University students, economics education, and self-interest. A systematic literature review

International Review of Economics Education 43 (2023) 100266 Available online 5 April 2023 1477-3880/© 2023 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). University students, economics education, and self-interest. A systematic literature review Cristina Miragaya-Casillas a , Raimundo Aguayo-Estremera b , Alberto Ruiz-Villaverde c , * a Department of Statistics and Operations Research, University of Granada, Campus Universitario de Melilla, Calle Santander, Melilla, Spain b Department of Psychobiology and Methodology of Behavioral Sciences,Complutense University of Madrid, Av. S´ eneca, Madrid, Spain c Department of Applied Economics, University of Granada, Campus Universitario de Cartuja, Granada, Spain ARTICLE INFO Keywords: Self-interest Systematic review Economic education University students ABSTRACT There is an open debate in academia about whether economics students behave in a more selfinterested manner than non-economics students. This debate is based on the assumption that economics students are exposed to the study of standard economic models. These models begin with a representative agent, the homo œconomicus, which is a rational optimizer that serves to satisfy their own self-interest. A systematic review was conducted to enhance this debate. Empirical studies that tested the existence of behavioral and/or attitudinal differences related to self-interest among university economics and non-economics students were included. The results provide evidence that economics students are more self-interested than non-economics students. This review has allowed us to highlight the limitations of the pre-existing scientific papers published to date. Primarily, the majority of studies have used cross-sectional data, and it is advisable to carry out more studies with longitudinal data. 1. Introduction In standard economics, the elaboration of models for economic analysis begins with a representative agent, the homo œconomicus, who is a rational optimizer that serves to satisfy their own self-interest. In economics, rationality is understood as the ability to properly analyze and process the information available. 1 Additionally, this agent is capable of optimizing decisions, combined with the ability to order preferences in a completely consistent manner (Klimczak, 2018; Rosengart et al., 2020; Urbina and Ruiz-Villaverde, 2019). Therefore, many researchers and social scientists have wondered whether exposure to the study of these types of economic models, based on the rational pursuit of self-interest, promotes a similar type of related behavior in students (Elegido, 2009; Ifcher and Zarghamee, 2018; Kirchg¨ assner 2005, Lanteri, 2008; Etzioni, 2015). In other words, students are often trained in economic reasoning, which cultivates and reinforces beliefs about the prevalence and power of rational self-interest (Gandal et al., 2005; Molinsky et al., 2012). Leavitt (1989, p. 39) went so far as to argue that business education, with its narrow focus on economics, creates “critters with * Corresponding author. E-mail address: [email protected] (A. Ruiz-Villaverde). 1 In intermediate microeconomics, (see for instance Chapter 3 Consumer Behavior of Pindyck and Rubinfeld, 2013; p.69), the assumption of operationalized is concretized through the utility function. This function expresses consumer preferences. Hence, a series of choice axioms are established: (1) preferences are complete; (2) preferences are transitive; (3) preferences are monotonous (no satiability); and (4) diminishing marginal rate of substitution (strict convexity). Contents lists available at ScienceDirect International Review of Economics Education journal homepage: www.elsevier.com/locate/iree https://doi.org/10.1016/j.iree.2023.100266 Received 1 April 2022; Received in revised form 28 February 2023; Accepted 3 April 2023 International Review of Economics Education 43 (2023) 100266 2 lopsided brains, icy hearts, and shrunken souls”. Attempts have been made to empirically test whether “Econ students” 2 deceive more, are less cooperative, more selfish, less honest, less altruistic, etc. Thus, “deviations from pure selfishness have been interpreted by suggesting that subjects are better people (i.e., more altruistic, or fair), but maybe they are just better economists” (Engelmann and Strobel, 2004; p. 868). Sawyer (1966) pioneered a study using surveys and found significant differences between business students and those from other degrees. Business students were more concerned about maximizing their well-being. In another study, Marwell and Ames (1981) were among the first to test the “free rider” hypothesis with public goods experiments under different conditions on economics students and participants from various subpopulations. They found that economics graduate students were much more likely to free-ride (i.e., contribute less) than any other group of subjects. Meier and Frey (2004) recorded the behavior of students in relation to the donations they made in real life to two social funds managed by the University of Zurich. They found no behavioral differences among economics students, but they did find that business students donated significantly less than other students. However, research findings on this topic are mixed. The debate presents a puzzle as to whether subjects only behave completely self-interestedly under certain circumstances, such as in competitive experimental markets with standardized goods, or in the final rounds of public goods experiments (Fehr et al., 2006). There are also studies that have found no differences in behavior between Econ and non-Econ students (e.g., Aarstad et al., 2011; Ahmed, 2008; Bauman & Rose, 2011; Brosig et al., 2010; Carter & Irons, 1991; Claiborne, 2002), and there are even studies that, contrary to expectations, have obtained evidence that Econ students are more willing to cooperate than non-Econ students (e.g., Hu and Liu, 2003; Yezer et al. 1996). Finally, other studies have observed that non-Econ students are more or equally selfish than Econ students (e.g., James et al., 2001). In this study, we aimed to present the results of a systematic review of the academic literature on this topic: Are Econ students more self-interested than non-Econ students? Additionally, this study allowed us to answer other research questions, such as how is “selfinterest” understood by researchers? What were the main methodological aspects of the reviewed studies? What were the main findings? This study has helped us refine the theoretical framework, as well as highlight the limitations of the published studies to date in order to guide future research. This paper examines the theoretical approaches, followed by the method used to carry out this systematic review, in accordance with the Preferred Reporting Items for Systematic Reviews (PRISMA) statement. 3 The results for the research questions are then presented. Finally, there is a discussion regarding the results obtained, and the study concludes by detailing recommendations for future research. 2. Theoretical approaches In order to understand the power that self-interest currently exercises in human action or motivation, one must go back to the ideological transformation of Europe in the 17th and 18th centuries from which capitalism emerged. According to Hirschman (1977), it was at this time that it was decided that taking advantage of passions would be more constructive than simply repressing them. Passions could be disciplined and transformed into constructive factors for the service of the common good. The search for self-interest began to be appreciated for its propensity to unite people as part of a social fabric based on mutual needs and exchange. The result was to be a smoothly functioning society populated by human actors seeking the “calm desire of wealth.” In the 18th century, the doctrine of self-interest was well established. Despite this, in the field of economic ideas, Adam Smith articulated his vision in his magnum opus, The Wealth of Nations (Smith, 1994). The doctrine of self-interest was then called “the invisible hand” and justifiably became the most influential economic metaphor. 4 In this way, in the study of economics, it is assumed—most often without criticism—that the pursuit of self-interest by economic actors produces the best social result, maximum national wealth. This paradox is possible because of the market’s competitive power (Urbina and Ruiz-Villaverde, 2019). Later, modern economic science focused on studying how individuals maximize their utility. According to Simon (1993), two important limitations arise from this: on the one hand, utility in neoclassical economics is restricted to desire for economic gain, which is questionable because it is not the only motive that guides human action; on the other hand, with the appropriate utility function, a person whose utility derived from giving to other people could—selfishly—give away “millions of dollars.” This invokes an open discussion of the definitions of self-interest and altruism. An individual is self-interested when they think, decide, and act according to their own satisfaction; however, they may care about the well-being of others to the extent that it affects 2 In general terms, we use the term “Econ students” to refer to university students who, during their studies, take courses such as Economics, Microeconomics, Macroeconomics, etc. They do not necessarily have to be students of Economics; they can be students of Business Administration, Accounting, Finance, Marketing, etc. 3 The Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) statement, published by Moher et al. (2009), was designed to help systematic reviewers transparently report why the review was done, what the authors did, and what they found. 4 The “invisible hand” is a concept that was first introduced by Adam Smith in The Theory of Moral Sentiments (1759). It is well known that, paradoxically, Adam Smith only uses the metaphor once in Book IV, Chapter II, paragraph IX of The Wealth of Nations. Following Tribe (2008), it is important to mention that “Das Adam Smith Problem” as the name given to an argument that arose among German scholars during the second half of the nineteenth century concerning the compatibility of the conceptions of human nature advanced in, respectively, Adam Smith’s Theory of Moral Sentiments (1759) and his Wealth of Nations (1776). Very succinctly, the invisible hand explains the rejection of a centralized direction of the economy, hence the preference for a market economy. The principle of sympathy (the ability to share the feelings of others) constitutes, in the Smithian structure, an essential prerequisite to ensure the pursuit of self-interest among a multitude of competing economic agents. C. Miragaya-Casillas et al. International Review of Economics Education 43 (2023) 100266 3 their own well-being (Urbina and Ruiz-Villaverde, 2019). More recently, from the perspective of behavioral economics, the study of social or other-regarding preferences has addressed the discussion of self-interested versus altruistic motivations. Rabin (1998) pointed out the existence of important literature in the fields of psychology, labor relations, and economics on the importance of equity, fairness, status seeking, and other deviations of self-interest in workers’ behavior. For this reason, behavioral economists have dedicated a great deal of effort, in the field of experimental research (through behavioral games), to interpreting the decisions of individuals (players in a laboratory) in relation to motivations based on altruism, reciprocal altruism, justice, etc. To a large extent the experiments have been devoted to testing the behavioral models developed by Fehr and Schmidt (1999), Bolton and Ockenfels (2000) and Charness and Rabin (2002). Within this framework of study, evidence shows that preferences that value others’ decisions depend on the intentions, behavior, and motivation of said others. Individuals cooperate with those who cooperate and punish those who do not cooperate, even if this implies costs in terms of material well-being. The study of self-interested behavior has also been approached from an axiological perspective by social psychologists. Shalom Schwartz’s theory of the universal content and structure of human values has become a reference for this type of study (Schwartz, 1992). In his value model, Shalom Schwartz defines values as goals and motivations which serve as guiding principles in people’s lives. First, he divides values into two categories according to whether they serve individual interests (e.g., power, achievement, hedonism) or collective interests (e.g., benevolence). Second, the goals and interests served by values can be either compatible or conflicting. Thus, two dimensions are proposed. The first dimension is called self-enhancement, in which individuals are motivated to follow their own intellectual and emotional interests (self-direction, stimulation, and hedonism values). The second dimension, self-transcendence, is defined as a preference for the status quo and the certainty provided by close relationships with other people, institutions, and traditions (tradition, conformity, and security values). Self-transcendence refers to the degree to which people are motivated to transcend selfish concerns and promote the well-being of others (including values of benevolence and universalism). Self-enhancement comprises values that motivate people to enhance their own personal interests, even at the expense of others (Schwartz, 1992). Other branches of psychology, consider self-interest to be one of the most important traits for understanding the behavior of individuals. However, to understand this construct, the variable is approached in a broader sense, and the behavior of individuals is related to selfishness and individualism. In a recent study, Gerbasi and Prentice (2013) defined self-interest as “the pursuit of gains in socially valued domains, including material goods, social status, recognition, academic or occupational achievement, and happiness”. The main explanation for people behaving in a self-interested manner starts with the assumption that other people will also act according to their own interest, and they can therefore justify their behavior for this reason (Holmes et al., 2002; Ratner and Miller, 2001; Miller and Ratner, 1998; Miller, 1999). Other reasons, such as altruism, conformity, or reciprocity, are taken seriously as long as they explain further variation in behavior beyond self-interest (Gerbasi and Prentice, 2013). When individuals are confronted by behaviors that appear to violate the unique central motive of self-interest, researchers are more likely to broaden the scope of this motive by considering other reasons, such as those mentioned above, than to question its influence on behavior and thus the non-centrality and uniqueness of self-interest (e.g., Cialdini, 1991; Omoto and Snyder, 1995). 3. Methods A systematic literature review has been carried out according to the recommendations of the PRISMA statement (Moher et al., 2009). Empirical studies which have tested behavioral and/or attitudinal differences related to self-interest among Econ and non-Econ students were included. 3.1. Search strategy The search was concluded in October 2022. Databases used in this review were Web Of Science, ProQuest and Scopus without restriction on search operators. Firstly, the search string was: ((‘Economic*’ OR ‘Business’ OR ‘Accounting’ OR ‘Finance’ OR ‘Marketing’) AND (‘student*’ OR ‘economic education’) AND (‘self-interest’ OR ‘selfish’ OR ‘selfishness’ OR ‘selfishly’)). According to previous studies found in this search (Aarstad et al., 2011; Beekun et al., 2017; Gerlach, 2017; Jamil et al., 2019; Meier and Frey, 2004; Stanley and Tran, 1998; Wang et al., 2011), a possible relationship exists between the constructs of self-interest and selfishness. Therefore, a second search was conducted including papers which measured the construct of selfishness and its derivatives. The previous search equation was repeated by adding the terms: ‘egoism,’ ‘egoistic behavi*’ and ‘egotism.’ All papers had to meet the same inclusion criteria. 3.2. Eligibility criteria This systematic review included empirical studies which met the following criteria: (a) must measure some behavioral and/or attitudinal variable based, motivated, or related to self-interest; (b) had a sample of Econ students; (c) included a comparison between Econ and non-Econ students; and (d) were published in Spanish or English. Conversely, studies were excluded for the following reasons: (a) did not measure any behavioral and/or attitudinal variable based, motivated, or related to self-interest; (b) did not include Econ students; (c) did not include a comparison with non-Econ students; and (d) did not analyze empirical data. C. Miragaya-Casillas et al. International Review of Economics Education 43 (2023) 100266 4 3.3. Selection process Papers were selected in four phases: (1) the title and abstract were read to identify records related to the research. Duplicate or full text records that were not available were excluded; (2) all complete texts were read with the exclusion criteria in mind; (3) a critical reading of the studies was carried out; and (4) a descending search of the papers references selected in previous phases was performed. Two team members independently performed the search, selection, critical reading, and extraction of information. The reliability of the information extraction was evaluated using the Kappa index. For each registered variable (see below), the degree of agreement between members was 100 %. 3.4. Data collection process A coding manual (available upon request) was used to record all variables and group them into four categories according to our research questions (see the Introduction), namely: (1) Conceptual aspects: conceptualization of the construct to study. (2) Construct and operational aspects: outcomes of the construct to study and measurement instrument. (3) Methodological aspects: data collection (transversal vs. longitudinal); sample size; variables of the characteristics of the sample such as types of degree; age and sex. Finally, (4) Conclusive aspects: statistical results of the degree to which the outcomes were studied (descriptive data, statistical tests, regression analysis, etc.); narrative conclusions and presence of theoretical framework: indoctrination vs. selection. 4. Results A total of 697 studies were retrieved from the databases: 202 studies from Web of Science; 261 studies from Scopus; and 234 studies from ProQuest. After a strict application of the search criteria, a final sample of k =29 was obtained. Fig. 1 shows the complete flow of empirical studies (Page et al., 2021). The results are presented by group in the aforementioned categories according to our research questions. Fig. 1. Flowchart of identified, excluded, and included studies. C. Miragaya-Casillas et al. International Review of Economics Education 43 (2023) 100266 5 4.1. Conceptual aspects. How did researchers understand “Self-interest”? When it comes to identifying the construct, it is worth mentioning that some papers did not define any of the terms in their research. However, the majority used the term “self-interest.” Some of these papers additionally also used other terms related to self-interest, such as “selfish,” “selfishness,” “selfishly”; and only a minority (3 papers) use other words related lexically to the above without mentioning “self-interest” (Dzionek-Kozłowska and Rehman, 2017; James et al. 2001; Sawyer, 1966). In addition, some papers used the concept of “self-interest” (and its derivatives) and the concept of “egoism” (and its derivatives) indistinctly. For this reason, the way in which the authors defined the object of study construct, self-interest, was reviewed to verify a possible relationship between these two concepts: self-interest and egoism. Aarstad et al. (2011) closely related self-interest to relativism and egoism because they believed that both motivated students to pursue goals focused on personal gain. According to Beekun et al. (2017) egoism would be a morally correct act if it promoted an individual’s long-term interest. Gerlach (2017, p.2) distinguished three theoretical mechanisms to explain egoism: (i) economics students were less concerned with fairness when making allocation decisions; (ii) they were equally concerned with fairness but had a different notion of what is fair when it came to allocations; and (iii) they expected others to behave more selfishly and, therefore, felt less obliged to behave fairly themselves. Jamil et al. (2019, p.98) believed that egoism was viewed as an individual action motivated by self-interest; and, therefore, egoistic individuals tended to act in a more self-interested manner and maximized their personal gains. Thus, the authors related self-interest to egoism when personal gains or benefits were central to Econ students’ decision-making. 4.2. Construct and operational aspects. How had authors related self-interest to other outcomes? All papers selected in the review tested behavioral differences related to self-interest between Econ and non-Econ students. However, some authors have used different outcomes when conceptualizing and operationalizing self-interest (Table 1). The explanation given by different authors on the choice of these outcomes was as follows. 4.3. Corruptibility and greed Frank and Schulze (2000) did not study self-interest in isolation but studied the degree to which self-interest dominated interest in others. The authors considered the outcome corruptibility met these requirements because it conflicted with generally accepted moral norms. Stanley and Tran (1998) described the theoretical-conceptual framework around the term self-interest. When presenting the results of their study, they introduced the term greed in order to make a comparison with a similar study. Thus, the constructs—corruptibility and greed—were outcomes used to test the self-interested behavior of university students. Table 1 Summary of construct and operationalization aspects. Study Construct Operationalization variable Aarstad et al. (2011) Ethical values Fairness, relativism, and egoism Ahmed (2008) Pro-social Cooperation Bauman and Rose (2011) Pro-social Giving behavior Beekun et al. (2017) Egoism Egoism, utilitarianism, relativism, justice, and ethical decision Blossiers and Soto (2008) Egoism Egoism Cappelen et al. (2015) Pro-social Pro-social Carter and Irons (1991) Self-interest Amount acceptable and amount kept Cox (1998) Cooperation Voting behavior, tipping behavior and investment behavior Dzionek-Kozłowska and Rehman (2017) Cooperation Cooperation Espín et al. (2022) Self-interest Envy, compassion, and unemployment Frank et al. (1993) Cooperation Cooperation and honesty Frank and Schulze (2000) Corruptibility Amount acceptable and amount kept Frey and Meier (2003, 2005) Pro-social Giving behavior Gandal et al. (2005) Self-interest Power and achievement Gerlach (2017) Self-interest Fairness Hole (2013) Self-interest Fairness Hu and Liu (2003) Altruism Cooperation Ifcher and Zarghamee (2018) Self-interest Self-interest James et al. (2001) Cooperation Cooperation Jamil et al. (2019) Egoism Egoism Lanteri and Rizzello (2014) Cooperation Cooperation Meier and Frey (2004) Pro-social Giving behavior Mertins and Warning (2014) Reciprocity Fairness Petersen and Ford (2019) Ethical values Personal values Sawyer (1966) Altruism Altruism Selten and Ockenfels (1998) Altruism Giving behavior Stanley and Tran (1998) Greed Fairness Wang et al. (2011) Greed Greed C. Miragaya-Casillas et al. International Review of Economics Education 43 (2023) 100266 6 4.4. Cooperation, pro-social preferences, altruism, self-enhancement, and reciprocity Most authors used the outcomes cooperation or pro-social preferences to contrast them with self-interested behavior and/or attitude when actions or attitudes promoted for personal gain come into play. Thus, “students who major in business and economics might act less cooperatively when that behavior leads to personal profit” (Cox, 1998, p.69) or emphasis on the self-interest model tends to inhibit cooperation (Dzionek-Kozłowska and Rehman, 2017). Some authors even considered pro-social preferences as a general outcome which included cooperative behavior or attitude (Ahmed, 2008; James et al., 2001; Meier and Frey, 2004). For instance, Ahmed (2008, p. 301) posited that “Economics students differ in their pro-social preferences and can be divided into selfish and cooperative types”. Thus, the common conceptual element which led self-interest to the opposing behaviors or preferences was due to the weight which people perceived the benefit of others to have in relation to their own. This classification for defining self-interest was expanded when we focused on the operationalization of these measurements. That is, the authors used operationalized variables different from those previously defined and new variables appeared, such as, fairness or power. 4.5. Methodological aspects. What were the main methodological aspects of the studies reviewed? When relating the outcomes which were identified with self-interest and the instruments used to assess them, a series of nuances must be noted. Authors measured the same outcomes with different instruments. For example, the variable self-interest was measured by way of three different questionnaires; likewise, it was measured by five different behavioral sets: dictator game, ultimatum game, public goods game, third-party punishment game, and prisoner’s dilemma game. The same instruments were also used to measure other outcomes. Authors used the ultimatum game to measure self-interest, greed, and pro-social behavior. Fig. 2 below shows the different outcomes employed by authors and their respective instruments. The instruments were classified as follows: questionnaires, economic experiments, and university records. In order to verify the main methodological aspects of the studies reviewed, it was necessary to check the following aspects listed in Table 2. Note. * PVQ: Schwartz Portrait Values Questionnaire (Schwartz et al., 2001); **SVS: Schwartz Value Survey (Schwartz, 1992). The sample size of the empirical studies presented in the papers was different depending on whether the design was cross-sectional or longitudinal. In the case of cross-sectional studies, the sample size was not homogeneous, with samples ranging from 16 participants (Stanley and Tran, 1998) to 600 (Espín et al., 2022), the average being 216 participants. In the case of the longitudinal studies, the samples were much larger because they used census samples. Sample sizes were 8743 (Bauman and Rose, 2011), 37,588 participants Fig. 2. Relationship between outcomes and instruments used to measure self-interest. C. Miragaya-Casillas et al. International Review of Economics Education 43 (2023) 100266 7 (Meier and Frey, 2004) and 96,783 (Frey and Meier, 2003, 2005). The article by Ifcher and Zarghamee (2018) had the smallest sample size, with a total of 276 participants in which the authors conducted a laboratory experiment where participants completed 6 experimental tasks before and after treatment. It is worth mentioning that the fourth longitudinal study (Frank et al., 1993) did not report sample size. Age and sex were recorded; however, nevertheless, they do not appear in Table 2 because more than 75.86 % of the papers did not report such information. Regarding the study population, two groups were considered: one group comprised Econ students and the other non-Econ students. The first group falls mainly in two degrees: Economics and Business. The majority (12 papers) used only Economics students and 7 papers only Business students. However, 7 papers used both degrees as a definition for the sample of Econ students. It is noteworthy that 2 papers also used Accounting and Operations Research students and students of Economic Pedagogy or Agricultural Economics as a sample of Econ students. Therefore, the sample as a whole was: 67.86 % of papers used Economics students; 50 % used Business students; 3.57 % used Accounting students; and 3.57 % Economic Pedagogy or Agricultural Economics students. In the case of Ifcher and Zarghamee (2018) students’ degrees were not specified. However, in the laboratory experiment they were asked if they had ever studied game theory or economics in general during any college course. For the second group, non-Econ students, there was greater diversity of degrees. As shown in Table 2, with the exception of 1 paper which also samples Police cadets, all were university students. The methodological aspects described above highlight the heterogeneity of studies on the subject. In addition, in order to study the quality of the evidence on this topic, the risk of bias of the selected papers in this review was analyzed (see Table 3). Selection, performance, detection, and attrition biases were considered. The conditions for establishing the risk levels of each bias were based on Cochrane guidelines (Higgins et al., 2019) and the methods section of the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE; Von Elm et al. 2014). According to the results, it could be posited that attrition bias is unlikely to affect the studies (high: 0 %, low: 55.17 % and undetermined: 44.83 %). However, most studies showed a high risk of selection bias (high: 72.41 %, low: 17.24 %, and undetermined: 10.34 %). With regard to performance bias (high: 20.69 %, low: 27.59 %, and undetermined: 51.72 %), and detection bias (high: 0 %, low: Table 2 Summary of methodological aspect. Study Data collection (Longitudinal vs. Transversal) Sample Size (N) Field of study (Econ students) (Non-Econ students) Aarstad et al. (2011) Transversal 168 Business Engineering Ahmed (2008) Transversal 90 Economics Humanities and police cadets Bauman and Rose (2011) Longitudinal (Panel, Census) 8743 Economics Arts and Sciences Beekun et al. (2017) Transversal 158 Business Sciences Blossiers and Soto (2008) Transversal 336 Business and Economics Medicine and Sociology Cappelen et al. (2015) Transversal 375 Business and Economics Humanities, Natural Sciences and Social Carter and Irons (1991) Transversal 92 Economics Unspecified Cox (1998) Transversal 301 Business and Economics Nurse and Education Dzionek-Kozłowska and Rehman (2017) Transversal 341 Economics Sociology Espín et al. (2022) Transversal 600 Business and Economics Medicine, Civil Engineering, Law, Anthropology and Psychology Frank et al. (1993)* Transversal 207 Economics Math, Computer Science, Engineering, Natural Sciences, Humanities, Architecture and Music Frank et al. (1993)* Longitudinal — Economics Astronomy Frank and Schulze (2000) Transversal 161 Economics, Economic Pedagogy and Agricultural Economics Unspecified Frey and Meier (2003, 2005) Longitudinal (Panel, Census) 96.783 Business and Economics Unspecified Gandal et al. (2005) Transversal 262 Economics Social Sciences Gerlach (2017) Transversal 165 Economics Arts and Sciences Hole (2013) Transversal 376 Business Engineering Hu and Liu (2003) Transversal 255 Economics Unspecified Ifcher and Zarghamee (2018) Longitudinal 276 Unspecified Unspecified James et al. (2001) Transversal 33 Economics, Business and Accounting Psychology Jamil et al. (2019) Transversal 146 Business Sciences, Social Sciences, Engineering and Computer Sciences Lanteri and Rizzello (2014) Transversal 190 Economics Nursing and Law Meier and Frey (2004) Longitudinal (Panel, Census) 37,588 Business and Economics Unspecified Mertins and Warning (2014). Transversal 359 Business Law Petersen and Ford (2019) Transversal 203 Business Engineering, Arts and Medicine Sawyer (1966) Transversal 122 Business Social Sciences and Social Services Selten and Ockenfels (1998) Transversal 118 Economics Unspecified Stanley and Tran (1998) Transversal 16 Economics Arts Wang et al. (2011) Transversal 112 Economics Education Note. *Frank et al. (1993) present two different studies in the same article. C. Miragaya-Casillas et al. International Review of Economics Education 43 (2023) 100266 8 44.83 %, and undetermined: 55.17 %), more than half of the papers did not report the information, therefore, it could not be concluded whether the risk was high or low. 4.6. Conclusive aspects. What are the main findings? Finally, this question was addressed using the variables included in the conclusive aspects (see Table 4). When considering self-interest, 62.07 % of the reviewed papers showed statistically significant differences between the two groups. Econ students presented more self-interested behavior than non-Econ students. However, 4 of these papers showed a number of nuances. Carter and Irons (1991), by way of the ultimatum game, determined that Econ students acted less pro-socially than non-Econ students. Only in the case of the group of students who proposed the distribution of a sum of money in the ultimatum game, were the results not maintained when selecting senior students. Cox (1998) found significant results for only two of the four outcomes studied. Likewise, Cappelen et al. (2015) found significant differences only in women. Finally, Espín et al. (2022) suggested that the differences were due to self-selection in preferences and indoctrination in students’ beliefs about the behavior of others. Only 2 papers have determined that Econ students are not considered less cooperative than non-Econ students (James et al., 2001, Hu and Liu, 2003). However, Hu and Liu (2003) argued that exposure to the self-interest model does not necessarily alter the extent to which people behave self-interestedly, despite the fact that Econ students were considered more cooperative when negotiating than non-Econ students. The remaining papers (17.86 %) did not show significant differences, but they did find that Econ students were more self-interested. Consequently, 86.21 % of the papers concluded that there were significant differences in the level of self-interested behavior according to the degree studied. 5. Discussion Currently, there is an open debate in academia on whether teaching standard economics promotes self-interested behaviors and/or attitudes (e.g., Ahmed, 2008; Haucap and Mu¨ller, 2014; Meier and Frey, 2004; Mertins and Warning, 2014; Hu and Liu, 2003). This is due, in part, to the fact that students are exposed to study models for economic analysis which start from a specific representative agent, the homo œconomicus, which is characterized by the rational search of their own self-interest (Urbina and Ruiz-Villaverde, Table 3 Risk of bias of the selected papers. Study Bias Selection Performance Detection Attrition Aarstad et al. (2011) High Low Undet. Low Ahmed (2008) High High Undet. Low Bauman and Rose (2011) Low Low Low Low Beekun et al. (2017) High – Low Undet. Blossiers and Soto (2008) High Undet. Low Undet. Cappelen et al. (2015) High Low Low Undet. Carter and Irons (1991) Low Undet. Undet. Low Cox (1998) Low Undet. Undet. Undet. Dzionek-Kozłowska and Rehman (2017) High Undet. Undet. Undet. Espín et al. (2022) High Undet. Low Undet. Frank et al. (1993)* Undet. High Undet. Low Frank et al. (1993)* Undet. Undet. Undet. Undet. Frank and Schulze (2000) High Undet. Undet. Low Frey and Meier (2003, 2005) Low Low Low Low Gandal et al. (2005) High Undet. Undet. Undet. Gerlach (2017) High High Low Low Hole (2013) High High Undet. Undet. Hu and Liu (2003) Low High Undet. Undet. Ifcher and Zarghamee (2018) Low Low Low Low James et al. (2001) High High Undet. Undet. Jamil et al. (2019) High Undet. Low Low Lanteri and Rizzello (2014) High Undet. Undet. Undet. Meier and Frey (2004) Low Low Low Low Mertins and Warning (2014). High Undet. Undet. Low Petersen and Ford (2019) High Undet. Low Low Selten and Ockenfels (1998) High Low Low Undet. Sawyer (1966) High Undet. Undet. Low Stanley and Tran (1998) High Undet. Undet. Low Wang et al. (2011) High Low Low Low TOTAL High: 72.41 % Low: 17.24 % Undet.: 10.34 % High: 20.69 % Low: 27.59 % Undet.: 51.72 % High: 0 % Low: 44.83 % Undet.: 55.17 % High: 0 % Low: 55.17 % Undet.: 44.83 % Note. High =High risk of bias; Low =Low risk of bias; Undet. =Undetermined risk of bias due to lack of information. *Frank et al. (1993) present two different studies in the same article. C. Miragaya-Casillas et al. International Review of Economics Education 43 (2023) 100266 9 Table 4 Summary of conclusive aspects. Study Statistical results Conclusions Hypotheses Aarstad et al. (2011) Ordinary least squares (OLS) b=0.220; p <.05; R 2 =.09 Econ students behave less ethically sensitive, on issues related to relativism and egoism, than non-Econ students (controlling for gender) Not tested Ahmed (2008) Chi-Squared test X 2 =17.8; p <.001 (prisoner’s dilemma) X 2 =15.7; p <.001 (stag hunt game) Econ students behave less cooperatively than non-Econ students Indoctrination Bauman and Rose (2011) Linear regression B=0.006; p <.001 (WashPIRG group) B=0.006; p <.001 (ATN group) Econ students behave less pro-social than non-Econ students (controlling for gender, subjects, age, black and nationality) Indoctrination vs. Selfselection Beekun et al. (2017) t-test t=4.05, p <.001 Econ students behave more selfish than non-Econ students Not tested Blossiers and Soto (2008) Descriptives (percentages) Econ students behave more selfish than non-Econ students Economic training Cappelen et al. (2015) DNS p=.07 (men, dictator) p<.05 (women, dictator) p=.40 (men, trust) p<.001 (women, trust) Econ students from the women’s group behave less pro-social than non- Econ students Not tested Carter and Irons (1991) Ordinary least squares (OLS) B= − 0.74; p <.05; R 2 =.05 (total responder group) B=0.71; p <.05; R 2 =.09 (total proposer group) B=1.47; p <.05; R 2 =.10 (senior responder group) B=0.18; p >.05; R 2 =.12 (senior proposer group) Econ students behave less pro-social than non-Econ students. However, in the case of the proposer group, the results are not maintained when only senior students were selected (controlling by course). Indoctrination vs. Selfselection Cox (1998) Ordinary least squares (OLS) B=0.05, p >.05 B=0.12, p >.05 B= − 0.13, p <.05 B=7.59, p <.05 Of the four questions (VDs), they found significant results in only two, in those the Econ students behave less cooperative than non-Econ students (controlling for gender, age and course) Economic training Dzionek-Kozłowska and Rehman (2017) Descriptive statistics (means) No statistically significant differences Indoctrination Espín et al. (2022) Probit regression p>.40 No statistically significant differences Indoctrination vs. Selfselection Frank et al. (1993)* Ordinary least squares (OLS) B=0.17, p <.05, R 2 =.20 Econ students behave less cooperatively than non-Econ students (controlling for game, gender, and course) Economic training Frank et al. (1993)* Descriptive statistics (percentages) Econ students behave less cooperatively than non-Econ students Economic training Frank and Schulze (2000) Probit regression B=0.64, p <.05 Econ students behave more corrupt than non-Econ students (controlling for gender, fix payments and course) Indoctrination vs. Selfselection Frey and Meier (2003, 2005) Probit regression B= − 0.10, p <.05 Econ students behave less pro-social than non-Econ students Indoctrination vs. Selfselection Gandal et al. (2005) MANOVA F=16.02, p <.001 (achievement) F=4.15, p <.05 (power) Econ students attach more importance to achievement and power values than non-Econ students Economic training Gerlach (2017) Tobit regression B= − 2.25; p <.001 Econ students behave more selfish than non-Econ students (controlling for gender) Not tested Hole (2013) Linear regression B= − 0.05, p =.04 Econ students behave more selfish than non-Econ students Self-selection Hu and Liu (2003) Chi-Squared test X 2 =11.39; p <.001 (First round) X 2 =14.31; p <.001 (Second round) Econ students behave more cooperative than non-Econ students (controlling for gender) Not tested Ifcher and Zarghamee (2018) t-test p=.005 Econ students behave more selfish than non-Econ students Self-selection vs. Economic training James et al. (2001) t-test t=1.98, p =.06 (Psychology) Econ students do not behave less cooperative than non-Econ students Not tested Jamil et al. (2019) t-test t=0.34, p >.05 (egoism) No statistically significant differences Economic training Meier and Frey (2004) t-test t=16.20, p <.001 Econ students behave less pro-social than non-Econ students Indoctrination vs. Selfselection (continued on next page) C. Miragaya-Casillas et al.