 2019; Kopp, Riekert & Utz, 2018). In contrast, our observations of real-life strategic and tactic managers indicate that they base their decisions on analyses of table data. In line with the argumentation above, we designed an experimental study to represent rather complex tasks, i.e., two versions of a task, one more complicated than the other, and to comprise a whole decision cycle. 42 MBA students were asked to make decisions related to managing a summer restaurant for 17 weeks. The goal was to maximise profit. The data presentation formats were graphs, tables and combined graphs and tables. The data have been collected using concurrent verbal protocols (Ericsson & Simon, 1980) The subjects made decisions using a spreadsheet model, and the results were saved. The subjects had the same background, except that 20 of them had an elective in business data processing. The data have been analysed using the principles of grounded theory (Glaser & Strauss, 1967; Strauss, 1987), that is, we have analysed the data in iterations, gradually enhancing our understanding of the relationships among the presentation formats, the information-processing strategies, and the results. The theoretic lenses were the following: We have extended the cognitive fit theory (Vessey, 1991) to complex tasks involving spatial and symbolic subtasks. The theory applied for assessing the subjects’ level of information processing was cognitive complexity theory (Schroder, Driver & Streufert, 1967;). Paivio’s dual-code theory (1971) motivated the design of our study by including the cooperation between the imagery and the verbal cognitive systems. The analysis revealed that the initial presentation format influenced the subjects’ mental representation of the decision problem and their information processing. This was particularly the case for subjects solving the complex version of our study. Subjects that were given the graphic presentation format, mainly used spatial processes, while subjects with the table format mainly used analytic processes. Subjects that used analytic processes attained higher profits (p < 0,001). In the second round of analyses, we looked particularly at the differences within each group to enhance our understanding of how the presentation formats were used during the decision-making process. First, we divided the decision process into two phases which we termed a problem-definition phase (PDP) and a problem-solving phase (PSP). This round gave the not-surprising result that subjects with a thorough PDP attained higher results than subjects who started solving the problem almost immediately. We detected, however, that few subjects changed the type of process they had chosen initially. Related to this finding, we observed that few subjects changed the initial presentation format. However, this partition into the two phases drew our attention to differences in the strategies that the subjects applied. We developed categories of strategies inspired by theories of general problem solving, such as trial and error, hill climbing and means-ends analysis (e.g., Kaufmann, 1988; Mayer, 1992; Anderson, 2015), but we adapted the categories to our context with focus on presentation formats. We found that subjects with perceptual and analytic processes applied quite different strategies, and that the strategies within each group also varied, leading to the development of sub-categories. We also found that the choice of sub-strategy influenced the results. Details on the strategies will be presented at the conference. Implications of our findings for the design of dashboards and decision support systems will be presented together with implications for managerial training. Limitations of the study will be discussed.

ABSTRACT. Imagine that you are screening candidates for an online service-desk operator position. You conduct an in-person interview to assess the candidates' support and communication skills. Since the interview is in-person, tall candidates tend to get higher scores, but height is irrelevant for success in the position. In the case of two candidates who got identical scores on the interview, one is tall and the other is short, whom would you choose? In normative decision models, irrelevant information should not influence the choice between competing alternatives. Nonetheless, irrelevant characteristics like the candidates' gender, race, and age affect decisions (Carlsson & Eriksson, 2019). An inadequate hiring process may increase the likelihood of high turnover, increase organizational misconduct, and lower morale (Bressler, 2014). According to one estimate, a poor hire adds up to about 30% to an employee's annual salary and replacing bad hire costs between $7,000 to a low-level employee and up to $40,000 for a senior manager (Kerr, Guide to Hiring Success). It is possible to reduce bias by blinding decision-makers to irrelevant information (e.g., Sah et al., 2015), but this is not always feasible, and decision-makers are usually reluctant to be blinded (e.g., Fath et al., 2021). Our study examines whether in such a case people are willing to have an algorithm help them choose between candidates. If one wants to control for the effect of an irrelevant attribute, she must adjust her assessment of the candidate accordingly. In the example above, the decision-maker needs to consider the influence of height when evaluating the candidates and reduce the evaluation of the tall candidate. Recent studies suggest that people have difficulty doing so (Rabinovitch et al., 2021). In formal models, such as linear regression models, an irrelevant characteristic can be included in the model and given a negative weight, i.e., as a suppressor variable (e.g., Darlington, 1968). Are people likely to accept advice from an algorithm that is based on a multiple regression model and follow it? It has previously been reported that people decline assistance from algorithms when choosing among candidates (e.g., Kaufman & Budescu, 2020), but appreciate it when making decisions in more scientific domains (Logg et al., 2019). Method. In four experiments (N = 857), we asked participants to select one of two candidates for a specific position according to a score achieved on a selection test (e.g., in-person interview). The test measures a relevant attribute (e.g., communication skills). However, it is also affected by another attribute that is irrelevant to the position (e.g., candidate’s height). One of the candidates is higher on this irrelevant attribute than the other. Thus, the lower candidate should be selected because she is higher on the relevant attribute the test measures. In Experiments 1 and 2, participants received advice regarding the normative choice from either an algorithm or a human expert (or no advice). They had to decide whether to follow the advice. In Experiments 3 and 4, they chose whether they wanted to see advice in the first place—either costly or for free—and were incentivized to choose optimally (by being given an opportunity to win a monetary bonus). The participants in all experiments provided their preferred method for selecting candidates (using an algorithm / relying on a human expert / deciding by common sense) after they made their selection. We hypothesized that: (a) A decision based on advice would be closer to the normative model than a decision made without advice; (b) Participants would regard advice from human experts more highly than algorithmic advice, both when following them and when choosing which type they want to receive. Results. We found that participants' decisions were more in line with the normative model when receiving advice (66-68%) than when not receiving advice (10-19%). Additionally, when faced with a single selection problem, participants followed either source of advice equally (Experiments 1 & 2) and were indifferent in regard to the type of advice they wanted to see (Experiments 3 & 4). Nevertheless, when asked for the best way to choose among candidates in general, participants showed a strong preference for human experts (51-62%) over algorithms (6-13%). This preference was present also for participants who opted to see the algorithm's advice and followed it. On average, about 50% of this group preferred to rely on a human expert, compared with 15% who preferred an algorithm. Third, 52-61% chose to see advice when it was free, but only 11-19% chose to see it when it was costly and lowered their potential payoff. Discussion. In many selection processes, human decision-makers make decisions without the assistance of a model. Height is just one of many potential irrelevant characteristics that could bias selection decisions. Most people struggle with correcting such irrelevant information intuitively. Such corrections can be made by algorithms (e.g., a regression model). We have found that providing advice from such a source drastically improves people's decisions. Despite this, people reject algorithms in general and prefer human experts. In other words, algorithms are effective but unappreciated. Participants follow algorithmic advice when it is free but are reluctant to use it when costly. The results indicate that people don't always recognize how complex the decision problem they face is. Understanding how to overcome the effects of irrelevant and potentially biasing information can improve selection processes. This improvement can help groups that are discriminated against by certain methods of selection, as well as organizations looking to hire the best candidates.

ABSTRACT. People sometimes avoid information about the impact of their action as an excuse to be selfish. Such “willful ignorance” reduces prosocial behaviors and has detrimental effects in many consumer and organizational contexts. We report the first meta-analysis on willful ignorance, testing the robustness of the phenomenon and quantifying its underlying motives. We analyze 33,603 decisions made by 6,531 participants in 56 different treatment effects all implying variations of an experimental paradigm assessing willful ignorance behaviorally. Meta-analytic results reveal the ability to avoid information about the impact of ones’ actions decreases prosocial behaviors by 28%, even if participants can easily find out this information. We estimate that about 40% of the observed ignorance is committed by reluctant altruists who use ignorance to excuse selfishness. We investigate the boundary conditions of willful ignorance and address the theoretical, methodological, and practical implications of our findings on who engages in willful ignorance, as well as when, and why.

ABSTRACT. As one of the biggest challenges humanity is facing today, climate change requires urgent action from all stakeholders. In addition to changes in regulation, provision of incentives and information campaigns, environmental policy increasingly includes green nudges to encourage individuals to reduce their carbon footprint (Schubert, 2017). This paper examines whether we can nudge individuals towards green behaviour by leveraging loss aversion. To this end, we run an economic experiment with a real effort task to test whether loss framing increases the likelihood of purchasing a voluntary carbon offset.

In line with prospect theory’s loss aversion hypothesis, the pain experienced from losses is relatively greater than the pleasure experienced from gains of equal value (Kahneman and Tversky, 1979); accordingly, loss-framed environmental appeals should be more effective in inducing behavioural change. Empirical evidence does indeed support the loss-aversion hypothesis with the majority of studies finding loss framed appeals to be more effective than gain-framed ones in driving green behaviour (Ropret Homar & Knežević Cvelbar, 2021). However, with only 7 of the (reviewed) studies measuring revealed preferences (the rest focusing instead on stated preferences), conclusions must be drawn with care and additional empirical research is warranted.

The motivation for measuring framing effects on the purchase of voluntary carbon offsets in particular is both practical and methodological. Representing 11% of all transport emissions (UNWTO, 2019), tourist air travel is an important contributor to climate change and the highest one-off emitter in an individual’s carbon footprint. Second to reducing the frequency of flying, individuals can purchase voluntary carbon offsets (VCOs) to reduce the net impact of their flight on the environment. In terms of methodological strength, the purchase of voluntary carbon offsets can be measured directly online, thus allowing for at the same time higher ecological validity (than typical laboratory experiments) and a more controlled environment than field experiments.

Following two rounds of pre-tests, we conducted an online experiment with 465 participants through the Prolific platform. The participants were divided evenly into a control group and two treatment groups: the loss treatment presented the negative consequences of not purchasing a VCO on the environment; the gain-framed text included positive consequences of purchasing a VCO on the environment, while the participants in the control group read a value-neutral description of voluntary carbon offsets. The subjects were then invited to an optional slider task (Gill & Prowse, 2012), consisting of 10 rounds to take approximately 5 minutes, with which they could earn up to £1 to purchase an offset. Their post-task decision on whether to collect this reward or allocate it towards a VCO constituted the real-behaviour component of the experiment. This was followed by survey questions capturing attitude towards voluntary carbon offsets, beliefs surrounding personal responsibility, self-efficacy and trust in VCOs, value orientation, environmental identity and behaviour, flying habits, familiarity with VCOs, and socio-demographics.

To minimise common method bias, the survey order was randomised, such that half of the participants followed the order specified above, while the other half first filled out questions on attitudes and beliefs, before being invited to complete the task and indicate how they would like to allocate their potential earnings (the behaviour component).

In total, 38.5% of participants chose to purchase a VCO within the experiment. The results show that once we account for an individual's beliefs related to voluntary carbon offsets, his value orientation, level of environmental concern and socio-demographic variables, the main effect of the frame on behaviour is not statistically significant. The interaction effect between the frame and survey order is, however, significant at the 5% level. Namely, a loss-framed description combined with behaviour-first survey order is associated with a higher likelihood of purchasing a VCO. Other statistically significant variables include frequent donation to environmental causes (negatively associated with behaviour) and a variable capturing environment-economic trade-off within Xiao and Dunlap’s (2007) Model for environmental concern, exhibiting an unsurprising strong positive effect.

The results imply that the loss frame had the theory-predicted effect, but only when it immediately preceded the behaviour. This suggests that the framing effect may have taken place as a ‘Type 1’ nudge, influencing a decision-making process that is automatic, subconscious and effortless (Wason and Evans, 1974). By contrast, by forcing the subject to reflect on their opinion of voluntary carbon offsets, questions on attitudes and beliefs transformed the decision on purchasing a VCO from a ‘Type 1’ to a ‘Type 2’ (that is, a deliberative and conscious) process, thereby diminishing the effect of loss framing. To policy makers and other choice architects, this finding may suggest that loss-framed environmental appeals will be more successful when placed alongside the desired behaviour (e.g. on recycling bins), rather than being part of, say, a wider information campaign.

Interestingly, more than half (54%) of respondents who expressed an intention to purchase a VCO within this experiment changed their minds once they were given the opportunity to do. One reason for this discrepancy may be that whilst the other studies examining real behaviour involve sacrifice in the form of effort (Ropret Homar & Knežević Cvelbar, 2021), here the cost is also monetary, suggesting that when decisions carry financial costs, people are less willing to act on their intentions.

Sources: Gill, D., & Prowse, V. (2012). A structural analysis of disappointment aversion in a real effort competition. American Economic Review, 102(1), 469-503. Homar, A. R., & Cvelbar, L. K. (2021). The effects of framing on environmental decisions: A systematic literature review. Ecological Economics, 183, 106950. Kahneman, D. & Tversky, A. 1979. Prospect Theory: An Analysis of Decision under Risk. Econometrica, Econometric Society, vol. 47(2), pp. 263-291. Schubert, C. (2017). Green nudges: Do they work? Are they ethical?. Ecological Economics, 132, 329-342. Wason, P. C., & Evans, J. S. B. (1974). Dual processes in reasoning?. Cognition, 3(2), 141-154. World Tourism Organization and International Transport Forum. (2019). Transport-related CO2 Emissions of the Tourism Sector–Modelling Results. Xiao, C., & Dunlap, R. E. (2007). Validating a comprehensive model of environmental concern cross‐nationally: A US‐Canadian comparison. Social science quarterly, 88(2), 471-493.

ABSTRACT. Introduction

Sustainable investments based on Environmental, Social and Governance (ESG) criteria are considered an important factor to mitigate climate change (Eurosif, 2018). To foster investments in activities that are sustainable according to the EU taxonomy, the European Union will make it mandatory for financial advisors to assess and consider investors’ sustainability preferences in the financial advice process. Informing investors about their ESG investment options is an important part of this new policy.

Previous studies show that providing information to clients affects investment decisions. While some experiments found that financial information regarding potential returns increases sustainable investments (Døskeland & Pedersen, 2016, 2021), other studies conclude that highlighting the ethical aspect of the sustainable investment increases ESG investments (Barreda-Tarrazona et al., 2011; Bassen et al., 2019; Bauer et al., 2021). Only few experiments (Døskeland & Pedersen, 2016, 2021; Glac, 2009) investigated whether financial or moral information is more effective, resulting in mixed outcomes. A reason for these mixed results might be related to the varying strength of manipulation. Whereas some studies work with different names of funds (Barreda-Tarrazona et al., 2011), others mark potential financial gains in a way that would most likely not be legal in EU countries (Døskeland & Pedersen, 2016, 2021).

Regarding ethical considerations, investors might be less influenced by moralizing investment behavior (what one should achieve) than by information on whether their investment behavior is actually effective and has a positive impact on the environment (what one could achieve). However, only few studies (Heeb et al., 2021; Siemroth & Hornuf, 2021), outline the environmental impact of investment decisions. Furthermore, no other study investigates whether financial-return and environmental-impact information interact with each other.

In this paper, we employ an incentivized online experiment to clarify the effect of financial-return information (stating that sustainable investments are profitable) and environmental-impact information (stating that sustainable investments have a positive impact on the environment) and the combination of both on sustainable investments. The paper further contributes to the literature by including individual characteristics (values, income, and financial literacy) as determinants of sustainable investments. These individual differences might also explain whether financial-return and environmental-impact information increases sustainable investments more effectively in specific subgroups.

We also investigate whether a general compared to a granular sustainability preference assessment affects sustainable investments. While the general assessment includes the percentage of investments that should consider ESG-criteria, the granular assessment additionally assesses the preferred focus: avoiding negative impacts and/or invest in activities that are sustainable according to the law. Moreover, our experiment investigates the effect of our treatment variations on customer satisfaction and the stability of the investment decision. Aside from the contribution to the academic literature, this study also provides insights on the impact of the new EU regulation. 

Method

We recruited a large and diverse sample of experienced investors and the general population (in total, N = 2254) in two survey waves. The sample was composed of 56.70% men with an average age of 48.20 years (SDage = 15.52). While 29.8% of the participants indicated more than 11 years of experience in investing, 34.1% reported no prior experience.

To test the pre-registered hypotheses, we conducted an investment experiment in which investment decisions of 15 randomly chosen participants were invested on the stock exchange and paid out after one year. All participants received an explanation of the term ESG and corresponding ESG factors (e.g., climate protection). As a treatment variation, we manipulated the type of information (none, financial-return, environmental-impact or both combined) and the sustainability preference assessment (general vs. granular). We informed participants, that the response in the preference assessment was used for a non-binding investment recommendation on the allocation of their money in the investment decision. Participants made the investment decision by allocating 600 Euro to four funds or by simply accepting the non-binding recommendation. The four funds (two conventional and two sustainable equity funds) were identical in terms of risk and performance but differed regarding ESG considerations and thus branches. A post-experimental questionnaire was included to survey customer satisfaction, individual differences such as altruistic and biospheric values (De Groot & Steg, 2007, 2008), financial literacy (Lusardi & Mitchell, 2008) and household income.

Results

The results show that financial-return as well as environmental-impact information increase sustainable investments compared to the control group. While environmental-impact information increases sustainable investments slightly more (not significantly, though) than financial-return information, presenting both pieces of information in combination has no additional positive effect. The results indicate that the effect of the environmental-impact information might be stronger when presented alone, compared to when presented in combination with financial-return information. Customer satisfaction remains mostly unchanged by financial-return or environmental-impact information.

The results also show heterogeneity effects in sustainable investments. Participants with high altruistic values or financial literacy react negative to financial-return information and increase sustainable investments less than those with low values or literacy. In general, higher biospheric values, financial literacy, and income are strong predictors of investing sustainably, while a higher degree of risk tolerance is related to lower sustainable investments. Moreover, women tend to invest more sustainably.

The granular sustainability preference assessment does not lead to less sustainable investments or satisfaction than the general assessment. In the granular assessment, about half of the participants focus on products that are sustainable according to law, e.g., the EU taxonomy, while the other participants either chose avoiding negative impact or both.

Policy implications for promoting sustainable investments are manyfold. We argue that information provided to clients should specifically address environmental-impact considerations. If investors have the impression that they can have environmental impact, they increase their sustainable investments. Financial-return information should be employed cautiously since information on financial returns was found to slightly counteract the environmental-impact information.

An interesting alley for future research could be to investigate the role of financial-return and environmental-impact information in stock market participation. Given the need for more sustainable investments and issues such as high inflation and old-age poverty, research should delve into the role of information as a measure to increase stock market participation and, at the same time, sustainable investments.

ABSTRACT. The mitigation of climate change is a crucial global challenge in the next decades. The overall reduction of greenhouse gas emissions requires swift global policy efforts, but will unlikely be successful without changes in private household consumption, which can be linked to almost two thirds of global greenhouse gas emissions. A substantial change towards more pro-environmental behavior (PEB) – including both the engagement in behaviors that benefit the environment and the omission of behaviors that harm the environment – is therefore essential to reach the required reduction in greenhouse gas emissions and mitigate climate change.

These behavior changes to reduce individuals’ environmental impact will have consequences on people’s daily lives and thus might affect individual well-being. In this paper, I use the concept of subjective well-being (SWB), which defines well-being from an individual’s subjective perspective on their own emotional and cognitive well-being. In the coming years, decision makers face the challenging task of balancing the well-being of citizens today with the well-being of future generations threatened by the ramifications of climate change. Consequently, the aim of environmental behavior change should be to achieve reductions of greenhouse gas emissions as large as possible – thus protecting the well-being of future generations – while causing as little as possible well-being losses today. To succeed at this goal, it is important to gain an understanding of the relationship between PEB and SWB.

Previous research generally finds positive correlations between PEB and SWB. This indicates that people who engage more in PEB seem to be happier and more satisfied with their lives than people who do less for the environment. However, recent debates in environmental psychology suggest, that PEB literature has been little concerned with the environmental impact of the behaviors measured until now – i.e., the actual impact a behavior has on the reduction of one’s greenhouse gas emissions or carbon footprint. Instead, the existing studies mostly focus on low-impact PEBs (e.g., stopping the water while brushing one’s teeth or bringing a bag to the grocery store) which may be a good indicator for measuring one’s general intention to behave pro-environmentally, but does not necessarily translate to lower ecological footprints. Only changes in high-impact environmental behavior will make significant contributions to mitigate climate change. Environmental impact is therefore crucial to take into account when exploring the relationship between PEB and SWB. This paper consequentially answers the following research question: Does the relationship between PEB and SWB differ depending on the environmental impact of the PEB studied?

To answer this research question, I systematically collected and analyzed studies that investigate the relationship between PEB of varying environmental impact with individual well-being. The strategy to identify relevant articles was fourfold: I searched for articles investigating behaviors termed as pro-environmental, behaviors in the context of climate change mitigation, behaviors aimed at reducing consumption in general and behaviors with particularly high impact on the environment in the high-impact consumption categories of food, mobility, and housing. After the article search, the studies were screened on a title-, abstract- and full-text-level based on pre-defined inclusion and exclusion criteria. The final sample includes 61 articles and 71 studies.

Based on studies assessing the environmental impact of individual PEBs, I systematically categorized the environmental impact of each PEB measured in order to test whether the relationship between PEB and SWB differs depending on the impact of the PEB on the reduction of greenhouse gas emissions.

First, I analyzed the overall relationship between PEB and SWB. The meta-analytical results reveal a very small positive association with an effect size of r = 0.081 (p < 0.0001). Second, I conducted an analysis of the moderation effect of environmental impact. The results reveal a negative moderation effect (B = -0.1619, p < 0.0001). I find that the relationship between low-impact PEB and well-being is positive and statistically significant, albeit small (r = 0.155, CI = [0.120, 0.190]), whereas the relationship between high-impact PEB and well-being is virtually zero (r = -0.005, CI = [-0.040, 0.028]).

The results for low-impact PEB go in line with previous research and indicate a small, positive and significant relationship between low-impact PEB and SWB. This means that, on average, people who engage more in low-impact PEB have higher well-being levels. To effectively mitigate climate change, however, an increase in high-impact PEB is needed. This paper provides the first systematic evaluation of the relationship between high-impact PEB and SWB and finds no evidence that they are connected to each other. These findings demonstrate that there is more ambiguity in the relationship between PEB and SWB than previously expected in the literature. While I find no evidence for a positive relationship between high-impact PEB and SWB, engagement in high-impact PEB does not seem to harm SWB either – contrary to what is suggested in climate change communication and public opinions.

This paper shows that the environmental impact of PEB matters when evaluating its relationship with SWB. Further research needs to ascertain how low-impact and high-impact PEB differ from each other and why this difference is relevant for individual well-being. Furthermore, further research is needed specifically on the relationship between high-impact PEB and SWB to increase understanding and design policy interventions that effectively decrease greenhouse gas emissions without sacrificing individual well-being.

ABSTRACT. •Motivation

Our planet is currently facing serious environmental challenges. One of the most challenging issues is plastic pollution, mostly caused by overconsumption and improper disposal of single-use plastic products (United Nations Environment Programme [UNEP], 2021).

Although institutions and researchers highlight the importance to take action (UNEP, 2021; Elliot & Webster, 2017), and even if people are aware of the problems, bridging the knowledge-to-action gap and motivating environmental behavior still remains a major challenge (Kollmuss and Agyeman, 2002).

•Topic and already existing literature

Non-governmental organisations play a key role in addressing such problems, but their activities mostly depend on voluntary donations. However, communicating about environmental issues - such as pollution, ocean acidification, or wildfires - is challenging because people rarely experience them directly and may therefore perceive them as psychologically distant (Trope & Liberman, 2010), remote and irrelevant risks (Carmi & Kimhi, 2015; Spence et al., 2012).

Although personal experiences of environmental threats are likely to bring these issues psychologically closer and increase perceptions of their risks (Akerlof et al., 2013; Van der Linden et al., 2015), such interventions would be complicated, lengthy, costly, and even dangerous.

One solution to overcome this gap could be communication through immersive media formats, such as Augmented Reality (AR) and Virtual Reality (VR). By providing sensory-rich experiences, this type of technology is able to create a sense of presence (Mol, 2019; Innocenti, 2017) and make users feel as if they are directly experiencing environmental issues.

• Research question & methods

However, to the best of our knowledge, there is no study examining the effects of AR on manipulating the psychological distance of environmental threats and motivating green behavior.

To fill this gap, we conducted a laboratory experiment to investigate whether AR is able to reduce psychological distance and motivate pro-environmental behavior.

• Hypotheses

We exposed participants to a 2-minutes AR experience, during which they observed what happens to some animals if they ingest or get entangled to plastic litter in the sea. We hypothesized that, in contrast to a control group, such experience would raise participants’ environmental attitudes and willingness to donate money to an environmental organization of their choice.

• Results By looking at aggregate data, AR does not seem to be more effective than simple persuasive messages both in terms of amount donated and psychological distance. Yet, AR does have a heterogeneous effect based on the sense of presence and environmental attitudes of participants in the experiment. Results suggest, therefore, that the sense of presence, generated by AR experience, is likely to impact users’ environmental attitudes, and raise concern about environmental issues.

Results suggest that the sense of presence, generated by AR experience, is likely to impact users’ environmental attitudes, and raise concern about environmental issues.

References:

Akerlof, K., Maibach, E. W., Fitzgerald, D., Cedeno, A. Y., & Neuman, A. (2013). Do people “personally experience” global warming, and if so how, and does it matter? Global Environmental Change, 23(1), 81–91. doi:10.1016/j.gloenvcha.2012.07.006 Carmi, N., & Kimhi, S. (2015). Further Than the Eye Can See: Psychological Distance and Perception of Environmental Threats. Human and Ecological Risk Assessment: An International Journal, 21(8), 2239–2257. doi:10.1080/10807039.2015.1046 Elliot, S., & Webster, J. (2017), Editorial: Special issue on empirical research on information systems addressing the challenges of environmental sustainability: an imperative for urgent action. Info Systems J, 27: 367– 378.

ABSTRACT. People sometimes evaluate an inferior option as better than a superior alternative when the options are presented separately, compared to when the options are presented jointly. Hsee (1998) explained this "less is better" phenomenon using the evaluability theory: Separate evaluations of options are often more influenced by attributes which are easier to evaluate, and less by more important attributes that are harder to evaluate. In this paper, we show an opposite pattern when making efficiency upgrade decisions: When considering whether to upgrade a product or a service that is measured by efficiency or speed, people overestimate the savings of large upgrade relative to a smaller one when both upgrades are presented jointly, but not in separate evaluation mode. Efficiency (or productivity) measures such as driving speed (e.g., mph/kmh), Internet speed (e.g, megabits per second – Mbps), and fuel economy (e.g., miles per gallon – MPG, or kilometers per liters), are very common, but people’s ability to evaluate changes in these measures are sub-optimal (Larrick & Soll, 2008; Svenson, 2008). Presented with information about the efficiency of products or services, people have been found to overestimate savings when performed from relatively high initial values. For example, people falsely believe that upgrading a car from 20 to 30 MPG would save more fuel than upgrading from 10 to 13 MPG, when the opposite is correct. This bias occurs when people follow non-normative rules that either favor the upgrade with the larger absolute difference (Difference heuristic) or the larger proportional increase (Proportion heuristic). Although both heuristics lead to biased estimations, the use of the Difference heuristic typically results in more biased evaluations relative to the Proportion heuristic (De Langhe & Puntoni, 2016). Because the Difference heuristic is simpler to follow, we hypothesized it would be more prevalent when evaluating two upgrades in the more cognitively taxing joint evaluation mode than in separate evaluation mode. Thus, we hypothesized that people would more frequently use the simpler Difference heuristic in joint evaluation of two upgrade options, and subsequently provide willingness to overpay for them. In contrast, in separate evaluation mode, people would more frequently use the Proportion heuristic, resulting in more calibrated willingness-to-pay (WTP). We examined these hypotheses in two experiments in which participants rated smaller vs. larger upgrade offers either in separate, or in joint evaluation mode. In Experiment 1 (N=298) participants expressed their WTP for hypothetically upgrading their Internet service’s or their printer’s speed, upgrade their car’s fuel efficiency, or increase their driving speed (by using a fast lane). We found that the WTP for larger vs. smaller upgrades were more biased, compared to the actual time/fuel savings, in the joint (vs. the separate) evaluation mode. For example, when evaluated separately, participants were willing to pay $17.33 to upgrade their Internet speed from 25 to 50 Mbps, and $21.97 (27% more) to upgrade from 25 to 100 Mbps. Although the actual time saved in the larger upgrade is 50% higher than the smaller upgrade, when presented jointly, participants were willing to pay 100% more for the larger upgrade ($32.26 vs. $16.11). Similar patterns were found for WTP for using a fast lane to increase driving speed (92% in the joint evaluation mode vs. 51% in the separate mode; the actual time saved in the larger upgrade is 50% higher than the smaller upgrade); upgrading to a faster printer (70% vs. 10%; the ratio of actual time saved is 33%); or upgrading to a more fuel efficient car (76% vs. 35%; the ratio of actual fuel saved is 33%). All of these interaction effects were statistically significant (p < .01). Experiment 2 (N=614) replicated these findings using a forced-choice (accept vs. reject the upgrade), using two scenarios from Experiment 1: The Internet speed upgrade priced at $17 for the smaller upgrade (25 Mbps to 50), and $30 for the larger upgrade (25 Mbps to 100). In the separate mode condition, 48% chose to upgrade the speed in the smaller upgrade condition, whereas 52% chose to do so in the larger upgrade condition. In contrast, in the joint condition most participants chose the larger upgrade (68%). A similar pattern was found for the fuel-efficiency scenario where the two upgrades were priced as $80 and $120: 50% and 49% chose to upgrade the smaller and larger fuel-efficiency (respectively), whereas 56% chose the larger upgrade in the joint condition. A Z-test for proportions showed that the difference in accepting the large upgrade between the joint and separate conditions was statistically significant in the Internet scenario, Z = 2.97, p < .01, but not in the fuel-efficiency scenario, Z = 1.26, p = 0.1. Our findings show that joint evaluation cause people to be willing to overpay for larger efficiency upgrades of products or services, whereas less bias accompanies separate evaluations. These findings are consistent with previous studies suggesting that people use heuristics rather than normative strategies when trying to estimate time or fuel savings (Larrick & Soll, 2008; Svenson, 2008). However, this research is novel in suggesting that people also revert to simpler heuristics over more complicated ones when the task at hand gets more cognitively demanding (e.g., the joint evaluation mode). Previous demonstrations of preference reversals in joint vs. separate evaluations have been explained by the evaluability of different attributes (Hsee, 1998). We suggest that the effect of separate vs. joint evaluation mode in the context of efficiency upgrades could be better explained by the different availability of cognitive strategies (heuristics) used for evaluating efficiency upgrades. Specifically, the availability of more biased heuristics (e.g., the Difference heuristic) increases when upgrade options are presented jointly, whereas the availability of less biased heuristics (e.g., the Proportion heuristic) increases when upgrade options are presented separately. Consumers should be aware of their biased WTP for efficiency upgrades in joint evaluation mode in order to lessen their susceptibility for this bias. Future research could examine other contexts in which more cognitively demanding tasks could increase the use of simpler heuristics, causing more biased evaluations and sub-optimal decisions.

ABSTRACT. A Middle Eastern marketplace bears many hallmarks of a perfectly competitive market. Row upon row of small-scale, often side-by-side, vendors sell identical tomatoes, cucumbers, eggplant, oranges and spices with their prices clearly displayed. These features make comparison-shopping easy for buyers, thus constraining both the prices set by myriad vendors and buyers’ scope for negotiation. We conduct a field experiment on these seemingly highly competitive Israeli produce markets to investigate the possibility of price discrimination based on buyers’ gender, perceived attractiveness and other perceived traits. We trained 90 buyers and sent them to produce markets across Israel. At each vendor, they verify the posted price and then ask for a discount on a one-kilogram or one-unit purchase. Vendors – predominantly males – employ third-degree price discrimination: women are offered larger and more frequent discounts than men, and the more attractive the female buyer, the larger and more frequent the discount offered. Male buyers do not benefit from this beauty discount. We focus on the gender and attractiveness of buyers as potential sources of price discrimination because they are two of the most salient and easily judged traits upon seeing someone briefly for the first time, as in a buyer-seller encounter in the produce markets we study. At the same time, we evaluate other traits of the buyers to allow for the possibility that they account for the observed price discrimination. Yet, no other perceived buyer characteristic influences the likelihood of obtaining a price discount: neither the buyer’s perceived wealth, kindness, intelligence nor ethnicity are significant predictors of the tendency to offer a discount or its magnitude. Although comparison-shopping of posted prices is straightforward, we argue that it is not costless, particularly when it comes to ascertaining final prices. The near absence of price-information sharing among buyers in these markets permits vendors to price at a markup and to grant individual requests for a discount without fear of inviting an onslaught of additional takers. The finding that the mostly male vendors in our sample willingly lower their posted prices in response to the buyer’s gender and, to a lesser extent, their looks attests to this markup and their willingness to employ it at their discretion. This paper contributes to the existing literature in several ways. First, based on evidence from 90 different testers, we demonstrate that discrimination can appear in a natural, competitive market setting. We thus contribute to the debate on the generalizability of laboratory experiments. Moreover, a large portion of beauty premium studies are based on survey data whereas our results, based on a field experiment, are complementary and can possibly be viewed as more reliable. Second, only a few audit studies provide evidence of price discrimination against customers. We provide further evidence, in a different setting, that is helpful to generalize the findings. Third, to the best of our knowledge, no audit study finds discrimination based on the customer’s attractiveness. While evidence of beauty-based discrimination exists, none in a consumer context. Finally, our research setting, a produce market, largely meets the assumptions of perfect competition, and therefore the likelihood of finding evidence of discrimination is ostensibly lower.

ABSTRACT. Introduction A business may have a strategy for sustainability in that it considers its social impacts and environmental impacts and prioritizes social-environmental consideration over financial considerations (Liu et al., 2021; Tiberius et al., 2021; Kazancoglu et al., 2021). A business may have a practice for sustainability in that its operations enhance social benefits and limit environmental harm (Liu et al., 2021). Strategy and practice in a business may be coupled in that strategy is implemented in practice, in the operations in the business (Liu et al., 2021). Businesses differ in that some elaborate strategies and practices for sustainability, while others care less. Recent research has found that family businesses are less environmentally oriented than non-family businesses (Miroshnychenko & De Massis, 2022). This frames our research question. In family businesses contrasted non-family businesses, what are the strategies and practices for sustainability, and what is the coupling between strategy and practice? This question is answered by analyzing a large sample of businesses in Spain, surveyed in 2021. The findings contribute to shedding light on differences and the importance of strategies and practices for sustainability, as derived in family businesses. Theoretical background and hypotheses Family owners can actually be agents of sustainable development (Ernst et al., 2022). The economic, environmental, and social sustainability dimensions play an important role in family businesses comparing non-family businesses (Fritz et al., 2021). Also, the attitude of family-owned businesses affects environmental sustainability intention (Singh, et al., 2021). It can be driven by socioemotional wealth (Gomez-Mejia et al., 2007, Kalm & Gomez-Mejia, 2016) referring to non-financial aspects of the firm and can be compared to the personal integrity of the owner-manager (Randerson, 2022). We specify three working hypotheses, H1. The strategy differs between family businesses and non-family businesses, in that family businesses draw up more sustainability strategies than non-family businesses. H2. Practice differs between family businesses and non-family businesses, in that family businesses are into more sustainable practice compared to non-family businesses. H3. Strategy and practice in a business are coupled positively, but only loosely. Research design A survey of representative samples of 1958 family businesses and 586 non-family businesses in Spain, conducted in 2021 by the Global Entrepreneurship Monitor. A family business denotes a business mostly owned and mostly managed by a family. A non-family business is a business that is not mostly owned or not mostly managed by a family. Solo-businesses are ignored. Strategy is measured by reported social and environmental considerations and prioritizing of social-environmental considerations. Practice is indicated by enhancing the social benefit and reducing environmental harm. The GEM survey enables us to control for several characteristics related to cause and effect. Results H1 posits that strategy differs between family businesses and non-family businesses, in that family businesses draw up more sustainability strategies than non-family businesses. This hypothesis is tested by the coefficient for governance. The significant and positive coefficient indicates that family businesses draw up more sustainability strategies compared to non-family businesses, controlling for other conditions. This supports Hypothesis 1. H2 posits that practice differs between family businesses and non-family businesses, in that family businesses are into more sustainability practice than non-family businesses. The positive and significant coefficient indicates that family businesses are into more sustainability practice compared with non-family businesses, which supports Hypothesis 2, controlling for other conditions. H3 posits that strategy and practice are loosely coupled, the coefficient for strategy is significant, supporting Hypothesis 3. The positive coefficient indicates that strategy and practice are coupled positively.

Table 1. Strategy and practice affected by family governance Strategy Practice Strategy 0.113*** Governance:Family Business 0.099† 0.054* Controls … … Intercept 3.410 -0.102 R-square 0.091*** 0.148*** N businesses 2,544 Linear regression, with metric coefficients. †p<.10 *p<.05 **p<.01 ***p<.001

Discussion The funding shows that family businesses elaborate more sustainability strategy and implement more practice than non-family businesses. However, in other research, family businesses take part less in green practices than non-family businesses (Miroshnychenko et al., 2022). Strategy and practice are coupled somewhat tightly both in family businesses and in non-family businesses, which is in line with Liu and colleagues’ findings in 2021. A noticeable difference between family businesses and non-family businesses about sustainability strategies and practices is recognized. It is a contribution to the current research. The results contribute to recognizing the importance of sustainability strategies and practices, as embedded in family businesses.

References

Ernst,R.A., Gerken,M., Hack,A., & Hülsbeck, M.(2022). Family firms as agents of sustainable development: A normative perspective. Technological forecasting and social change, 174, 121135. Fritz,M.M.C., Ruel, S., Kallmuenzer, A., & Harms, R. (2021). Sustainability management in supply chains: the role of familiness. Technological Forecasting and Social Change, 173, 121078. Gómez-Mejía,L.R., Haynes,K.T., Núñez-Nickel, M., Jacobson,K.J., & Moyano-Fuentes, J. (2007). Socioemotional wealth and business risks in family-controlled firms: Evidence from Spanish olive oil mills. Administrative science quarterly, 52(1), 106-137. Kalm,M., & Gomez-Mejia,L.R. (2016). Socioemotional wealth preservation in family firms. Revista de Administração, 51(4), 409-411. Kazancoglu,Y., Sezer,M.D., Ozkan-Ozen,Y.D., Mangla, S.K., & Kumar, A. (2021). Industry 4.0 impacts on responsible environmental and societal management in the family business. Technological Forecasting and Social Change, 173, 121108. Liu,Y.,M. Samsami,H. Meshreki,F. Pereira, and T. Schøtt. 2021. Sustainable Development Goals in Strategy and Practice: Businesses in Colombia and Egypt. Sustainability, 13(22), 12453. Miroshnychenko,I., & De Massis,A. (2022). Sustainability practices of family and nonfamily firms: A worldwide study. Technological Forecasting and Social Change, 174, 121079. Randerson,K. (2022). Conceptualizing family business social responsibility. Technological Forecasting and Social Change, 174, 121225. Singh,G., Sharma,S., Sharma,R., & Dwivedi, Y.K. (2021). Investigating environmental sustainability in small family-owned businesses: Integration of religiosity, ethical judgment, and theory of planned behavior. Technological Forecasting and Social Change, 173, 121094. Tiberius,V., Stiller,L., & Dabić,M. (2021). Sustainability beyond economic prosperity: Social microfoundations of dynamic capabilities in family businesses. Technological Forecasting and Social Change, 173, 12

ABSTRACT. Background and theoretical framework: Our field of research is that of self-employment with a French umbrella company (Kantorowicz, 2015; Loufrani, 2015). Self-employment with umbrella company is a specific French status (portage salarial) allowing a self-employed worker, or freelancer, to carry out his activity in the form of an employee under private law (Louvion, 2019). It is a hybrid status where he is an employee of the umbrella company that provides administrative and legal support for his activity. Ottmann, Mahut and Gazier (in press) addressed the issue of the place of umbrella companies in freelancers’ career. They identified, by quantitative methods applied to the databases of an umbrella company, characteristics of the freelancers’ activity allowing different profiles to be drawn. Among these profiles, those of "established" and "quasi-established" have emerged. An established individual is defined as a self-employed person who has succeeded in establishing a stable and remunerative activity. The "quasi-established" individuals have statistically the same characteristics but with lower revenue, a less regular activity and a faster exit from the umbrella company. This study is specifically about the established and quasi-established freelancers. Our research questions are: (1) what is the place of collaboration with umbrella companies in the careers of established freelancers and, (2) what entrepreneurial work allows them to establish a remunerative and stable freelance activity? Our assumptions are those formulated by Ottmann, Mahut and Gazier (in press) concerning the category of established freelancers at the end of their article: - On average, they have more stable family situations than the other categories, i.e. probably spouses who are themselves financial supporters of households. - These are successful transitions from salaried employment to freelance work, particularly in end-of-career situations. The transition could have been initially desired but also involuntary, especially in situations where the contract was B2B (Business to Business). - A "drop" in the quantity of activity after three years was statistically highlighted. This could be explained by exits from freelancing and/or self-employment or by a decrease in the time worked: - First possibility: the "established" freelancers capitalize on their previous expertise and professional network and both would " expire" after three years. - Second possibility: people reach retirement age and continue their activity but with a gradual decrease.

Methodology: Semi-structured interviews of about one hour with: - Ten established freelancers, including two women (42 and 55 years old at the time of entry into self-employment with umbrella company) and eight men (33 to 63 years old). - Five quasi-established freelancers including three women (53 to 57 years old) and two men (39 and 46 years old). Moreover, 32 additional interviews were conducted with other profiles. Their data shed light by comparison on certain dynamics of the established and quasi-established profiles.

Results: The hypotheses deduced from the statistical analysis of the databases are globally confirmed by the qualitative interviews. These are individuals in the second half of their career, some of whom have chosen to be self-employed, while for others self-employment was a default choice after losing their job and failing to find another stable salaried job. Among those for whom self-employment was a choice, it could be in a long-term or even permanent perspective, or temporary, with some describing self-employment as a break in their career. Nevertheless, despite the diversity of motivations, it was indeed an entrepreneurial work on their part that allowed them to succeed in their transition to self-employment. They capitalized on their networks and skills acquired in their previous salaried jobs. Subsequently, the hypotheses concerning the drop and the exit from self-employment are confirmed. There is indeed depletion of the professional network after a few years and difficulties in renewing this network, which would lead to a gradual retirement. Entrepreneurial work then seems inaccessible, which raises the question of the entrepreneurial activity experienced and perceived, in the context of freelance collaboration with umbrella company, by integrating the automatisms that could have been built up in the first part of the career in salaried employment and that re-emerge in freelancing with umbrella companies due to the hybrid nature of the status. The return of these automatisms forged in salaried employment can be an obstacle to the perpetuation of an entrepreneurial activity.

References Kantorowicz, B. (2015). Droit et pratique du portage salarial. LexisNexis. Paris, France. Loufrani, Y. (2015, juin). Le portage salarial enfin sécurisé. Revue Française de Comptabilité, (488), 7‑8. Louvion, A. (2019). Profession : tiers employeur. Quand les entreprises de portage salarial vendent du salaire. Sociologie du travail, Vol. 61 - n° 4. Ottmann, J-Y., Mahut, D., et Gazier, B. (in press). Entre emploi atypique et entrepreneuriat : le portage salarial comme « forme d’emploi de transition » ?

ABSTRACT. During the past decades, there has been a massive shift in the childcare responsibility from the family to the local community. In Norway, the publically provided childcare coverage for 1-5 year olds has increased from 20 to 90 percent since 1980. Based on the existing literature, we would expect this development to reduce the influence of family background on school performance. We show in this paper that the opposite has happened. The trend towards universal childcare has been accompanied by an increase in the correlation between parents’ earnings rank and their offspring’s rank in the primary school performance (GPA) distribution. In the present paper, we examine how the social gradients in primary school performance have developed in different parts of the country, and exploit this spatial variation to identify and estimate causal relationships. In particular, we seek to identify the roles of factors such as local childcare coverage, income inequality, and the degree of residential segregation in explaining the influence of family background on primary school performance.

Equality of opportunity is a widely accepted aim of economic and social policy. From an intergenerational perspective, equal opportunities imply that offspring born into poor families have the same chances in life as those born into richer families. The empirical literature on intergenerational earnings correlations points to Norway and the other Nordic countries as being among the most socially mobile societies in the world. However, although the literature on mobility trends in these countries shows mixed results, recent empirical evidence from Norway suggests that intergenerational mobility has come under pressure, particularly at the bottom of the socioeconomic class distribution. As intergenerational earnings mobility metrics typically require earnings data for both parents and offspring at mature age, there is so far no evidence covering offspring born after the early-mid 1980s. As a result, existing studies have not been able to capture any shift in mobility trends arising from the expansion of publicly provided childcare that has taken place from around 1980.

In a previous study (Markussen and Røed, 2020), we have examined intergenerational mobility trends in Norway for offspring cohorts born between 1952 and 1975. A major finding was that overall economic mobility declined over this period, and that the decline was particularly large at the bottom of the parental earnings distribution. A problem with earnings-based analyses of intergenerational mobility, is that individual earnings-potentials are reveled at a relatively late stage in the lifecycle; hence, based on today’s data it is difficult to examine intergenerational mobility for cohorts born after the early 1980s. Hence, if we want to study trends in the influence of family background on offspring’s adult economic outcomes for more recent birth cohorts, we need to study outcomes that are likely to be highly correlated with adult economic outcomes, yet measured and recorded at a much earlier age. One such outcome is the grade point average (GPA) obtained from lower secondary school, typically at age 15 or 16. This outcome has the great advantage that it can be mapped into a rank outcome which has the exact same distribution as the earnings rank outcomes. Hence, it is possible to study trends in intergenerational mobility for cohorts born up to the beginning of the 2000s on a scale similar to that used for earnings outcomes.

Preliminary results indicate that the trend toward lower intergenerational mobility has continued into the new century. The correlation between parents’ earnings rank and offspring GPA-rank has increased almost year-by-year for cohorts born between 1987 and 2001. Although we cannot be sure that trends identified from GPA scores will be mirrored in earnings-based measure when these offspring grow up, these patterns suggests that the expansion of publically provided childcare has not been sufficient to offset other mechanisms inducing a stronger relationship between family background and early educational performance.

In the remainder of the paper, we examine the relationships between GPA rank and parental earnings rank at local levels, such that we can exploit the spatial variation in potential explanatory factors. This will be done within the framework of regression models.

ABSTRACT. Introduction Child poverty is rising in Norway. The proportion of children who live in persistently low income households has increased from 7.7% (2008-2010) to 10.7% (2015-2017) 1. Growing up poor is associated with poorer physical and mental health, more developmental delay, and lower school achievement2, 3. These adverse associations of low income have also been documented in studies of Norwegian youth 4-6. Poverty may influence health and development through factors operating at several levels (i.e. individual, relational and institutional) 7, but services to children and families with low income are often poorly coordinated8. Building on the existing literature and attempting to provide better services to low-income families, the New Patterns project has been developed. New Patterns recruits families with low income and need for long-term welfare services. Included families receive integrated welfare services from a permanent family coordinator (FC) who coordinates services from different sectors (culture, education, welfare, health- and social services) and voluntary organizations, and supports all family members for five years (see figure 1). Families are recruited from different municipality services: kindergarten, school, public health clinics, general practitioners, the Norwegian Labour and Welfare Administration (NAV), child protection services, and mental health services. FCs tailor services to needs identified by the end-users through developing a family plan where the question “what is important for you?” is key, and by keeping the end users’ perspectives as a foundation when developing measures and aims for achieving a better situation for the families. New Patterns aims to prevent transfer of poverty and the accompanying social, economic, and health-related consequences over generations. Aims The aim of the current study is to evaluate how integrated and coordinated services provided over an extended period by a family coordinator affect children and their parents. The effectiveness of New Patterns will be assessed using several indicators of socioeconomic status and living conditions, service use, mental health and health related quality of life, self-efficacy, school performance and leisure time activity. Methods The study design was informed by the Medical Research Council’s guidance on the development and evaluation of complex interventions 9. Using a mixed methods approach utilizing survey and register data, as well as information gathered through interviews and shadowing we will describe and evaluate outcomes associated to the intervention, and the process of implementation. Quantitative data is gathered at baseline and annually throughout the study. We develop a panel dataset based on repeated measures of the same outcome variables over time for children and parents, i.e. income, education, living condition, leisure activity, work, health and well-being. This allows us to estimate effects of the intervention and provide in-depth knowledge of the families’ living conditions. Participants are identified to facilitate longer-term follow-up using register data, this will make identification of a potential control groups possible, i.e. matching on observable characteristics 10, 11. Participants New Patterns recruits families with children aged 0-17 years, with household income in a period of at least 3 years lower than 60% of the equivalized median income in the population. In addition, family members are in need for long-standing welfare services. Families are not eligible for the project if the child protection service is considering taking over the daily care of the child/children at the time of recruitment. However, involvement of child protection service as a support in the family is not an exclusion criterium. All families recruited for New Patterns are asked to participate in the research project- Based on the plans for enrollment, at least 200 families will be participating in the New patterns. Family characteristics for those already enrolled suggest an estimated range of 400-600 children and 200-300 adults in the project at large. Results We have developed a detailed mapping that has been implemented in New Patterns and will present baseline data describing the socioeconomic conditions and health-related measures of families included. Conclusion We have limited knowledge about life circumstances for poor families with children growing up in a high-income country like Norway. Hence, this study will provide new knowledge about a population segment which is underrepresented in surveys. In addition, the study will provide in-depth knowledge about how children and adults in low-income families experience their life circumstances, health and well-being.

ABSTRACT. In this study we examine individual financial self-efficacy accuracy, to further understand self-efficacy and its connection to the gender gap in financial literacy. Self-efficacy accuracy refers to how well self-efficacy aligns with actual performance (Stankov and Lee 2017, Talsma, Schüz et al. 2019), on an individual level. A large number of quantitative studies have found associations between higher self-efficacy and performance in areas such as work-related tasks, entrepreneurial activity, creativity, physical activity, and academic achievements, including financial literacy. Further, there are clear gender differences in level of self-efficacy across areas of knowledge and skills (Rimm and Jerusalem 1999, Stoet and Geary 2018). While males tend to overestimate own abilities, females tend to underestimate their abilities. We aim at conducting individual-level investigations of financial self-efficacy among adolescents, with an emphasis on gender differences. Investigating self-efficacy on an individual level is important in order to obtain details about the mechanisms in self-efficacy and performance. Individuals with similar skills, or the same person on different occasions, may perform very different depending on their level of self-efficacy (Bandura 1982). Further, in the current study we examine self-efficacy towards not-yet-experiences tasks. Novel situations may affect self-efficacy differently than familiar tasks, given previous findings regarding the relationship between general and domain-specific self-efficacy (Miyoshi 2012). Novel situations are also expected to produce increased gender differences. Also, economics as a novel subject, implies that we are testing the more academic aspect of financial self-efficacy, which may unfold and manifest itself differently than financial self-efficacy regarding more practical financial tasks. The reason for targeting adolescents is to explore whether, and potentially how, individual differences in financial self-efficacy appear prior to adulthood. It is a preliminary study in which a recently developed general self-efficacy measure (Nielsen, Dammeyer et al. 2018) is tested, in combination with validated measures of financial self-efficacy and financial literacy. We invited four 9th grade classes (approximately 80 students) to participate in a 45-minute paper-and-pencil questionnaire and test, which consists of 4 blocks. Block 1 is a general self-efficacy test + one domain-specific financial self-efficacy item. Based on the argument that generalized self-efficacy is the aggregate of specific past experienced successes and failures (Sherer, Maddux et al. 1982) and assumed to be a predictor of performance in new situations, we presented this measure prior to the financial literacy test. Block 2 is an objective financial literacy test (OFK) drawn from the PISA assessment 2018. Block 3 is one subjective financial knowledge (SFK) question. SFK is also an indication of self-efficacy when compared to OFK, and previous studies (Usher and Pajares 2008, Fong and Krause 2014) report that own mastery experiences are particularly important contributors to academic self-efficacy. We therefore placed SKF after the OFK to enable comparison of self-efficacy before and after the OFK test. Block 4 is a locus of control question. We aim at testing self-efficacy accuracy by comparing the results from performance on the literacy test with the self-efficacy results, on an individual level. Individual variations in self-efficacy measured before and after performance test will also be explored. Further analysis will examine if there are gender differences in financial self-efficacy accuracy in such a novel setting of adolescents being introduced to a new academic subject.

References Bandura, A. (1982). "Self-efficacy mechanism in human agency." American Psychologist 37: 122-147. Fong, C. J. and J. M. Krause (2014). "Lost confidence and potential: a mixed methods study of underachieving college students’ sources of self-efficacy." Social Psychology of Education 17: 249–268. Miyoshi, A. (2012). "The stability and causal effects of task-specific and generalized self-efficacy in college." Japanese Psychological Research 54(2): 150-158. Nielsen, T., et al. (2018). "Gender Fairness in Self-Efficacy? A Rasch-Based Validity Study of the General Academic Self-Efficacy Scale (GASE)." Scandinavian Journal of Educational Research 62(5): 664-681. Rimm, H. and M. Jerusalem (1999). "Adaptation and validation of an estonian version of the general self-efficacy scale (ESES)." Anxiety, Stress, and Coping 12(3): 329-345. Sherer, M., et al. (1982). "The self-efficacy scale: Construction and validation." Psychological Reports 51: 663-671. Stankov, L. and J. Lee (2017). "Self-beliefs: Strong correlates of mathematics achievement and intelligence." Intelligence 61: 11-16. Stoet, G. and D. Geary (2018). "The gender-equality paradox in science, technology, engineering, and mathematics education." Psychological Science 29(4): 581-593. Talsma, K., et al. (2019). "Miscalibration of self-efficacy and academic performance: Selfefficacy≠ self-fulfilling prophecy." Learning and Individual Differences 69: 182-195. Usher, E. L. and F. Pajares (2008). "Sources of Self-Efficacy in School: Critical Review of the Literature and Future Directions." Review of Educational Research 78(4): 751–796.

ABSTRACT. Welfare states rely on human resources to provide the social services available. Social workers are often responsible for linking services provided and the families in need. In this paper we quantify the differences in effectiveness across social workers, and study the implications of these estimates for their optimal allocation across families. To do so, we use data from a two year home visitation program for the most deprived in Chile, where social workers were (quasi)-randomly assigned across families. We start by estimating the social workers’ fixed effects, that reveal substantial heterogeneity in the value-added by social workers. Second, we find that there is heterogeneity in performance within social workers, across different families. For example, a higher value added in the work domain weakly correlates with the higher valued added in the domains of education and housing. Third, we find a weak correlation between the supervisor’s and own social worker evaluation and her estimated value added. Finally, a high value added social worker has a small but significant impact which lasts up to four years after the home-visitation program.

ABSTRACT. The capacity to absorb a financial shock is a key principle of financial wellbeing, yet many adults report that they would struggle to do so without borrowing. There is little research, however, on how to encourage saving for emergencies in high income countries. Most field trials of saving behaviour focus on retirement or lower income countries where access to banking technology is less pervasive. We present a pre-registered field trial that tested two ways to encourage the uptake of a “rainy day” savings account, among a sample of high-street bank customers in Ireland. The study was commissioned by a national consumer protection body.

The first intervention was an electronic direct mail (EDM) campaign that communicated the risk of financial shocks to customers. The main idea was that, while the probability of facing a specific financial shock (e.g. a car breakdown) might be low, low probabilities accumulate. Hence, the probability of facing any financial shock (e.g. a car breakdown or household repair or medical bills or technology replacement) is much higher. The EDM featured an animated GIF and included a link to a “Savings Need” quiz. This intervention was informed by two surveys of nationally representative samples of consumers (N1 = 1,500; N2= 1,000). The “behavioural EDM” was compared against a “standard EDM” that encouraged customers to save to improve their financial wellbeing and featured a link to a “financial wellbeing health check.”

The second intervention consisted of changes to the bank’s existing savings account application form. Changes were informed by a review of over 50 experiments on saving and consisted of (1) altering the name of the savings goal (to “unexpected need”), (2) providing tips on how much to save and when to start, (3) an interactive calculator to help customers understand the relationship between amount saved per month and when they would hit their target and (4) a soft-commitment pledge to withdraw savings only for specific reasons. The control condition was the bank’s pre-existing application form. We hypothesised that our interventions would (i) increase savings account uptake and (ii) affect amount saved.

Method The study ran for 26 weeks from mid-May 2021. The design was 2 (EDM: behavioural vs. standard) x 2 (application form: behavioural vs. standard) between groups, giving four experimental treatments. Participants were 138,568 of the bank’s customers who met inclusion criteria and were randomised into one of the four treatments. Inclusion criteria were that the customer has a current account with the bank, was not in arrears, did not already have the savings account and consented to receiving marketing materials from the bank. An additional 22,732 customers formed a “no contact” control condition, who were otherwise eligible for the trial but did not receive any communications nor have access to the behavioural application form. We also ran an additional A/B test of the application form with customers who did not receive communications but opted to take out the savings account over the study period (“organic” uptake, n = 993).

Results Both interventions led to more account openings (EDM p = .069; application p = .027), although there was no interaction (p = .257). Compared to the Standard EDM-Standard Application group, there was a 26% increase in account opening among groups who received at least one experimental treatment. The effect persisted when the small number of account closures were factored in. There was no difference between the no-contact control group and the Standard EDM-Standard Application group. Socio-demographic characteristics were available for a subset of customers (n = 61,610). When these were added to logistic regression models predicting accounts opened, the effects of both interventions strengthened (EDM p = .009; application p = .004; interaction p = .118). There were no interactions between the interventions and the socio-demographic variables, implying similar effects across groups.

Participants in the Behavioural EDM-Standard Application group saved slightly less on average, although neither intervention led to statistically significant differences (EDM p = .827; application p = .527; interaction: p = .867). Turning to the organically opened accounts, the conversion rate (completed applications as a percentage of started applications) showed a 39.5% increase in completed applications for the behavioural form compared to the standard form (Figure 1; p = .022). Those who completed the Behavioural Application had saved marginally less on average than those who completed the Standard Application at the end of the study period (Mdn = €600 vs. €840; p = .088). However, those who completed the Behavioural Application were significantly less likely to be in the highest income grouping (earning more than €40,000 per year; p = .009). Hence the difference in savings balances may be attributed to different sociodemographic characteristics of savers, rather than a decrease in willingness to save; the effect of the Behavioural Application on amount saved is non-significant when income is added as a control (p = .892).

Of course, saving more does not necessarily lead to better financial wellbeing, particularly if it leads to increased borrowing. We also assessed combined credit card and loan balances and current account overdrafts among those who opened savings accounts and found no differences between experimental treatments.

Discussion The results show that both interventions increased the rate at which people save for financial shocks. An EDM campaign that communicated the accumulation of the probability of shocks and a behaviourally-informed application form led to independent increases in the number of customers who opened savings account. The behaviourally-informed application form seemed to be particularly effective at encouraging saving among those with lower incomes, who may be a higher risk of borrowing if they face a financial shock. Importantly, none of these interventions affected indicators of debt, implying that customers did not save at the expense of necessary spending. Together, the interventions highlight the potential for behaviourally-tested interventions at improving financial wellbeing.

ABSTRACT. Hełka and Wójcik in qualitative study (2019) identified the social norms essential for the process of incurring and repaying financial liabilities for Poles. They found number of differences between people with various indebtedness experiences in terms of shared norms and the way they construct their financial reality. Model borrowers have a distinctly different approach to debt than unreliable debtors and non-borrowers. Model borrowers seem to be oblivious of the dark sides of loans. They also indicate less reasons for not repaying obligations than others. Non-borrowers emphasize that credit brings higher standard of living but also permanent stress and psychological burden. Unreliable debtors treat loans as an immediate way to solve their financial problems and have their own private rules for borrowing. By following the SAGE model (Power et al., 2018) and current standards in economic psychology (Ranyard, 2017) we conducted quantitative study to verify the results on a larger sample. At the turn of 2018 and 2019, 235 Poles (55.8% woman) between 19 and 80 years old (M=39,52, SD=13,31) completed consciously and voluntarily the anonymous survey online or (if they prefer) in the p&p version (N=28). Among them were 129 model borrowers, 64 unreliable debtors and 42 non-borrowers. Respondents express their agreement with the 26 statements relating to social norms concerning financial liabilities as well as reasons and justification of the deviations from repaying financial liabilities and credit use association revealed in the previous study (Hełka & Wojcik, 2019) on the 10-point scale from 1 - I totally disagree to 10 - I fully agree. They also answer question concerning: their previous indebtedness experience and borrowing plans, material situation and demographic data. In line with previous qualitative study numerous differences between Poles with various indebtedness experiences have been found (see Table 1) and will be discussed in the course of presentation.

Table 1 Descriptive statistics for the level of agreement with the statements relating to social norms concerning financial liabilities as well as reasons and justification of the deviations from repaying financial liabilities and credit use separately for non-borrowers, model borrowers and unreliable debtors Item Non- borrowers M(SD) Model borrowers M(SD) Unreliable debtors M(SD) Nowadays, having a loan is the norm 6.31(2.70) 7.24(2.44) 7.66(2.07) In order to buy a car, furniture, home appliances, etc., I would take a loan 2.55(2.07) 5.60(3.05) 5.59(3.50) When I take out a loan, I know exactly how I will spend the money I borrow 8.33(2.64) 9.33(1.28) 8.73(2.09) When I borrow money I know where I will get money to pay back the loan and that I will pay the loan back on time 8.17(2.58) 9.31(1.31) 8.33(2.25) Credit is a method for me to solve a problem quickly 2.50(2.32) 3.57(2.72) 4.13(3.20) Sometimes I spend more than I earn using a credit card or other type of loans 1.67(1.43) 2.61(2.46) 3.73(3.18) Loans are taken for the purpose of investing in something to earn and are not within the reach of the average citizen 4.26(2.44) 4.16(2.81) 4.41(3.10) I would only take a loan as a last resort if I had no other way out of, for example, if I needed money to treat myself or my relatives 7.88(2.59) 6.87(3.06) 6.19(3.15) I would not live beyond my means by borrowing money that I will not be able to pay back on time 9.45(1.44) 8.09(2.98) 7.45(3.19) I would not borrow money from someone who does not have much of it himself 8.57(2.21) 7.98(3.10) 8.03(2.94) I would borrow money knowing that I would not repay the loan 1.50(1.09) 1.40(1.25) 2.50(2.67) If I had any unpaid commitment, I would not avoid contacting the creditor, but rather try to solve the problem 8.48(2.06) 8.78(2.21) 7.62(2.74) Sometimes I forget to pay some liabilities (rent, utility costs, invoice, ticket, etc.) on time 2.69(2.49) 3.32(2.90) 5.41(3.09) Paying back each loan is a priority for me 8.90(1.39) 8.80(1.91) 7.72(2.68) I am in arrears with some fees, because others are in arrears 1.52(0.89) 1.42(1.22) 2.95(2.71) If I had any obligations not repaid on time (bills, debts, etc.), I talk to my relatives about it, 6.62(2.89) 6.11(3.15) 5.47(3.25) Failure to pay small liabilities or loans is normal 1.60(1.21) 1.91(1.93) 3.11(2.67) Our debts are a taboo topic that you don't talk about with friends 5.48(2.93) 4.73(2.93) 4.63(2.86) If the creditor does not claim payment, I may not pay the obligation 1.79(1.51) 1.50(1.12) 2.23(2.17) If I am dissatisfied with the service or purchase, I may not pay 2.81(2.20) 2.65(2.35) 2.53(2.12) I may not pay back a loan that I did not take out myself, but only inherited 4.05(3.08) 4.17(3.23) 3.58(3.08) I may not repay the loan that I did not take out, but only gobbled. 2.57(2.09) 2.98(2.38) 2.48(2.23) I avoid taking loans, because living with a loan would be too stressful 8.24(2.58) 5.91(3.13) 5.23(3.31) Loans are expensive and only available to wealthier people 5.40(2.88) 4.73(2.71) 4.86(2.94) Credit is a source of stress, nerves and discomfort for me 8.00(2.18) 5.76(2.95) 5.63(3.11) Despite unpaid financial liabilities, you can live a normal life 3.12(2.32) 4.24(2.84) 4.70(2.94)

References Hełka, A. M., & Wójcik, M. (2019). Social norms in the process of incurring and repaying financial liabilities among Poles with various indebtedness experiences. Polish Psychological Bulletin, 1, 54-62. DOI - 10.24425/ppb.2019.126019. Power, S. A., Velez, G., Qadafi, A., & Tennant, J. (2018). The SAGE model of social psychological research. Perspectives on Psychological Science, 13(3), 359-372. Ranyard, R. (Ed.). (2017). Economic Psychology (Vol. 2380). John Wiley & Sons.

ABSTRACT. In a series of seven experiments, conducted in both field and laboratory settings, we demonstrate in the context of small informal loans between friends, that individuals operate under different mindsets depending on their role in the loan (lender vs. borrower), which consequently leads to misaligned repayment expectations. Lenders, operating under a communal mindset, do not expect borrowers to repay small amounts that borrowers, operating under an exchange mindset, intend to repay. Paradoxically, we also find support that lenders (vs. borrowers) hold stronger memories of microloaning incidences. We explain how the two-mindset hypothesis can reconcile these seemingly contradicting results, as well as why many small informal debts remain unpaid. We discuss the contribution to social and economics literature, implications for interpersonal relationships, and directions for future research.

Using real world social interactions with close friends, Study 1 (92 confederates providing 179 episodes of borrowing large and small amounts money from friends) tested whether the friends will request a repayment. The results show that borrowed amount is positively associated with the likelihood of a repayment request. In Study 2 (n=358) participants were assigned to one of three conditions (borrower vs. lender vs. observer), read a scenario about a lending situation between friends and indicated the minimum perceived amount for returning the loan. We found that borrowers indicated a smaller amount of money from which loans should be returned, than lenders had indicated while observers indicated intermediate amounts. Study 3 (n=500) further manipulated, instead of measuring, the cost of the meal (i.e., loan size) with the following values: $10, $20, $50, $100 and demonstrated that the repayment expectation gap diminishes as loan amount increases. In Study 4 (n=400) participants were assigned to one of four experimental conditions; (role: borrower vs. lender) x (loan-size: $5 vs. $500), imagined that they borrowed money from/to one of their friends, and reported the extent to which they would expect to return /receive the loan. Participants also reported the extent to which they perceive the situation as “a loan or a friendly-gesture”. The study revealed a significant indirect effect of perception suggesting that the effect of role on expectation was partly driven by the extent to which individuals perceived the situation as a social exchange. Study 5 (n=401) provided further support for the mediating role of mindset (economic vs. social) by manipulating it directly. It demonstrated that the gap can be remarkably reduced when borrowers’ and lenders’ mindsets converge. Study 6 (N=303) provided evidence for the moderating role of norm conformity by showing that the effect is also qualified by individuals’ personality traits (e.g., communal and exchange orientations). In Study 7 (N = 100) participants played 10 rounds of a game while in each round they rolled two standard dice and reported their outcome for money. In each round, participants had to decide whether they preferred to convert their points into money or purchase a lottery ticket that would cost them 7 NIS (if they managed to roll 7 points). Participants could ask for a loan from the other player if they rolled an outcome lower than 7 points or could give a loan to their game partner if they managed to roll more than 7 points. After participants completed 10 rounds of the game, they were asked to recall, for each round, whether they gave a loan to the other player, received a loan from the other player, and the amounts given or received. We found that people are more likely to remember details related to microlending than details related to microborrowing.

Contribution. The research provides a demonstration of a systematic gap between borrowers and lenders in the extent to which they expect small informal loans between friends to be paid back. This gap, which stems from borrowers and lenders operating under different mindsets, together with the absent of a repayment enforcing mechanism, may explain why a large portion of micro loans remain unpaid. Anecdotal evidence points to the existence of an immense market of unpaid small debts between friends and acquaintances. These transactions are likely to involve psychological factors that are absent in large loans governed by formal agreements and rules. Surprisingly, however, social scientists have made little attempt to explore this ubiquitous phenomenon. Across individuals from three different countries, using lab and field data, and observations of nearly 3,000 hypothetical and consequential decisions, this research provides converging evidence for a gap between borrowers and lenders’ perceptions of small interpersonal financial assistance. Borrowers, governed by an exchange mindset, expect to repay monies that lenders, governed by a communal mindset, don’t expect them to repay. These differences in mindsets may not only explain why individuals are willing to loan small amounts despite the credit risk involved, but also why microlenders are unwilling to request repayment when recipients fail to repay, which consequently increases the probability of the loan remaining unpaid. This result has important implications for interpersonal relationships.

ABSTRACT. The purpose of this study is to investigate what kind of saving and spending motives customers of a digital bank have for their accounts, and whether there is a relationship between motives and financial difficulties. We have chosen to call this kind of division of funds “digital mental accounting” since we are looking at earmarking of money in a digital technology setting.

In a time marked by considerable changes in the financial environment and a growing number of opportunities for spending, having control over personal finances can be difficult. This is reflected in the vast amount of consumer debt and loan defaults we see today (Finanstilsynet 2021). Consumers may gain more control over their finances by using budgets and set saving goals (Rabinovich and Webley 2007), but research shows that a minority of households use budgets (OECD 2016). One reason people tend to report that they do not use budgets, is that digital technology has transformed the way consumers plan their finances. One tool that can be used when planning, and which also has been found to reduce self-control problems related to spending (Soman and Cheema 2011) is mental accounting. This refers to cognitive operations used to organize and keep track of financial activities (Thaler 1999). One way of using mental accounting is by categorizing or labeling household income into spending or saving categories (e.g., “holiday”, “entertainment", “health”, and so on). Research has shown that when money has been divided into multiple accounts where each account has a specific purpose, this will increase the psychological cost of spending money that was set aside for a different purpose (Soman and Cheema 2011). In this study, we use field data from a digital bank to explore how widespread the use of digital mental accounting is, the characteristics of people using several bank accounts as a form of digital mental accounting, and what specific purposes is the most typical among those who use digital mental accounting. We also explored if there are any relationships between the use of digital mental accounting and the labels of digital accounts and financial difficulties.

Building on the work by Xiao and Noring (1994) who are using Maslow’s pyramid of needs as a conceptual framework when studying saving motives, we use the following categories: Daily (account for daily expenses), Purchase (account for purchase plans (e.g. car, travel, second home, and so forth), Emergency (account for an emergency (e.g. Buffer, health)), Retire (account for retirement), Child (account for children or grandchildren) and Grow (account for a better life, advanced standard of living, or other intangible reasons).

Our analyses are utilizing data from a survey conducted on customers of a Norwegian digital bank (n = 594) and included 3 questions measuring self-control, 1 on savings motives, 2 on personal finances and 2 on mental accounting along with 11 on socio-economic characteristics.

Preliminary analyses identified a gender difference in digital mental accounting: women are more likely to have earmarked accounts for retirement and children, whereas men are more likely to have separate accounts for activities supporting growth. These gender differences in motives differ from previous studies which have found men to be more likely to save for retirement, children, and growth, while women typically earmark their income for “daily expenses” (Xiao and Noring 1994). This finding suggests that more women than before are aware that their lower income and lower contributions to the public and private pension systems require personal saving to uphold their standard of living in retirement.

Overall, the younger generations used digital mental accounting more than the older generations and were more likely to have separate accounts for daily expenses, purchases, and emergencies. This does not necessarily mean they use mental accounting more often, but they are better at making use of new technology when planning their expenditures and saving. This was not surprising, seeing as they have grown up with technology and therefore are likely to be more comfortable using these kinds of tools. The results showed that the likelihood of having a separate account for emergencies decreased with increasing levels of self-control. One explanation for this finding could be that people with high self-control are better at saving in general and do not need an emergency account. As expected, we found that the probability of having financial difficulties increased as level of self-control decreased. But, controlling for self-control level, people with separate accounts for purchases, retirement or “other” were less likely to have financial difficulties. This finding suggests that digital mental accounting may be a helpful tool for keeping track of and planning personal finances so that financial difficulties are avoided.

The findings suggest that digital mental accounting may have replaced traditional budgeting as a tool for keeping track of and planning spending and saving. We find gender differences in mental accounting, and the results suggest that digital mental accounting may be a useful tool that reduces the likelihood of financial difficulties. This is important insight for financial advisors and educators.

References: 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/.

OECD (2016). OECD/INFE International Survey of Adult Financial Literacy Competencies. OECD. WWW.oecd.org, OECD.

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.

Soman, D. and A. Cheema (2011). "Earmarking and partitioning: increasing saving by low-income households." Journal of Marketing Research 48(SPL): S14-S22.

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. Conclusions of behavioral economics show that an individual may not act optimally and rationally in his decision-making. However, those findings, coming mainly from laboratory conditions, need to be verified in the real environment. This paper examines the presence of the representativeness heuristic in the environment of fantasy sport. The data come from a fantasy league according to National Hockey League in season 2015-16. Its users come mainly from the Czech Republic. The results show a moderately strong presence of representativeness. Nevertheless, the irrationality manifests for example by higher demand for the same nationality hockey players. However, the main influence of demand is still the athletes’ performance.

ABSTRACT. Insufficient exploration of one’s surroundings is at the root of many real-life problems, as demonstrated by many famous biases (e.g., the status quo bias, learned helplessness). The current work focuses on the emergence of this phenomenon at the strategy level: The tendency to under-explore the set of available choice strategies. We demonstrate that insufficient exploration of strategies can also manifest as excessive exploration between options. In such cases, interventions aimed at improving choices by reducing the costs of exploration of options are likely to fail. In Study 1, participants faced an exploration task that implies an infinite number of choice strategies and a small subset of (near) optimal solutions. We manipulated the amount of practice participants underwent during the first, shorter game and compared their performance in a second, longer game with an identical payoff structure. Our results show that regardless of the amount of practice, participants in all experimental groups tended to under-explore the strategy space and relied on a specific strategy that implied over-exploration of the option space. That is, under-exploration of strategies was manifested as over-exploration of options. In Study 2, we added a constraint that, on a subset of practice trials, forced participants to exploit familiar options. This manipulation almost doubled the per-trial average outcome on the comparable longer second game. This suggests that forcing participants to experience the effects of different (underexplored) strategy components during practice can greatly increase the chance they make better choices later on.

ABSTRACT. We examine high-stakes strategic choice using more than 40 years of data from the American TV game show The Price Is Right. In every episode, contestants play a simple sequential game of perfect information for which the optimal strategy can be found through backward induction. We find that contestants systematically deviate from the unique subgame perfect Nash equilibrium and from the equilibrium of an agent quantal response model. Omission bias cannot explain these departures from optimality. Instead, the behavior of contestants is well described by a model of limited foresight, where a sizable proportion of the contestants myopically consider the next stage of the game only.

ABSTRACT. Introduction

While the quantity of household waste is increasing worldwide in total and per capita, the quality (i.e., purity of waste streams) is declining with substantial adverse implications for resource recovery and climate change. This insight is hardly new – household resource consumption and waste behavior have long been focus areas of behavioral and policy research. This research suggests that interplay between convenience factors, information and knowledge, and social norms are key to influencing behavior. Although strategies to influence household resource consumption, such as social norm-based interventions (Wolske et al., 2020), provision of feedback and information (Karlin et al., 2015), or increasing convenience (Schultz, 2014) have repeatedly been tested, the existing research suffers from several shortcomings that limit explanatory power and ability to produce actionable insights under realistic conditions: Short study periods; reliance on self-reported, grouped, or otherwise estimated data; and use of practically unscalable interventions (e.g., staff-intense procedures or manual distribution of material).

In this study, we exploit the latest technological developments to study waste behavior in a granular, long-term, and realistic setting: We influence household recycling behavior using a scalable and innovative combination of physical and digital interventions, and measure treatment effects longitudinally over the course of one year, using objective outcome measures with high data granularity under ecologically valid conditions.

Method

The study is set in the newly developed Stockholm Royal Seaport district in Stockholm, Sweden, where mixed municipal waste (MMW), plastic waste (PW), paper and newspaper waste (NPW) are disposed through an underground pneumatic waste system. The system’s inlets allow for household-specific, real-time weight measurement of waste, similar to smart electricity meters. Our study includes 154 households from three multi-family dwellings. We collected weight data at household-level for MMW, PW and NPW fractions over 12 months, starting June 2020.

We developed behavior-based interventions to improve recycling and waste sorting, specifically concerning PW and MMW. Our goal was to (i) investigate short and long-term effects of our interventions on households’ waste and recycling behavior in a real-world setting, and to (ii) assess whether app use moderates this effect.

We combined an information-based awareness campaign delivered through physical measures (i.e., posters; floor stickers; social event) with a digital social norm-based intervention delivered through a purpose-built app (personalized (social comparison) feedback; recycling information focused on PW). While we can reasonably assume that all households experienced the physical interventions, only households voluntarily joining the app were exposed to its content.

Results

Throughout the 12-month experimental period, around 40% (n=62) of households joined the app and roughly half of them used the app at least twice (n=34).

Due to space constraints, we focus our analysis on a simplified comparison between a pre-intervention period (10-week average prior to intervention) and a joint post-intervention period (average of all weeks after intervention), using three 2 (app user: yes vs. no) × 2 (time: pre-intervention vs. post-intervention) ANOVAs with app-user as between-subjects factor and time as within-subjects factor. We also analyzed our data in a more granular manner using six ten-week time periods (reported in full paper).

We find a significant main effect of time on plastic (p < .001) and paper waste (p = .02), but not on mixed municipal waste (p = .33). Average weekly waste amounts increased especially for plastic, with households disposing about twice as much plastic waste per week after (M = 0.19 kg, SD = 0.22) compared to before the intervention (M = 0.09 kg, SD = 0.17). For plastic waste, we also find a significant main effect for app user (p = .03) and an app user × time interaction (p = .03). Importantly, app users’ and non-users’ amount of plastic waste differed significantly after the intervention (p < .01; Mapp user = 0.25 kg, SD = 0.25 vs. Mno app user = 0.15 kg, SD = 0.19), but not before (p = .30; Mapp user = 0.11 kg, SD = 0.16 vs. Mno app user = 0.08 kg, SD = 0.17). In terms of paper and municipal waste, there was no significant difference between those households using and not using the app, as both the effects of app user and the two-way interactions were non-significant for both waste fractions. Our more granular analysis shows that these effects are stable over the 12-month period.

Discussion Based on an intervention focusing on household waste behavior, using data from 154 households collected over an entire year, the results revealed that all households threw significantly more PW and NPW post-intervention, although MMW levels remained stable across the study period. Moreover, those households also using the app generally threw more plastic than non-users post-intervention, indicating a particularly pronounced effect on app users’ PW sorting.

This, however, does not explain why MMW did not decrease accordingly. Considering the circumstances, this could either be an indication of a general increase in waste volumes during the Covid-19 pandemic or the result of a shift in disposal routines. Given that some research views recycling and reducing waste as related but different behaviors (Ebreo & Vining, 2001), it remains a question for future research whether one set of interventions can change both behaviors simultaneously.

In conclusion, this study presents longitudinal evidence that combined behavior-based interventions can significantly influence household waste behavior in real-life settings, and that these effects remain stable over extended periods of time. Further, it suggests that recent technology can strengthen the behavioral impacts of physical interventions even under conditions with low experimental control. Future research should continue to explore the potential of combined digital-physical interventions and novel technologies for improved data collection under ecologically valid conditions.

Ebreo, A., & Vining, J. (2001). How Similar are Recycling and Waste Reduction? Environment and Behavior, 33(3), 424–448. https://doi.org/10.1177/00139160121973061

Karlin, B., Zinger, J. F., & Ford, R. (2015). The effects of feedback on energy conservation: A meta-analysis. Psychological Bulletin, 141(6), 1205–1227. https://doi.org/10.1037/a0039650

Schultz, P. W. (2014). Strategies for Promoting Proenvironmental Behavior. European Psychologist, 19(2), 107–117. https://doi.org/10.1027/1016-9040/a000163

Wolske, K. S., Gillingham, K. T., & Schultz, P. W. (2020). Peer influence on household energy behaviours. Nature Energy, 5(3), 202–212. https://doi.org/10.1038/s41560-019-0541-9

ABSTRACT. Theory and research question Correct waste disposal behaviour is a central factor of the circular-economy. Waste that is not littered but correctly disposed of and separated in waste bins allows recycling and a re-use of natural resources which in turn protects the environment and the climate. While many people separate their waste in their homes, waste separation in the public sphere (or ‘on-the-go’) is less common. Related to that, there are only few studies that investigate waste separation in the public sphere. However, many studies investigate the related behaviour of littering in public. One main factor that might reduce littering relates to attentional cues that increase visibility of waste bins (Gerlach et al., 2018). For instance, the results of a field-experiment suggest that attentional cues such as watching eyes and nature pictures reduced littering more than educational interventions about the consequences for finance or health (Gangl et al., 2020). The aim of the present study was, first, to examine whether separate waste collection qualitatively works in the public sphere or not and second, whether an improved visual design of the collection bins can further foster correct waste disposal. In addition, we wanted to test whether meaningful attention (i.e., thematically related bin design with beautiful nature pictures) can increase correct waste separation and reduce littering more compared to neutral attention (i.e., colourful bin design with high contrast pattern). Method: To test our hypothesis, we conducted a field experiment in three Austrian cities (Krems, Steyr, Leoben) involving 90 collection points, where normally only one residual waste bin was located and no collection of targeted recyclable material (plastic, metal or composites) took place. We randomly assigned four experimental conditions, stratified for different characteristics of the location: control condition with a residual waste pin and no possibility to separate waste, additional standard collection bin to separate waste, additional attention-neutral collection bin to separate waste, additional attention-nature collection bin to separate waste. To control for time effects such as weather, we applied a pre-post design: we assessed waste behaviour two weeks before and two weeks after the new bins were installed.

As dependent variables, we assessed 15 different categories of collected waste materials in Kilogram and 15 different categories of litter counted 10 meters around the collection pin over a period of four weeks (three times a week). In addition, we assessed as number of situational variables (e.g., proximity of other waste bins, location appearance and lighting) and time variables (e.g., busyness, temperature). Results Waste separation in public was successful. On average, the separate waste bin contained more of the targeted material (plastic, metal, composite) than the adjacent residual waste bin. Looking at the category-level, in particular the collection of plastic worked. The collection of metal was only successful in pedestrian areas and the collection of composites rarely worked. Correlational analyses show that in general, separate waste collection worked best in pedestrian areas compared to other areas (e.g., stations). In addition, locations next to fast-food restaurants collected large amounts of the targeted material which was, however, contaminated. Thus, the present results suggest that separate waste collection is most likely to be successful in pedestrian areas. However, with some behavioural interventions (e.g., a reminder to separate waste by staff) waste separation next to fast-food restaurants could possibly also be improved. The different experimental conditions did not differ in terms of waste separation. In terms of total waste volume, results also show that the attention-nature bin condition led to the collection of more waste than the control condition, but that there was large variation across specific measurement days. Further analyses indicate that the attention-nature bin might be especially effective in “grey” public areas compared to “green” locations such as parks. This confirms that visibility and thus attention for the bin, indeed explain the successful collection of waste (that is at least partly, lower likelihood of subsequent littering). The result that only the nature bin but not the high-contrast bin had a strong effect provides some evidence to the hypothesis that attention which conveys meaning might offer some additional positive effect. Results also showed that littering of cigarette stubs occurs relatively frequent whereas other waste is seldomly littered. However, correlational analyses show for instance that littering is reduced in observable and lighted places. The different experimental conditions had no sizable effect on littering. To sum up, the present research shows that separate waste collection also works in the public sphere, particularly in pedestrian areas. However, further targeted interventions are required to possibly improve separate waste collection (e.g., next to fast-food restaurants or for other materials than plastic). In addition, designs which create visibility and in addition also attention to the cause, particularly, through nature pictures, might increase the amount of collected waste. References: Gangl, K., Grosch, K. & Walter, A. (2021). Mehr Sauberkeit im Gemeindebau. Ergebnisse eines verhaltensökonomischen Feldexperiments in den Müllbereichen. [Increasing cleanness in social housing. Results from a behavioural economics’ field-experiment in waste disposal areas.] Research Report of the Institute for Advanced Studies (IHS). https://irihs.ihs.ac.at/id/eprint/5654. Gerlach, R., Beyer, R., Foerges, R., van der Meer, E., & Nimke-Sliwinski, B. (2018). Evaluation von zielgruppenspezifischen Antilittering-Maßnahmen im Feld mit Hilfe des Einsatzes von Nudging. [Evaluation of target group specific anti-littering interventions in the field based on nudging.] In Beyer, R., Krause, B. & Nachtwei, J. (Hrsg.). Empirische Evaluationsmethoden Workshop 2018 (pp. 5–22). ZeE Verlag. https://www.researchgate.net/profile/Jens-Nachtwei/publication/336902731_Empirische_Evaluationsmethoden_Band_23/links/5db9d597299bf1a47b02555c/Empirische-Evaluationsmethoden-Band-23.pdf

ABSTRACT. A pressing problem of today’s civilization are on the one hand the mountains of trash that are increasing tremendously and on the other hand the valuable raw materials that are mined for countless consumer products in huge quantities so that very soon not much will be left. One solution to both problems certainly is the systematic separation of waste so that the valuable materials can be recycled and used for new products. Although recycling in the western world has started decades ago, it often focuses rather on waste in the private household than on waste in public and semi-public space. The increase of waste separation in (semi-)public space is a next step to save valuable resources and reduce mountains of trash. But how can consumers be motivated to separate their waste in the public space? The current study investigates not only psychological factors but also environmental cues that motivate as well as hinder waste separation in the public and semi-public space. Methodologically, we have a multimethod approach and use (1) observations and (2) interviews with consumers. Starting with observations of consumers in the street and office buildings disposing of their waste, we investigate how these persons behave with the waste in the (semi-)public room and whether they are willing to separate this waste. Beside the observations of human behavior also waste bins that allow for separation are studied. We investigate where waste bins are situated and how they are designed. Also, we examine how these features are related to the visible amount of waste that is collected with these waste bins. These observations yield a first base for the development of research questions to be answered with a further investigation. Based on the observations and on literature research, interviews are conducted with 20 consumers using regularly public transportation in the Austrian capital Vienna and the Austrian country town Krems. The interview guideline comprises several aspects: (a) A very detailed description of a situation, in which the interview partner had to dispose of an item in the (semi-)public space (environment, feelings, values, etc.), (b) aspects that facilitate or hinder the separation of waste in the (semi-)public space (e.g., labels & pictograms on waste bins, familiarity with the situation, etc.), (c) an assessment of the motivating effect of new designs (color, labels, pictograms, etc.) of waste bins to separate different fractions of waste. In addition to the consumers, four cleansers in the respective city or town responsible for the emptying of the waste bins close to public transport are interviewed. From them it is requested (a) to describe a typical situation when cleaning and emptying the waste bins (design of waste bin, attitudes, etc.), (b) an evaluation what facilitates the emptying of the waste bin and what hinders it (design of waste bin, experience, (c) an assessment of new designs for waste bins (color, labels, pictograms, etc.) to facilitate emptying them. All interviews last about 60 minutes and are recorded and transcribed. The transcripts are analyzed with Grounded Theory and comprehensively reveal the factors that facilitate but also hinder the separation of waste in the (semi-)public space. Based on these findings, recommendations for waste industry but also legislators are highlighted and discussed.

ABSTRACT. Organizations and policymakers increasingly rely on economic incentives to prompt participation in activities amongst those who were previously not engaged. We ran a field experiment with a recycling program to examine incentives’ effectiveness to motivate new behavior—i.e., attract non-recyclers. We compared standard contingent incentives (payment contingent on recycling) to non-contingent incentives (upfront unconditional payment) of different sizes. A high contingent incentive was as effective as a non-contingent incentive (of any size) in attracting people to the program, but this masked differences in who participated. Across incentive sizes, people who had never recycled were 5.8 times more likely to begin recycling with the program when given a non-contingent incentive (20.2%) than when offered a contingent one (3.5%). A second experiment conceptually replicated this effect in an online job market, showing that non-contingent incentives were substantially more effective in attracting previous non-compliers.

See paper: Córdova, A., Imas, A., & Schwartz, D. (2021). Are non-contingent incentives more effective in motivating new behavior? Evidence from the field. Games and Economic Behavior, 130, 602-615.

ABSTRACT. Many individuals travel between countries as part of their professional routines. How do they perform during those short trips abroad? To begin to answer this question, I analyzed the outcomes of over 5 million chess games played around the world. Importantly, tournament chess provides a clean setting in which location-dependent factors are mostly irrelevant; the audiences are quiet and the referees make hardly any judgments. Controlling for differences in chess skills, I found enhanced performance among players who were competing outside of their home countries. This finding was robust to additional controls such as age, sex, and skill momentum or game practice, and to the inclusion of individual or country fixed effects. This advantage, an approximately 2% increase in game outcome, suggests that traveling has a positive effect on performance.

ABSTRACT. Building on seminal research, the topic of “bridging the gap” between utility and psychological motivations is now a common one in the transportation literature (for a survey see De Vos et al. 2016). Part of this research is focused on the relationship between transport and emotion, and in particular the influence of emotion on users’ decision-making regarding mode choice, independently of standard economic factors. The objective of this paper is to integrate emotion directly into individual economic preferences and to show how significant emotions are compared to standard transport-related economic variables. Based on data collected in Lyon (France) in 2015 by the Household Transport Survey (HTS), we first define and classify affective and economic variables. In order to measure each variable’s impact on mode choice, we estimate a binomial logit model based on standard economic variables and on the new affective variables. Our results show that basic emotions do affect individual decision-making at a significance level of 5%, but that the user’s current emotional state does not. Along with the quality variable, basic emotions appear surprisingly to be more determinant of transport-mode decision-making than standard price variables.

ABSTRACT. Darwin (1872) hypothesized that crying is a cultural phenomenon. Survey-based studies that have tested this hypothesis have been subject to self-reporting bias whereas criticisms towards experimental studies have mainly focused on their inability to elicit emotions sufficiently. In this paper, we address the methodological limitations associated with self-reporting and non-sufficient emotional elicitors by exploring data on gold medalists of all 450 individual events at the 2012 and 2016 Summer Olympic Games at the end of the medalists’ respective competitions and during the medal ceremonies to investigate the effect of individual, cultural and economic variables on crying. We find that age, gender, geographical region, belonging to the host country, religious fractionalization, and stereotypic gender roles (proxied by labor force participation rate of women in the athlete’s home country) are prominent predictors of crying. The logarithm of GDP per capita has no significant effect on crying. These findings suggest that emotional crying is not only a biological feature, but also a cultural phenomenon, in line with Darwin’s hypothesis.

ABSTRACT. Objectives and motivation: Psychometric data of bipolar scales yield bivariate complementary information. The order of magnitude of agreement towards an item assertion implies information about the order of magnitude of disagreement. The complementarity induces a compositional data structure which heavily differs from an interval or ratio scale (the so-called Simplex). The Simplex heavily affects the set of statistical procedures applicable and compositional data must not be evaluated using standard statistical procedures without proper preparation. Most standard statistical procedures (e.g., pearson correlation analysis, t-tests, linear regression) require interval scaled or ratio scaled data. Additionally, the psychometric limit of quantification (LOQ) of bipolar scales should be considered. While the existence of the LOQ seems reasonable its magnitude is scale specific.

Methods and results: For proper data analyses, the isometric log-ratio (ilr) transformation is suggested to overcome the limitations caused by the Simplex. Applying the ilr transformation to psychometric data provides interval scaled values which can be evaluated using standard statistical procedures. By means of a simulation study, we show that an increase in the statistical power of the well-known correlation test based on Student's t-distribution can be achieved. Depending on the specific parameters (magnitude of LOQ, dispersion etc.), up to an 18% increase in statistical power can be achieved indicating the superiority of the ilr approach compared to classical correlation analysis of psychometric bipolar data. As correlations are often used to assess criterion-related validity or convergent validity, the results of this study affect a large number of psychometric scales and analyses. For example, testing the hypothesis of equality of the two means of independent samples using a t-test based on Student's t distribution is equivalent to testing the hypothesis of a null-correlation between the binary grouping variable and the dependent variable. Additionally, as the coefficient of correlation contributes to the slope of a regression line the ilr approach also affects linear regression hypotheses testing. That is, the results of the simulation study generalize from correlation analysis towards a larger set of standard statistical procedures. Applying the ilr approach to real data we demonstrate that up to 21% of non-significant results can be reversed towards significance, avoiding the statistical type-II-error. Thus, less effects and relations among psychometric variables are overseen improving the statistical analysis of psychometric data. Alternatively, reducing the sample size is possible while retaining equal statistical power.

ABSTRACT. Ample academic work has linked conspicuous consumption to certain reproductive benefits; however, few studies have explicitly tested whether variables linked to intimate interpersonal relationships can predict conspicuous consumption patterns. Grounded in theories from evolutionary psychology, the present research investigated the extent to which conspicuous consumption is sex-specific, and whether relationship status plays a prominent role in shaping such consumption responses. Across two studies, including over 100,000 representative observations of conspicuous consumption from American households (Study 1) and a high-powered preregistered experiment (N = 420 corresponding to 8,400 data points) among American men (Study 2), we find converging evidence for the thesis that single men are particularly prone to consume conspicuous goods, both compared to mated men and women, irrespective of their relationship status. Importantly, these findings are not restricted to men’s actual relationship status but apply even by the mere perception of being single, as induced through short text-based scenarios, underscoring the managerial relevance of the current research to advertising and marketing practice. We conclude by discussing theoretical implications and offering future research directions.

ABSTRACT. Abstract

This study explores social representations of Bitcoin among French population. The aim of this paper is to assess the nature of Bitcoin) depending participant’s detention of Bitcoin. The financial crisis has led to a widespread loss of trust in financial institution such that central banks or banks. Some authors indeed suggested that the 2008 financial crisis triggered a crisis of trust (Sapienza and Zingales, 2009 ) and of confidence (Tonkiss, 2009 ). Yet, crypto-currencies (such as Bitcoin) are based on a technology (block chains) of exchange which is able to carry out secure transactions without the need for a – trusted – third party. Shortly, this technology enables to avoid financial intermediations in finance transactions. This reason could explain the enthusiasm for crypto-currency. More precisely, Bitcoin is generally defined as a decentralized electronic cash system initially designed and developed by Satoshi Nakamoto in October 2008. It is a “good” that exists only in digital form whose quantity is, by definition, artificially limited and whose property, by construction, can easily be transmitted by internet. The most important issue is to determine the economic interpretation of Bitcoin. Indeed, the status of Bitcoin as an alternative currency or a speculative asset is still unclear. To be considered as a currency, Bitcoin should fulfill three mains characteristics: store of value, unit of account and a medium of exchange. While individuals and some organization have begun to accept Bitcoin and other crypto-currencies as valid form of payment, fluctuations in Bitcoin prices challenge its role of store of values. Cheah and Fry (2015) have, for example, showed that the fundamental value of Bitcoin is zero. In this regard, as the speculative aspect of Bitcoin dominates its other functions, Bitcoin price appear to be at a mercy of nothing more that market sentiment. Then, some governments such as Germany, France, Korea and Thaïland have indicated a repudiation of Bitcoin as a currency. The academic literature in economy on crypto-currency has recently emerged especially about legal aspect, regulation and role in the payment system. However the literature on the aspect of the representation of the crypto-currency by laypeople is quasi-inexistent according to our knowledge. We decided to mainly focus this study on the representation of Bitcoin since it is the most well-known crypto-currency for people. Indeed, in December 2017 the value of one Bitcoin was equivalent of 13.889 dollars. It remains (until today) the highest value of Bitcoin. Since then the topics of Bitcoin begin to be a mass media topics and enable us to collect data over population. Data, using on online questionnaire, was collected between January 2018 and end of February 2018. We used a survey questionnaire which requires answers over a free association task over the target word “Bitcoin”, a financial literacy questionnaire (Lusardi and Mitchell, 2014 ), a risk-aversion questionnaire (Arrondel et al, 2004) and a socio-demographic questionnaire. Our main hypothesis is Bitcoin owner particiapnts have a different social representation of Bitcoin than participants who do not possess Bitcoin. We expected an effect of financial literary and risk-aversion. Indeed, Lusardi et al (2018) found that higher financial literacy is negatively associated with using mobile payments. We also made the hypothesis that Bitcoin owner participants have a low degree of risk aversion relative to as other participants (since the volatility of the price of Bitcoin was high when data was collected). Overall, 481 participants completed entirely our questionnaire among them 293 declared to own Bitcoin at that time. We found two social representations depending on participant’s detention of Bitcoin. According to the structural approach (Abric, 2001 ) the two identified social representations are different since their presumed central core are not similar. Consequently, For participants who detain Bitcoin, the central core is formed by the words “money”, “freedom”, “decentralized”, “future” whereas for participants who do not possess Bitcoin the central core is formed by the word “ money”, “virtual” , “crypto” and “cash”. Differences in the first periphery among participants highlight issue about legal regulation, financial bubble and trust in financial institutions. This article raises also the consumer protection issue in the financial sphere.

Sapienza, P., & Zingales, L. (2012). A trust crisis. International Review of Finance, 12(2), 123- 131. Tonkiss, F., 2009. Trust, Confidence and Economic Crisis. Intereconomics. 44, 196-202. Cheah, E-T., Fry, J. (2015). Speculative bubbles in Bitcoin markets? An empirical investigation into the fundamental value of Bitcoin. Economics Letters, 130, 32-36. Lusardi, A., & Mitchell, O. (2014). The economic importance of financial literacy: Theory and evidence. Journal of Economic Literature, 52(1), 5–44. Arrondel L., Masson A. et Verger D. (2004), « De la théorie à une enquête méthodologique originale », Economie et Statistique, n°374-375, p. 21-51. Lusardi, A., C. B. Scheresberg and M. Avery. 2018. Millennial Mobile Payment Users: A look into their Personal Finances and Financial Behaviors. GFLEC Insights Report, GFLEC. Abric, J.-C. (2001). L’approche structurale des représentations sociales: développements récents. Psychologie et société, 4(2), 81–103

ABSTRACT. Introduction. The activity of Russian citizens in modern financial markets has been growing steadily in recent years, although the degree of involvement is still significantly inferior to developed Western countries. Psychological and social aspects of Russian consumers' attitudes toward digital technologies are characterized by a fairly high predictive potential, which allows not only to develop diagnostic procedures and tools to identify the structural and dynamic characteristics of perception and readiness to use modern technologies, but also to form recommendations for its construction. The empirical study was conducted in May 2021, at a time when the COVID-19 pandemic not only increased economic and political uncertainty around the world, but also provoked an acceleration in the pace of digitalization in all areas of people's lives. The study focused not only on the factors of financial literacy and digital competence, but also on the behavioral or psychological factor of users' financial behavior. Method. The research included the development of a methodology to study the population's financial behavior when using FinTech, an online survey aimed at studying the financial behavior of Russians when using FinTech during the period of relative economic stability after two waves of COVID-19 virus - May 25-27, 2021 in Russia. The research was focused on the analysis of behavioral attitudes and reactions of the population and focused on comprehending the problems of economic security of individuals in the conditions of transition to the digital economy. We proceeded from important determinants or indicators influencing consumer attitudes in the situation of uncertainty in the financial sphere, which have already been studied in previous measurements on Russian samples: physiological manifestations of anxiety on financial stimuli, financial optimism, perception of insurance risks, experience of money shortage and financial̆ uncertainty, financial confidence (optimism) (Medianik, 2020). The inclusion of statements in the questionnaire construct to assess attitudes related to the successful use of digital technology, personal financial management, and responsible financial behavior became new markers in assessing the impact of FinTech on financial behavior. In our research we were faced with the need to formulate a definition of Digital Transformation Behavior - it is a change in consumers' financial behavior and adaptation to the requirements of the new digital economy. Construction of Digital Financial Behavior questionnaire. In this study, the author's questionnaire "Digital Financial Behavior", which included: a questionnaire of economic attitudes (Deineka, Zabelina, 2019); financial anxiety questionnaire (Medyanik, 2016); author's questionnaire of digital technology preference (Medyanik, Malyutina, 2021) was used for approbation. The questionnaire was compiled with the help of the staff of the Sociological and Internet Research Resource Center of the St. Petersburg State University Science Park. Сharacteristics of the sample. The reliability of the questionnaire was checked by Cronbach's Alpha coefficient, which was equal to 0.930. Thus, the fit of the model of 62 statements was determined to be good. The sample was 1,045 people (465 men, 578 women). Hypothesis making. Demographic variables served as guiding elements in the analysis and allowed for possible differences in the sample in terms of age, gender, employment, place of residence, and availability of economic education. We focused on testing the following risks of financial behavior as users transition to digitalization: 1. The risk of "unsuccessful" transition to the digital economy due to a lack of digital and financial competencies; 2. The risk of the influence of personal psychological barriers in the use of digital technology; 3. The risk of gender and age differences in the use of digital technology. Thus, the following hypotheses were put forward. The main hypothesis: with the approval and successful use of FinTech there is a new model of financial behavior - "digital financial consumer". We defined the "digital financial consumer" as a user with minimal knowledge and competence in the digital economy, which allows him to successfully adapt and use most digital financial instruments on a regular basis. This model is certainly a transformation of the consumer's financial behavior in a digitalized environment with new profiles and directions of types of financial activities. Results. Our research has shown that not only the structure, but also the very nature of digital financial consumption in Russia is changing under the influence of the ongoing digital transformation. Financial attitudes, digital skills, preferences and habits of Russian consumers are formed during their active digital socialization and largely depend on digital and financial literacy. This transformation of new "digital" skills and habits allows us to talk about the emergence of a new model of financial behavior of the "digital financial consumer. An important indicator of this model of consumer behavior in the digital reality is the successful use of the Internet to make online purchases of goods and services. But we are not talking about mass use of various FinTech platforms. In terms of engagement, this is only about half of our study sample. The main results of the study, which we would like to draw additional attention to: - most Russians already approve of the use of specific digital platforms, but few use them in practice; - women more often use FinTech platforms in everyday life than men (mostly purchasing goods and services through apps and paying via quick payments system) - More often the services of online payment services are used by citizens aged 26-35 years old. In the same group the indicator of digital financial literacy is significantly higher; what is noteworthy is that young people 18-25 years old, consider investing in cryptocurrencies through the application very effective in contrast to other groups. - Consumers with an economic education are more frequent and confident in using most FinTech products; - the "financial digital consumer" model has financial behavior profiles depending on the type of financial activity: "confident digital consumer," "investing digital consumer," "saving digital consumer," and "responsible digital consumer. The other profiles have not yet been discovered because FinTech platforms are not common in Russia.

ABSTRACT. Motherhood is not only children caring practice, but also a stage in the life path that affects of woman professional and economic behavior . Temporary disability, change in the usual rhythm of life and new items of expenditure form alternative economic models of behavior. Of course, motherhood is a universal value for all cultures, but in each country there are various socio-economic and social mechanisms to support women during this period. To identify the repertoire of filling the concept of "non-maternal practices of motherhood", microsociological theories are of particular importance (symbolic interactionism - J. G. Mead, I. Goffman; phenomenological sociology - P. Berger, T. Luckman; ethnomethodology - A. Schutz, G. Garfinkel) focusing attention on everyday social interaction mediated by personal meanings. So, a young mother is trying to find a part-time job, create an account on social networks, take refresher courses, etc. guided not by abstract judgments about the ideals of neoliberalism, but by a specific life situation, including the need to pay off a loan and her child's sleep schedule. I. Hoffman's concept of presenting oneself to others, which describes the mechanisms of strategic interaction, can be useful for understanding the features of mothers' self-presentations focused on projecting their motherhood (for example, studying Instagram activities). In Russia, maternity leave period is one of the longest in the world. It averages 140 days (70 days before delivery and 70 after), in the absence of complications in childbirth. Further, parental leave is granted until the child reaches the age of three years. A monthly allowance for a child caring up to 1.5 years old is paid in the amount of 40% of the average salary to the person caring for the child, it can be used by the mother, and most often in Russia it is she. It also provides for a one-time payment "Maternity capital", which the family can spend for specific purposes in accordance with the law. However, not every woman has the opportunity to fully use maternity leave, since in fact expenses exceed income. In addition, our study of scientists-young mothers allows us to talk about a surge in the professional activity of women after the birth of children. This thesis is also true for the Russian academic environment. We tried to understand whether this trend continues in other populations. We organized an online survey involving more than a thousand mothers with children aged 0 to 3 years (N=1321 as of January 2022, the survey is ongoing). Current results show that more than 75% of women are economically active during their maternity leave. Lack of money is the leading motive, but the next most popular answer is the desire for professional development. In Russia, more than 58% of specialists with higher education are women. In combination with traditionally active life attitudes, the desire for self-development looks convincing. The main types of economic activity were distributed as follows (in descending order): 40% entering a previous job, 9% entering a new job, 22% freelancing or starting their own business for the first time. It should be borne in mind that in Russia there are no tax benefits for entrepreneurs with children. The procedure for registering an economic entity is simple and fast, but the process of starting a business requires significant time and special knowledge. Nevertheless, young mothers decide to take this step, choosing as business ideas not only the areas labeled as "female" - the production and sale of clothing, cosmetology services and cooking, but also the production of furniture and equipment, high technologies. Young mothers receive online training, receive additional skills for secondary employment or a change of profession. Of course, the pandemic played a significant role, the labor market has changed - many business processes have moved online, which allowed young mothers to participate more actively in them (68% of young mothers who continue their professional activities have moved online). We found that more than half of the women surveyed did not conduct active economic activity online before the birth of a child (82%), did not receive additional income, except for the main place of work (94%), did not plan to financial position for the period of pregnancy and the beginning of motherhood (86%) - this indicates a low level of financial literacy, a short horizon of financial planning. At the same time, a woman's age, her level of education, and the area of residence have a significant influence. Thus, women aged 20-25 showed the greatest activity - they began active economic activity in social networks to promote their services or goods. Women over 35 have started their own business or entered official employment. Younger (under 20) and more mature women (over 45) showed lower economic activity. Most of the participants in the study 66.7% rely on the income of their spouse, because. state support remains extremely insignificant and cannot afford to support themselves and the child. At the same time, anxiety grows in the first year, with an understanding of the cost of the necessary basket of goods. Women indicate the need to look for additional sources of income, even at a normal level due to the unstable financial situation of the spouse. This situation forces a woman to take on the activity despite the pressure from society and the need to spend maximum time with a child. Early motherhood is one of the most financially vulnerable stages in a woman's life. Often, providing for the child becomes a task that the mother must solve on her own. This forms new practices of economic behavior of women who seek to balance between “good”, socially approved motherhood and career. Also, the activity of women creates new norms of motherhood and influences the structure of gender roles in society.

acknowledgement:rsf russian science foundation projekt №22-28-00636 "Modern "non-maternal" practices of young mothers: repertoire, potential and social risk"/