**1. Introduction**

Sustainable education involves an active academic engagemen<sup>t</sup> intended to create economic, social, and environmental programs able to improve living standards, generate empowerment, and respect interdependence [1]. To achieve this, a teaching approach rooted in ethics, values, social responsibility, and sustainability is required. In addition, it is necessary to revise the learning process of students. Deep learning techniques are aimed at helping students to develop a critical spirit and to reflect on concepts and paradigms in such a way that they are able to understand drawbacks, similarities, and complementarities among paradigms from different fields [2]. Consequently, activities promoting the deep learning of students aim to illustrate interconnections and interdependences, highlight dynamics rather than fix structures, and develop skills for critically assessing concepts such as Equity [2].

Following this approach, we present an educational activity that links economics with social values. Furthermore, this is a pioneering initiative to teach social values to first-year Business Administration (BA) and Tourism students at the University of Valencia (Spain). It was implemented through a transversal-skills module, just after the introduction of the "Ethical project: training ethical professionals", designed as a progressive teaching process throughout a degree. In this way, ethics is taken into consideration as a transversal competency in the student's curriculum. Specifically, our proposal revolves around the values of trust and reciprocity. In a nutshell, trust and reciprocity are two sides of the same coin: trust grows (declines) over time as a result of the other's choice (not) to reciprocate cooperation [3]. They form the core of human relationships, essential for improving living standards and may also act as seeds of sustainability for the socio-economic system [4]. For this reason, the concepts of trust and reciprocity require special attention from social science scholars as key values for a sustainable economic model.

Trust may be understood as one's expectations about the goodwill of others to meet their commitments and to not cause harm to others [5]. As a mental process, trust is based on the other's reputation, honesty, morality, and current and future circumstances to anticipate their actions [6]. Thus, this type of cognitive trust is configured with information, repeated interactions, past experiences, and history of cooperation with others [7]. As a voluntary decision, one may learn to trust others, unlike irrational trust that determines involuntary decisions [8]. Particularly in economics, research on social behavior focuses on the cognitive component of trust [9].

Reciprocity may be loosely defined as an obligation to respond in kind to what has been received, including gift exchanges within marriage or kinship relations. It is a broad concept that involves the exchange of goods or services, whether immediate or deferred, for psychological (social relations) or economic reasons (availability of resources) [10]. Thus, reciprocity governs the relations between di fferent generations and statuses [11] and is regulated by the recognition of others and the concern for satisfying their needs for existence [12]. It is also understood as an intrinsic human behavior [13,14] as an instrument for maximizing economic and social advantage.

Experimental economics provides a large body of research on *social preferences,* such as fairness, trust, reciprocity [15] and other concepts, related to people's concerns about the well-being of others. Following a methodology based on laboratory experiments makes it a versatile tool to be used when implementing active learning strategies in social sciences. Compared to the traditional lecture method, the e ffectiveness of classroom experiments on student learning is widely studied with mixed findings. Some works do not find significant di fferences between experiment-based teaching and lecture-only teaching on student performance [16]. In contrast, other papers show such positive e ffects on student learning, performance, and attitude [17–20], that they encourage teachers to incorporate experiments into their teaching even if some extra work is necessary [17].

In this paper, we describe an educational activity that is an attempt to help introduce the teaching of social values as a transversal issue in social science faculties from the very beginning of a degree. From this starting point, a multi-task activity is integrated into the transversal-skill module of the first academic year, which is named *Incorporation to University Study*. It is designed with the goal of motivating students to reflect on social values, specifically trust and reciprocity, as necessary pillars for developing sustainable social and economic systems. To this end, students are encouraged to read and write on humanistic economics [21] to train autonomous and critical thinking and formal writing. Trust and reciprocity values are worked by using experimental games: students are involved in a trust game where they make decisions affecting their own and others' results. Reflection and discussion have a different tempo. Students spend a long time producing a progressive work of reflection. However, the discussion part, where they share and debate their opinions with others, takes place during one classroom session.

Game Theory o ffers the possibility to analyze a decision making context under the hypothesis of the rationality of players. The research question is focused on students' decisions in the trust game, a game created to measure trust in economic decisions. Rational players in this game should not trust nor reciprocate the partner. Taking this theoretical equilibrium as a reference point, we propose a finitely repeated trust game to study the educational learning of students in this context. Specifically, we designed an experiment that includes two treatments. The first is a control treatment: Students are paired to make individual sharing decisions in a trust game played several times with a di fferent partner each time, and they are only informed of their own earnings, which depend on both players' decisions. In this way, their current partner's previous decisions should not a ffect the player's current decision. In the second treatment, however, players are informed of their partner's accumulated earnings before making a new decision. Allowing them to know their own and their partner's accumulated earnings may have an e ffect on their future sharing decisions. This gap may be interpreted as "endogenous income inequality", such that the following research question makes

sense: *Does the information about the inequality in accumulated earnings have an e*ff*ect on students' observed trust and reciprocity?*

## **2. Materials and Methods**

#### *2.1. A Modified Trust Game*

To create trust and reciprocity dynamics in the classroom, we ran a trust game experiment. Our experiment closely replicated one of the treatments in Berg, Dickhaut, and McCabe [22]. Specifically, both a sender and a receiver are endowed with the same amount of money *E*. The sender (trustor) decides which part *x* ∈ (0, *E*) of the endowment to send to an anonymous receiver (trustee). The amount *x* is then multiplied by *n* = 3 in the receiver's hands. In our MTG, the receiver then decides which amount *y* ∈ (0, E + 3*x*) to return to the sender. Consequently, the final pay-off for the sender is π*s* = *E* − *x* + *y*, and that of the receiver equals π*r* = *E* + 3*x* − *y*. Figure 1 shows the extensive form of our modified trust game (MTG).

**Figure 1.** Extensive form of the one-shot modified trust game (MTG).

The MTG has a unique subgame perfect Nash equilibrium in ('*no trust', 'no transfer'*) and therefore, neither trust nor reciprocity is a possible result under the assumption of rational *homo economicus*. However, in lab experiments, results that are not rational emerge involving the exchange of money between parties. It is observed that individuals were willing to share their money to improve their partner's outcome. For instance, in Berg, Dickhaut, and McCabe [22], the average amount sent by the sender was 51.6% of the maximum amount, and the average payback from the receiver w 15.53% of the maximum amount. Definitively, the role played by the sender is crucial, since the value creation (VC) eventually depends on his decision: *VC* = (*n* − <sup>1</sup>)*x* and therefore, on his belief about the receiver's reciprocal behavior. There is also an opportunity for both parties to enhance existing wealth whenever the receiver transfers back the amount received plus an extra (even) small amount ε: *y* = *x* + ε. In such a case, the sender holds or increases his own wealth (since π*s* = *E* + ε) and, in turn, also the receiver's (<sup>π</sup>*r* = *E* + 2*x* − ε). The maximal value creation is reached when the sender sends all his endowment to the receiver: *VCmax* = (*n* − <sup>1</sup>)*<sup>E</sup>*. In such a case, from a social viewpoint, the sender trusts the receiver the most when sharing all his endowment with the receiver. The receiver reciprocates to the sender when transferring back all the amount received from the sender plus an extra reward sufficient to maintain the egalitarian condition.
