Basic human values

Basic human values are considered a major element of individual identity and in human life [16]. In Schwartz’s model they are defined as “trans-situational goals, varying in importance, that serve as guiding principles in the life of a person or other social entity.” In 1992, and relying on earlier approaches [17, 18], Schwartz developed a theory specifying the meaning, catalog and measurement of values. The first version of the theory from 1992 proposed 10 values and the revised version of the theory included 19 more specific values [4, 19]. In both versions of the theory there are four higher-order values: self-transcendence, self-enhancement, conservation, and openness to change. Self-enhancement, which reflects achievement or power values, stands in opposition to self-transcendence, which reflects the values of universalism or benevolence. Openness to change, which includes the values of stimulation or self-direction, stands in opposition to conservation values, which include tradition, security, and conformity. The four higher-order basic human values are underlined by different motivations underlying human behavior. People who attribute high importance to self-transcendence values are concerned with the welfare and interests of others. By way of contrast, people attributing high importance to self-enhancement values find it important to pursue their own interests and relative success and dominance over others. Individuals scoring high on openness to change values find autonomy, independence of thought and action or excitement in life important. Conservation stands in opposition to openness to change. People for whom conservation is important emphasize safety, acceptance of customs and restraint of actions or impulses that may violate expectations of others’ order, self-restriction, preservation of the past, and resistance to change [7]. The source of these values is, according to theory, embedded in three universal needs of individuals: needs as biological organisms, need for coordinated social interaction, and survival and welfare needs of groups [4]. The model has been tested and validated in hundreds of samples and studies across different countries and continents [20–25].

The theory also suggests that the values be arranged in a (semi-)circle which reflects relationships between the values. This structure is observed not only among younger and older adults, but even among adolescents or preschool children [26]. When values in the circle are opposite one another, pursuing these values at the same time may create a conflict. When their motivations are similar, values are located closer to each other on the circle and pursuing them simultaneously may be compatible. These values have been shown to be associated with and influence a wide range of attitudes (e.g. toward immigration, political orientation, or identifying as a European, to name but a few (see, inter alia, [27–33]) and behaviors [13, 29, 33–35].

In the current study we focus on the four higher-order values. Schwartz et al. [4] suggested that we may use broader or more specific values, depending on our research purpose. This study focused on the higher-order values, because as explained above they entail the basic motivations of the theory. Nevertheless, we also present the empirical analysis on the lower level of values in the Supporting Information.

Economic preferences

In the current study we are examining potential relations between the four aforementioned higher-order basic human values and the following key and commonly used economic preferences captured in the Global Preference Survey [8]: risk taking, trust, altruism, and negative and positive reciprocity. The survey also contains the preference for patience, but as we did not expect, hypothesize or pre-register any theory-derived relations between patience and any value. Thus, we did not include patience in our analysis.

Risk seeking (or risk aversion) is usually defined based on the expected utility people get from different outcomes that involve uncertainty. For example, a risk-neutral individual would be indifferent about choosing between receiving $100 for certain, or an equal probability of receiving $50 or $150. A risk-averse individual would tend to select the safe option ($100), and a risk-seeking individual might prefer the thrill of choosing the risky option ($50 or $150 with equal probabilities) [36]. The concepts of risk aversion and risk seeking have played a prominent role among economists, because they were able to explain many phenomena, like the risk premium of stocks, variations in consumption and career paths, and potentially even smoking [37]. The modeling of risk in economics has gone through a transition based on findings in behavioral economics, away from orientation towards wealth alone, and towards loss aversion relative to specific reference points [38, 39]. In essence, willingness to take a risk entails a person’s evaluation of a certain outcome relative to an uncertain outcome. Similar choices in more or less similar forms are present in everyday life, for example in career choices, decisions to immigrate, parking, etc. [40, 41], with newer research questioning how stable risk preferences are [10, 42].

Trust [43, 44] is defined in economics as the expectation that when in a vulnerable situation, others won’t take advantage of it [45, 46]. For example, in a trust game, the trustor gives resources to a trustee, which are tripled by the experimenter (to incentivize the behavior). The trustee can then freely decide whether to give something back to the trustor. The expectation that the trustor will give something back is typically the measure of trust [47]. Trust is particularly important in non-perfect market setups, because contracts in such setups can never be 100% complete and enforcement of contracts is costly. Trust is necessary for actors to be able to rely on their counterparts upholding agreements, and that such agreements will not need to be enforced, thus incurring considerable costs. Trust is desirable, because it can thus reduce transaction costs [48, 49]. Trust (in people) is also strongly correlated with risk preferences [13] but clearly distinct from them [50].

Altruism is defined as the costly act of giving something beneficial to others without expecting anything in return [51, 52]. Altruism is often used to reflect caring about others and their outcomes. Traditionally, economists from the neoclassical economic perspective had been puzzled by the behavior in the so-called dictator game, where completely anonymous individuals decided to share money with an unknown second person they would never know. This behavior has traditionally been called altruism in economics (although later tests refine the understanding as to how much this behavior really does measure a trait of altruism [53, 54]. Altruism has been linked to general pro-social behaviors [55, 56] but also environmentally friendly behaviors [57, 58].

Finally, reciprocity is one of the earliest and arguably most important behavioral economic concepts [59, 60]. It is defined as the willingness to reciprocate an action in kind without enforcement by an institution. Reciprocity is closely related to trust. Whereas trust is the belief or expectation that other people will act in a trustworthy way, reciprocity is more dynamic. Positive reciprocity describes the act of returning a favor. By way of contrast, negative reciprocity has been proposed as a primary driver of sustaining costly punishment of norm violators. Such punishment is described as a major driver of cooperation. For example, both positive and negative reciprocity are important factors shaping workers’ behavior towards their employer, either when returning the favor of higher pay with extra effort, or engaging in backlash against lower-than-expected pay [55, 61–63].

Economic preferences are usually not related to each other in studies, that is, each measure is a stand-alone measure often used as an explanatory variable [13, 44]. Therefore, no mapping of the measures similar to that of the values has been proposed so far, to the best of our knowledge.

Mechanisms underlying possible relations between basic human values and economic preferences

Whereas previous literature has examined the relations between economic preferences and psychological predictors of behavior such as the Big Five personality traits or locus of control [1, 64, 65], to the best of our knowledge there are hardly any studies examining the relations between economic preferences and basic human values (however, see [66]). Some studies in business economics and marketing examined possible connections between values and economic preferences (e.g. [67–69]). However, these studies typically focused on specific subsets of values and/or economic preferences in the marketing context rather than on abstract motivations relevant for a wider range of behaviors. Below we describe potential mechanisms that may account for relations between economic preferences and basic human values, and formulate empirically testable hypotheses about these relations based on the proposed mechanisms.

1. Risk seeking (in contrast to risk aversion) represents the willingness of an individual to explore potentially beneficial but risky situations. Conservation values emphasize quite the opposite, namely, the importance of safety, preservation of the past, and resistance to change. Thus, (H1) we expect risk seeking to be negatively associated with conservation.

Openness to change underlines the importance of adventures and excitement in life, autonomy and independence in thoughts and action. Thus, and by way of contrast, (H2) we expect risk seeking to be positively correlated with openness to change.

Trust represents the belief in other people’s fair behavior. It describes the active process of allowing oneself to be vulnerable to other people’s exploitation. Self-transcendence is defined as the importance of being concerned with the needs of other people. People with high levels of self-transcendence prioritize the needs of others, hoping that their helpful actions will not be exploited. Therefore, (H3) we expect a positive association between trust and self-transcendence.

By way of contrast, self-enhancement prioritizes individuals’ own needs over those of the others. People scoring highly on self-enhancement strive for success and power over others, which may come at the cost of the wealth of others. Considering others as competitors is inherent in this value and may induce distrust. Consequently, (H4) we expect a negative association between trust and self-enhancement values.

Both altruism and self-transcendence entail a motivation to support others and give up something for the benefit of others. Thus, (H5) we expect a positive association between altruism and self-transcendence. Since the motivation underlying self-enhancement is precisely the opposite, that is, seeking one’s own benefit at the cost of others’ well-being, (H6) we expect a negative association between the values of altruism and self-enhancement.

Reciprocity is a socially relevant motivation. Not only does it represent the importance of doing something good and promoting the well-being of others–it also helps to maintain a social status quo in a functioning society, providing an element of security that favors will not be forgotten, the latter underlying both the values of self-transcendence and conservation. Therefore, (H7) we expect self-transcendence values and (H8) conservation values to be positively associated with positive reciprocity.

As mentioned earlier, reciprocity is a tool to maintain the social status quo, societal trust, and a functioning society, which are important goals for people who score highly on conservation. However, whereas self-transcendent individuals may be happy to express positive reciprocity, they may be hesitant to express negative reciprocity, as it would go against their motivation to do good. Therefore, (H9) we expect conservation (but not self-transcendence) to be positively associated with negative reciprocity.

Negative reciprocity could also be considered an expression of power and control over others. It entails an element of punishing or retaliating against improper or undesired behavior. (H10) We thus expect self-enhancement values to be positively associated with negative reciprocity. Table 1 summarizes our hypotheses (S1 Table summarizes them on the lower values level). We preregistered a variation of our expectations (though more detailed on the specific values rather than the higher order value level) and the procedure at the American Economic Association’s registry for randomized controlled trials https://www.socialscienceregistry.org/trials/6483) under Scharfbillig [70]. In the next section we will present the methodology underlying our hypotheses, and the empirical examination thereof.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Summary of hypotheses.

https://doi.org/10.1371/journal.pone.0296852.t001

Participants

We collected data using convenience samples in two countries, Poland and Germany: a non-student sample of 212 from Poland (123 females, 89 males, mean age 34.0 years, all of Polish nationality) and a student sample of 127 participants from the University of Cologne, Germany (90 females, 36 males, 1 other, mean age 21.2 years, 120 of German nationality and 8 of a different nationality). See full sample descriptions by country in the Supporting Information, S2, S3 Tables. Convenience samples were sufficient for our purpose, because we were not interested in descriptive evidence but in associations, for which we did not necessarily need representative or particularly diverse samples. Several studies have shown that associations between behaviors or attitudes are similar even when comparing them across convenience samples (including student ones) and representative samples [71]. The sample participants were recruited using the snowball technique, and participation was not remunerated. In order to partially compensate for the limitations of the use of a convenience sample, we examined our hypotheses in the two samples separately to gain external validity for the results. In addition, we controlled for sociodemographic characteristics in our regressions to account for potential compositional effects. Ethical approval was obtained from the Joint Ethics Committee of the University of Frankfurt and the University of Mainz, Germany (“Connecting Basic Human Values and Behavioural Economics”). Written informed consent was obtained.

Survey design and measurements

To ensure that the approaches were as comparable as possible, the same methodology was used for the data collection in the two samples and for both basic human values and economic preferences. As we are using the method of the global preference survey, several of the economic preferences are–in contrast to economic tradition–asked about, rather than derived from a revealed choice. However, the survey still contains several hypothetical choices and the final preferences are calculated based on them. Values are typically measured using self-rated survey items, so we chose this methodology–also for assessing the economic preferences. The most notable way to measure economic preferences via a survey comes from the experimentally validated preference survey module [8]. Participants in both countries answered an online survey self-administered in German or Polish, respectively. The survey was administered using the online survey tool, LimeSurvey.

Correlations

All the analyses were performed using Stata version 18. Since the findings in the multivariate analyses were similar in the two countries, we are reporting the results for the total sample beginning with the descriptives. We performed a measurement invariance test for the values and find full metric invariance and partial scalar invariance, thus confirming that the values are comparable between the two countries, see S8 Table. We further performed a MANOVA test for sample differences in the main dependent variables, economic preferences and values (λ (10,288) = 0.76, p<0.01). We found that whereas there were significant group differences in the human values (λ (4,325) = 0.77, p<0.01), economic preferences were rather similar (λ (6,293) = 0.99, p = 0.99). However, the overall proportion of the variance attributable to the group difference is only about a quarter. In the pooled analysis we controlled for nationality. S3, S4 Tables provide both the descriptives and the results of the multivariate analysis for the two samples separately. Table 2 presents the means and standard deviations of the higher order values and the economic preferences for a pooled sample of the two countries.

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Descriptive statistics.

https://doi.org/10.1371/journal.pone.0296852.t002

S1 Fig in the Supporting Information shows the correlations between values and economic preferences overall and per country. We see clear correlations profiles which seem to be mostly consistent in each of the two countries. The Pearson correlations between economic preferences and values were weak or moderate [74]. S5 Table presents the Pearson correlations between the four higher order values and the six economic preferences. In line with H1 and H2, openness to change was significantly positively correlated (r = 0.23, p < 0.000) and conservation significantly negatively correlated (r = -0.24, p < 0.000) to risk seeking. In other words, individuals attributing higher importance to the value of openness to change also exhibited a stronger risk seeking preference, while individuals attributing higher importance to conservation values exhibited a weaker risk preference.

In line with H3 and H4, trust was correlated positively to self-transcendence (r = 0.16, p = 0.01) and negatively to self-enhancement (r = -0.25, p < 0.00). It was also correlated positively to conservation values (r = 0.14, p = 0.01) and negatively to openness to change values (r = -0.11, p = 0.06), for which we had not proposed any hypotheses.

The associations of higher order values with altruism provided empirical support for H5 and H6. Self-enhancement was negatively correlated to altruism (r = -0.28, p < 0.00), and self-transcendence positively (r = 0.32, p < 0.00). In other words, individuals attributing higher importance to self-enhancement values were less likely to display altruism, whereas those attributing higher importance to self-transcendence values were more likely to display altruistic preferences. Altruism also displayed a negative correlation to openness to change values (r = -0.19, p < 0.00) and a positive correlation to conservation values (r = 0.10, p = 0.06), for which we had not proposed any hypotheses.

The correlation between self-transcendence and positive reciprocity was positive (r = 0.16, p < 0.00) but its correlation with conservation was insignificant (r = 0.04, p = 0.46). Thus, empirical support was provided for H7 but not for H8.

Finally, negative reciprocity displayed, as expected in H9, a positive association with self-enhancement values (r = 0.26, p < 0.00). However, and in contrast to H10, it was negatively correlated with conservation values (r = -0.18, p = 0.00).

In order to correct for multiple hypothesis testing, the conservative Bonferroni approach would require adjustment of the significance level to 0.005 (0.05/10). With this correction, out of the above 8 supported hypotheses, only H4 had to be rejected–and even then only just (p = 0.0056). S1 Text also shows the analysis for correlations per country separately.

Multivariate analysis

Next, we delved deeper into the relationship between values and economic preferences. Correlations may be subject to multiple omitted variables bias. One way to address this is to use a regression analysis with all theoretically relevant values as predictors simultaneously, while controlling for sociodemographic characteristics. We avoided including higher-order values in the regressions that were not hypothesized to be associated with the respective economic preference, as this would likely have caused a multicollinearity problem. The reason for this is that values are theorized to span over a two-dimensional circular space [51]. All values that were hypothesized to be related to a specific economic preference were included in the regression. Furthermore, although we used economic preferences as dependent variables and values as predictors, we did not assume any causal relationship between them, but rather merely associations. While there were some differences in effects sizes in the reversed regressions, none of the associations was strong, lending further support to the result suggesting that the values and the economic preferences do not show considerable overlap. Table 3 presents the regression coefficients of these models (S7 Table provides the result for lower level values, S2 Fig show the regression lines in a scatter plot).

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. Standardized OLS regression coefficients.

https://doi.org/10.1371/journal.pone.0296852.t003

Supporting H2, risk seeking displayed a negative and significant association with conservation. Openness to change was no longer significant, but according to the motivational trade-offs between openness to change and conservation, it is unlikely that both values would be simultaneously significant in the same regression as they are located on opposite sides of the value circle. The second regression for trust provided support for H4. The regression showed a negative and significant association between trust and self-enhancement. Also in this case, the lack of positive and significant correlation with self-transcendence was likely a result of the trade-off between self-enhancement and self-transcendence, which are located in opposition to each other on the value circle. In line with H5, altruism turned out to be positively and significantly associated with self-transcendence, but not with the opposing value of self-enhancement.

Positive reciprocity did not display any significant association with self-transcendence or conservation when included together in the model. However, negative reciprocity and self-enhancement (but not conservation) showed a positive and significant association, lending support for H10.