Do Sacrificial Moral Dilemmas Measure True
Utilitarianism?
Izabella Engelmoera,1 aUniversity of Amsterdam
This manuscript was compiled on June 2, 2017
Research into the psychological and neural basis of moral decision-making generally assumes that hypothetical sacrificial moral dilem-mas are a valid measure of utilitarian decision-making. However, it re-mains unclear whether utilitarian decisions in these dilemmas trans-late to a genuine utilitarian impartial concern for the greater good. Therefore, we conducted two studies to assess whether sacrificial moral dilemmas are a valid measure of utilitarian decision-making. For this purpose, participants were presented with several moral dilemmas in which they could choose whether or not to endorse a utilitarian outcome, and were asked to make decisions in a novel economic game. Our results show that utilitarian decision-making on moral dilemmas was associated with psychopathic traits. Sec-ond, we found no association between utilitarian decision-making on moral dilemmas and real life utilitarian behavior (making a utilitarian choice in the economic game). Taken together, these results suggest that there is reason to doubt whether sacrificial moral dilemmas that dominate current research are a valid measure of a true utilitarian moral viewpoint.
Moral dilemma | Utilitarianism | Psychopathy | Dictator Game
F
or centuries, human morality has been an important topic of interest in philosophy. More recently, however, human morality has also been studied in psychology and neuroscience (Sun, 2012; Tomassello & Vaish, 2013). The scientific interest in human morality is not surprising because the capacity to make moral judgements is commonly viewed as an essential condition for the functioning of modern societies and as a fundamental part of the human mind (Tomasello & Vaish, 2013; Haidt, 2001; Mikhail, 2007). Research into the psychological and neural basis of moral decision-making generally assumes that sacrificial moral dilemmas, in which participants have to decide between two morally opposing options (i.e., a utilitarian and a deontological option), are a valid measure of moral making and, more specifically, utilitarian decision-making (Kahane et al., 2015; Christensen & Gomila, 2012). In this study, we aimed to assess whether this crucial assumption is valid.In sacrificial moral dilemmas, participants are presented with a hypothetical scenario in which they must decide whether to sacrifice one person’s life in order to save more persons’ lives (Kahane et al., 2015). If they do not endorse harming the individual, then they are seen as endorsing a deontological decision. Deontology holds that the morality of an action depends on the intrinsic nature of the act itself (Bialek & de Neys, 2016). From this perspective, if harming is wrong, then it is wrong regardless of the potential benefits that could be derived from harming. If, however, one endorses harming the individual, then that decision aligns with a utilitarian view (Kahane et al., 2015). Utilitarianism holds that the morality of an action is determined by its consequences (Bialek & de Neys, 2016). Therefore, harm is acceptable if it increases the
well-being of a greater number of people (Bialek & de Neys, 2016).
The dual-process model of thinking can explain the choice for one of the two opposing moral views (Greene, 2003; Greene & Haidt, 2002; Greene, 2009). According to this model, cog-nition is an interplay between a bottom-up process, which is automatic, fast and emotional (also known as System 1) and a top-down process, which is controlled, slow and relatively unemotional (known as System 2; Evans, 2008; Kahneman, 2011; Chaiken, 1999; Greene, 2009). In order to give a util-itarian response to moral dilemmas, one must override the intuitively cued emotional System 1 process, which biases us against harming others (and therefore would make us choose the deontological option). This is done by allocating cognitive resources and thus activating System 2 processing (Bialek & de Neys, 2016; Greene, 2003; Greene, 2009).
Accordingly, research demonstrates System 1 and System 2 processes are engaged differently depending on the type of the moral dilemma (Gleichgerrcht & Young, 2013; Greene et al. 2011; Valdesolo & DeSteno, 2006). If presented with an impersonal dilemma, which lacks salient emotional content (e.g., would you push a button to direct a trolley onto one person and away from five other people?), most people would give a utilitarian response (Gleichgerrcht & Young, 2013). By contrast, if one is presented with a personal dilemma (e.g., would you push a person off a bridge to stop a trolley from killing five other people?) emotions are more strongly engaged, increasing the difficulty to override the intuitively cued emo-tional process by System 2 processes (Greene et al., 2004). This leads the majority of people to reject harming one person for the greater good (Gleichgerrcht & Young, 2013). This is corroborated by research showing that, in impersonal dilem-mas, brain areas associated with deliberative reasoning are more active while personal dilemmas predominantly activate regions associated with emotion (Greene et al., 2004). Per-sonal moral dilemmas might more strongly activate emotions because humans are argued to have an instinctual aversion to intentional, up and close, or direct hurting of others than more distant, less intentional and less direct killing (Sun, 2012).
Thus far, researchers have predominantly viewed utilitar-ian decisions as more favourable than deontological decisions. Deontological responses are shared by our great ape relatives and are thought to be devoid of moral reasoning, because they are said to be driven by System 1 processes (Greene et al. 2001; Haidt, 2001). By contrast, utilitarian reasoning involves abstraction and cognitive control, a process unique to humans (Greene et al., 2004). Nevertheless, recent find-ings paint a more complex picture of supposedly utilitarian
decision-making. One study found that patients with vmPFC damage (compared to healthy participants) show a more utili-tarian response to moral dilemmas (Koenig et al. 2007) that was associated with a reduced emotional response, but also reduced social emotions such as compassion, empathy and guilt (i.e., emotions associated with moral values). Another study found that a utilitarian response on moral dilemmas might not only result from enhanced cognitive control and abstract reasoning but also from reduced empathic concern (Gleichgerrcht & Young, 2013).
Considering these findings, it might be expected that in-dividuals with reduced empathy – such as inin-dividuals who score highly on psychopathic traits – are more likely to make utilitarian decisions. Psychopathy is conceptualised as a two-factor construct, each of which comprises two facets Lykken, 1995). Factor 1 psychopathy consists of the interpersonal and affective facet which are associated with decreased empathy, stress reactivity, higher callousness, and a manipulative use of others (Lykken, 1995). Factor 2 consists of the lifestyle and antisocial facet and is related to reduced impulse control and emotion regulation, and increased aggression (Lykken, 1995; Cornell et al. 1996). The majority of studies indeed demonstrate that scoring highly on psychopathic traits is as-sociated with utilitarian choices in moral dilemmas (Bartels & Pizarro, 2011; Glenn, Koleva, Iyer, Graham, & Ditto, 2010; Koenigs, Kruepke, Zeier, & Newman, 2012; Kahane et al., 2015; Crockett, Clark, Hauser, & Robbins, 2010; Gao & Tang, 2013). However, in other studies this association was not found (Cima, Tonnaer, & Hauser, 2010; Glenn, Raine, & Schug, 2009;
Pujol et al., 2012).
If one assumes that utilitarian decisions of people higher in psychopathy traits are reflective of a ‘true’ utilitarian decision-making process, then these findings would suggest that indi-viduals with higher psychopathic traits feel greater concern for the maximization of the greater good for others. However, this seems implausible because such individuals have a greater tendency to prioritize their self-interest over the interest of others and are less concerned with others’ well-being (Kahane et al., 2015). Thus, even if they choose a utilitarian option in impersonal moral dilemmas, it is likely that these individuals exhibit selfish behavior in real life that deviates from a utili-tarian moral perspective. Therefore, an association between utilitarian decision-making in sacrificial moral dilemmas and psychopathy traits might suggest that sacrificial moral dilem-mas are not a valid measurement of utilitarianism, or at least of truly utilitarian reasoning (Kahane et al., 2015).
The association between utilitarian decision-making in sac-rificial moral dilemmas and psychopathy traits seems plausible considering that those who score highly on psychopathy traits show reduced harm aversion (Crockett et al., 2010; Cushman, Gray, Gaffey, & Mendes, 2012). Reduced harm aversion is associated with reduced empathy, a component of Factor 1 psychopathy (Sinter et al. 2004) and entails that such indi-viduals feel lower aversion when harming other persons in an up-close and personal way (Greene, 2008). Reduced harm aversion therefore might direct people highly in psychopathic to even give utilitarian responses on personal moral dilemmas (Gleichgerrcht & Young, 2013; Greene et al., 2001). This would contrasts them with healthy participants who are thought to generally choose the deontological option on personal dilemmas because of their instinctual aversion against directly
inflict-ing harm on another person (Gleichgerrcht & Young, 2013; Greene et al., 2001). If one does not feel harm aversion, then choosing the utilitarian option might simply be a cost-benefit analysis in which saving four people is better than letting one live. In fact, using a cost-benefit analysis to guide decisions is an important feature shared by utilitarians and egoists (i.e., individuals who take only into account their self-interest and not the interest of others; Sidgwick, 1907). However, whereas the utilitarian cares about maximizing the greater good even if this does not benefit himself, the egoist only cares about his own good.
Because reduced emotion and empathy might underlie the possible bias for a utilitarian decision, one would expect that Factor 1 is associated with utilitarian decisions in people higher in psychopathy traits rather than Factor 2. However, the involvement of Factor 1 psychopathy in utilitarian decisions is currently unclear. Two studies show conflicting findings. One study found that Factor 2 traits of psychopathy contribute more strongly to utilitarian responses (Gao & Tang, 2013), while another found that utilitarian decisions were associated with low anxiousness, a trait of Factor 1 psychopathy (Koenigs et al., 2012). Therefore, in this study, we wish to investigate how Factor 1 and Factor 2 psychopathy are involved in moral decision-making.
A further way to assess the validity of sacrificial moral dilemmas is by investigating how it correlates with moral behaviour, which is thought to measure ‘true’ moral judgement (Kahane et al., 2015). If sacrificial moral dilemmas are a valid measurement of utilitarianism, we would expect that individuals who make utilitarian decisions in sacrificial moral dilemmas also show utilitarian behaviour (e.g., impartially maximize the greater good). However, if there is no association between utilitarian decisions on sacrificial moral dilemmas and utilitarian behaviour, then this potentially provides another reason to question whether sacrificial moral dilemmas measure ‘true’ utilitarian moral reasoning. One way to measure moral behaviour is by using a dictator game (Jakiela, 2012).
In a traditional dictator game, participants divide money between themselves and another anonymous person within the experiment (e.g., keep €30, give €20). Because the anonymous person cannot directly influence the decision of the person dividing the money, a dictator game is thought to provide an unconfounded measure of altruism (Jakiela, 2012). So far, however, research investigating the relationship between psychopathy and utilitarian decision-making did not include a dictator game. Therefore, for the current study, a modified dictator game was designed to measure utilitarian decision-making. A utilitarian way of dividing money would be to maximize total payoffs, even if this means keeping less money for oneself. For instance, a utilitarian would choose to keep €10 and give €50 instead of keeping €30 and giving €10. This is because the aim of the utilitarian is maximizing the total sum of the payoff. One could expect that persons scoring highly on psychopathy traits make more selfish decisions in the monetary task. This is because underlying their utilitarian decisions in the moral dilemmas lies a less empathetic and more egoistical motivation (Bartels & Pizarro, 2011; Glenn, Koleva, Iyer, Graham, & Ditto, 2010; Koenigs, Kruepke, Zeier, & Newman, 2012; Wiech et al., 2013, Kahane et al., 2015; Kahane et al., 2015).
dilemmas by researching the association between psychopa-thy traits and utilitarian decision-making and the association between utilitarian decision-making and real life utilitarian behaviour. We hypothesize that individuals scoring highly on psychopathic traits will choose the utilitarian decision more often than individuals lower in psychopathy traits on personal and impersonal dilemmas. We expect this association to be stronger for the personal dilemmas. Additionally, because we believe reduced harm aversion and empathy might underlie the possible association between utilitarian decision-making and psychopathy traits, we expect that Factor 1 psychopathy con-tributes to utilitarian responses but not Factor 2 psychopathy. By means of a behavioural experiment, we will also investigate whether utilitarian decisions in sacrificial moral dilemmas are associated with utilitarian decisions on a modified dictator game. Finally, we hypothesize individuals higher in psychopa-thy traits will choose a more egoistical way to divide money in the dictator game.
Methods
Participants
The current study comprises two experiments. Participants in the first study (N = 257) were undergraduate psychology students at the University of Amsterdam (see Table 1). All participants partook in the first study as part of a mandatory battery of psychological experimental tasks and questionnaires. The participants were compensated with research credits for partaking in the study. We excluded 27 participants from data-analysis because they did not pass the attention check or correctly follow the study procedure. Participants in the second study (N = 30) were recruited through an advert on the university lab web page. Participants were undergraduate psychology students at the University of Amsterdam (see Table 1). Participants could choose between a compensation of one research credit or €10. One participant was excluded from data-analysis for not passing the attention check. Both participants from the first and second study were informed that – if they were chosen at random – they would receive the money they had earned in the modified dictator game, and their donations would also really be made.
Table 1. Demographic characteristics of participants Study 1 (N = 257) Study 2 (N = 30) Mean age (SD) 20.0 (2.5) 21.7 (3.2)
Gender (N) Females 176 16
Males 72 14
*Nine participants did not provide gender.
Materials
Sacrificial moral dilemmas. Six sacrificial moral dilemmas were
presented to the participants, three were impersonal while the other three were personal dilemmas (see Appendix A for the selected dilemmas). Three dilemmas (Fumes, Motorboat and Trolley) were used in previous studies and left unaltered while two (Vitamines and Earthquake) were altered to fit our criteria (see Appendix B). One dilemma (Mineworkers) was designed by the researchers. The dilemmas varied in the number of people dying when choosing not to sacrifice the individual. The number of people dying could be four, five or six. Participants
rated to what extent they found a utilitarian action acceptable on a 6-point Likert scale (-2.5 ‘Completely unacceptable’, 2.5 ‘Completely acceptable’). The total score ranged between -15 and 15 with positive scores signifying a utilitarian outlook.
Self-Report Psychopathy Scale – Short Form (SRP-SF; Dutch translation). The SRP-SF is a measure of the
interper-sonal, affective, lifestyle and antisocial factors of psychopathy (Welker, Lozoya, Campbell, Neumann, & Carré, 2014). The questionnaire comprises 29 items on a 5-point Likert-scale (1 “Strongly disagree”, 5 “Agree strongly”; see Appendix C for the items). The SRP-SF has a score range from 29 to 145. A higher score indicates higher psychopathic traits. The SRP-SF is a valid and internally consistent scale to measure psychopathy traits in university students (Neal & Sellbom, 2012; Williams, Paulhus & Hare, 2007). Cronbach’s α for the four factors ranges between .62 and .92, indicating acceptable reliability (Welker et al., 2014; Jones, 2014).
Modified dictator game. As a behavioral measure of
utilitari-anism, we designed two monetary dictator games. Participants in Study 1 were asked to divide money between themselves and a anonymous randomly selected other participant or a charity organization of their choice. In Study 2, participants were asked to divide money between themselves and another partic-ipant they met in the waiting room before taking part in the experiment. We chose to incorporate this social component in Study 2 because we believe that seeing the other person might increase ecological validity, since there aren’t many occasions in real life in which people divide money between themselves and one anonymous other person. For both Study 1 and 2, the modified dictator tasks featured two separate sliders where the options the participants could choose from were differentially utilitarian. On one of the sliders, participants could give away a maximum of €75 and keep a maximum of €37,5. On the other slider, the maximum amount of money to give away was €100 and the maximum amount to keep €50. The sliders were designed in such a way that every euro donated would be worth twice as much for the other participant, while every euro taken from the donation will be half as much worth for oneself. Therefore, the most utilitarian option was to donate everything. The amounts of money were chosen in such a way that they would make a meaningful difference for the participants (i.e., one could earn or give a substantial amount of money).
Procedure
In Study 1, participants took part as part of a mandatory battery of psychological experimental tasks and questionnaires. Participants were presented with six moral dilemmas and two dictator games (in randomized order). We used pseudo-randomization to prevent influence of the order in which moral dilemmas (personal vs. impersonal) were shown and the order in which moral dilemmas and monetary task were shown. After the moral dilemmas and dictator game, participants filled out the SRP-SF and a demographics questionnaire.
In Study 2, participants were recruited online and tested in the university lab. In this study, participants were unknow-ingly paired with another participant or an actor-participant to participate in the study. We chose to include an ‘actor-participant’ in order maximize our sample size and statistical power. For the actor-participant, we designed a script to en-sure standardization in the interaction. Both participants were asked to read the information brochure while the researcher
left them for two minutes to retrieve the informed consent forms. We did this to give participants an opportunity to socialize with each other. In the brochure, participants are informed that one of them would be randomly selected to perform an extra task (the modified dictator game) based on the color of the card in the envelope they drew from a jar. In reality, both participants were assigned to this “extra” task, but they did not know this of each other. Its purpose was to make participants believe that they would divide money between them and the other/actor participant. Once the researcher returned with the forms, both participants were asked to draw an envelope from a jar and asked to wait with opening the envelope until they were inside their test cubicle. In the cubicle, participants were presented the moral dilemmas and the dictator games (in randomized order). Subsequently, participants were presented with the SRP-SF, LSAS and a demographics questionnaire.
In both studies, participants were informed before making their decision on the dictator game that two randomly chosen participants would be selected whose choices on the dictator game would be used to allocate real money.
Results Study 1
Confirmatory analysis
In order to examine the effects of Factor 1 and Factor 2 psy-chopathy on the choice made in personal dilemmas (deontolog-ical vs. utilitarian), we conducted a multiple regression with Factor 1 and 2 psychopathy as independent variables and per-sonal dilemma choice as dependent variable. All assumptions of multiple regression were met (i.e., multicollinearity, ho-moscedasticity, uncorrelated residuals, linearity and normally distributed residuals). This general linear model significantly explained the data, F(2, 254) = 5.25,p = .006, R2 = .04. Factor 1 psychopathy was significantly related to the choice on personal dilemmas, b = 0.024, SE = 0.01, t(254) = 2.43, p = .008. The positive value for b signifies that as one scores higher on Factor 1, the more likely one is to choose the utili-tarian decision on personal dilemmas. Moreover, as expected, Factor 2 was not significantly related to the choice on personal dilemmas, b = .00, SE = 0.01, t(254) = 0.03, p = .49.
To examine the effects of Factor 1 and Factor 2 psychopathy on the choice made in impersonal dilemmas (deontological vs. utilitarian), a multiple regression with Factor 1 and 2 psychopathy as independent variables and impersonal dilemma choice as dependent variable. All assumptions of multiple regression were met. This general linear model significantly explained the data, F(2, 254) = 3.84, p = .02, R2 = .03.
Factor 1 psychopathy was significantly related to the choice on impersonal dilemmas, b = 0.02, SE = 0.01, t(254) = 2.02,
p= .02. Higher scores on Factor 1 psychopathy are associated
with more utilitarian decisions. However, as with personal dilemmas, Factor 2 was not significantly related to the choice on impersonal dilemmas, b = .00, SE = 0.01, t(254) = 0.11, p = .46.
To investigate if the association between utilitarian decision making and Factor 1 psychopathy was stronger for personal dilemmas we computed confidence intervals for the effect sizes and looked for overlap. The confidence interval for personal dilemmas was BCa 95% CI [.23, .58] whereas it was BCa 95%
CI [.16, .51] for impersonal dilemma. This shows that the confidence intervals indeed overlap.
Two simple regression analyses were conducted to research the relation between sacrificial dilemmas and sum of money donated. Assumptions for both simple regression analyses were met (e.g. homoscedasticity, uncorrelated residuals, nor-mally distributed residuals and linearity). Choice on personal dilemmas was not significantly related to sum of money do-nated, b = 0.00, SE = 0.00, t(255) = -0.18, p = .43, nor was choice on impersonal dilemmas, b = 0.00, SE = 0.00, t(255) = -0.46, p = .33.
To examine the effects of psychopathic traits on sum of money donated, we conducted a multiple regression with Fac-tor 1 and 2 psychopathy as independent variables and sum of money donated as dependent variable. All assumptions of multiple regression were met. The general linear model did not significantly explain the data, F(2, 254) = 1.22, p = .30,
R2 = 0.01. Factor 1 psychopathy was not significantly related
to the sum donated, b = -0.23, SE = 0.56, t(254) = -0.40, p = .35. Factor 2 also was not significantly related to sum donated on the modified dictator game, b = -0.68, SE = 0.78, t(254) = -0.88, p = .19.
Exploratory analysis
An exploratory analysis was conducted to investigate the rela-tionship between the facets comprising Factor 1 and decisions made in personal and impersonal dilemmas. For this purpose, we conducted two multiple regressions with the interpersonal facet and affective facet as independent variables and deci-sions made on personal or impersonal dilemmas as dependent variables. For both regressions all assumptions were met.
First, we investigated the relationship between the interper-sonal and affective facet and choice on perinterper-sonal dilemmas. The general linear model significantly explained the data, F(2,254) = 5.50, p = .005, R2 = .04. The interpersonal facet was
significantly related to decisions on personal dilemmas, b = 0.04, SE = 0.02, t(254) = 1.99, p = .03. This means that the higher one scores on the interpersonal facet in Factor 1, the more likely one is to choose the utilitarian decision on personal dilemmas. However, the affective facet was not significantly related decisions on personal dilemmas, b = .01, SE = 0.02,
t(254) = 0.51, p = .31.
Second, the relation between the interpersonal and affective facet and choice on impersonal dilemmas was investigated. The general linear model significantly explained the data, F(2, 254) = 4.13, p = .02, R2 = .03. The interpersonal facet was
significantly related to choice on personal dilemmas, b = 0.03,
t(254) = 1.85, p = .04. The affective facet was not significantly
related to choice on personal dilemmas, b = 0.00, SE = 0.02,
t(254) = 0.28, p = .39.
Study 2
In order to research the association between utilitarian decision-making and psychopathy traits, we conducted a multiple re-gression with Factor 1 and 2 psychopathy as independent variables and personal dilemma choice as dependent variable. All assumptions of multiple regression were met. The general linear model did not significantly explain the data, F(2,27) = 1.12, p = .34, R2 = .08. Contrary to the hypothesis, Factor
1 psychopathy was not significantly related to the choice on personal dilemmas, b = -0.02, SE = 0.03, t(27) = -.82, p =
.42. Moreover, Factor 2 was also not significantly related to the choice on personal dilemmas, b = -0.02, SE = 0.03, t(27) = -.45, p = .33. To examine the relationship of Factor 1 and Factor 2 psychopathy and choice made in impersonal dilem-mas, a multiple regression with Factor 1 and 2 psychopathy as independent variables and impersonal dilemma choice as dependent variable. All assumptions of multiple regression were met. This general linear model did not significantly ex-plain the data, F(2, 27) = 0.91, p = .42, R2 = .06. Factor
1 psychopathy was not significantly related to the choice on impersonal dilemmas, b = 0.04, SE = 0.04, t(27) = 1.26, p = .11, nor was Factor 2 psychopathy, b = -0.02, SE = 0.04,
t(28) = -0.41, p = .34.
Two bootstrapped simple regression analyses were con-ducted to investigate the relation between sacrificial moral dilemmas and donation. A bootstrapped method was chosen because the data were not normally distributed. All other assumptions were met. Choice made on personal dilemmas was not a significant predictor for sum of money donated,
F(1, 28) = 0.05, p = 0.95, R2 = .00. Choosing the utilitarian
option on personal dilemmas was not significantly related to sum of money donated, b = 0.74, SE = 11.14, t(28) = 0.07, p = .47. Choice made on impersonal dilemmas was also not a significant predictor for sum of money donated, F(1, 28) = 0.12, p = 0.73, R2 = .004. Choosing the utilitarian option on
impersonal dilemmas, was not significantly related to sum of money donated, b = -2.81, SE = 9.27, t(28) = -0.35, p = .39. To examine the relation between psychopathic traits and sum of money donated, we conducted a bootstrapped multiple regression with Factor 1 and 2 psychopathy as independent variables and sum of money donated as dependent variable. We bootstrapped because data were not normally distributed. The general linear model did not significantly explain the data,
F(2, 27) = 0.99, p = .91, R2 = .01. Factor 1 psychopathy was
not significantly related to the sum donated, b = -0.44, SE = 1.61, t(27) = -0.29, p = .27, nor was Factor 2 psychopathy, b = -0.15, SE = 2.39, t(27) = -0.08, p = .47.
Lastly, a bootstrapped simple regression analysis was con-ducted to research the relation between sum of money donated and likability and social attractiveness of the other partici-pant. Again, a bootstrapped method was chosen because of non-normally distributed data. The general linear model did not significantly explain the data, F(1 ,28) = 0.85, p = .37,
R2 = .17. Likability and social attractiveness of the other
participant was not significantly related to the sum donated,
b= -21.69, SE = 17.20, t(27) = -0.92, p = .18.
Discussion
We conducted two studies to assess the validity of sacrifi-cial moral dilemmas in measuring utilitarian decision-making. First, we investigated how psychopathy traits are related to utilitarian decision-making on moral dilemmas. Second, we in-vestigated the association between utilitarian decision-making on moral dilemmas and real life utilitarian behavior. In Study 1, we found that utilitarian decision-making was associated with stronger psychopathic traits. The more highly one scores on psychopathy traits, the more likely one is to choose utilitar-ian options on sacrificial moral dilemmas. We also investigated how Factor 1 and 2 relate to this association, because research has been inconclusive in this respect (Gao & Tang, 2013;
Koenigs et al., 2012). As expected, we found that Factor 1 was related to utilitarian decision-making while Factor 2 was not. The association between utilitarian decision making and Factor 1 psychopathy was not stronger for personal dilemmas. Furthermore, results showed no association between utilitarian decisions and money donation to another anonymous partici-pant or a charity in the modified dictator game. Our data did not support the hypothesis that psychopathy traits and dona-tion rates are negatively associated. In Study 2, no support was found for our hypotheses. It should be noted, however, that study 2 was underpowered to detect the relatively small effects observed in study 1. For the following discussion, we will therefore focus more on study 1.
The results of Study 1 further call into question the validity of sacrificial moral dilemmas as a measurement of a utilitarian moral view. First, this study corroborates previous research by finding an association between utilitarian decisions on moral dilemmas and psychopathy traits. Second, we found no association between utilitarian decisions on moral dilemmas and money donation.
A reason for the absent association between utilitarian de-cisions on moral dilemmas and money donation might have to do with how sacrificial moral dilemmas are constructed. Moral dilemmas reflect an extreme type of life or death decision which is very rarely encountered in life (Kahane, 2015). Further-more, it is assumed that if one rejects harming another in this unusual context, this automatically represents endorsement of a utilitarian moral view. However, rejecting a deontologi-cal decision does not necessarily mean the endorsement the utilitarian aim of impartially maximizing aggregate well-being (Bartels & Pizzarro, 2011; Kahane, 2015; Kahane & Shackel, 2010). This is underlined by prior studies showing that utili-tarian decisions on sacrificial dilemmas are not reflective of an impartial attitude, seeing as utilitarian decisions are strongly influenced by whether dilemmas involve sacrificing (or saving) foreigners or persons from one’s own country (Swann et al. 2010), strangers or family members (Petrinovich, O’Neill, & Jorgensen, 1993) or animals and humans (Petrinovich et al., 1993). If utilitarian decisions on sacrificial dilemmas would truly measure a utilitarian moral view, the content of these dilemmas should not matter. A true utilitarian would consider options from a maximally impartial perspective that gives equal weight to interest of all sentient beings (Singer, 1979). Thus, sacrificial moral dilemmas might not be a valid way to measure a utilitarian view, not only because of its association with psychopathy, but also because utilitarian choices on moral dilemmas may not relate to actual utilitarian moral behavior.
Interestingly, in contrast to our expectation, we found no relation between Factor 1 and 2 of psychopathy and dona-tion rates. This calls into quesdona-tion whether our measurement of moral behavior is valid. However, a potential explana-tion for this lack of associaexplana-tion could be that females were overrepresented in our sample (176 females versus 72 males). Psychopathy is known to manifest differently in women than in men (Hare, 2003, pp. 224), so perhaps this also translates to different donation patterns in males and females high in psychopathy traits. Unfortunately, due to the limited number of male participants, we could not investigate such differences with sufficient statistical power. Therefore, we recommend that future studies include more males in their sample.
option in our monetary dilemma may not differentiate well enough between utilitarianism and deontology. That is, donat-ing more money does not go against the deontologists principle of not using individuals as a means to an end (Bialek & de Neys, 2016). For future studies, it is therefore advised to design a monetary dilemma which pits utilitarian outcomes against deontological outcomes. For example a monetary dilemma where one could take money from another participant anony-mously and allocate the greater amount gained to charity. Taking money would violate the rights of the individual, and thus would go against deontological principles (Bialek & de Neys, 2016). At the same time, taking money would maximize total payoffs and is therefore justified from a utilitarian point of view. Another limitation of this study might be that we only included undergraduate psychology students. Therefore, the results might not generalize to the general population. However, other studies researching this subject found similar results while having a more diverse subject pool (e.g., Kahane, 2015; Gleichgerrcht & Young, 2013).
It should also be noted that variance explained by Factor 1 psychopathy was very small for personal dilemmas and for impersonal dilemmas. It seems therefore that even though the results were significant and the effect sizes were small to medium, Factor 1 psychopathy only explains choosing a utilitarian option for fairly small amount. Nevertheless, the association between psychopathy and utilitarian decisions calls into question the validity of sacrificial moral dilemmas.
In our exploratory analysis of Study 1, we found that the interpersonal facet was associated with utilitarian decisions, whereas the affective facet was not. This result was surprising, because previous research provided reason to assume that lack of empathy – which is part of the affective facet – could explain the association between Factor 1 psychopathy and utilitarian decision-making. However, as mentioned earlier, the inter-personal facet comprises amongst other things interinter-personal manipulation for personal gain (Hare, 2003, pp. 234) and is also a predictor of instrumental aggression (Vitacco et al., 2006). Arguably, when one uses instrumental aggression, this indicates lack of empathy for others. In fact, social emotions such as empathy, guilt, and remorse, which psychopaths lack, discourage instrumentally aggressive acts (Glenn & Raine, 2009). Since sacrificial moral dilemmas involve other people’s fate, it is still plausible that a lack of empathy underlies the re-lation between Factor 1 and utilitarian judgment on sacrificial moral dilemmas. However, this does not explain why we found no association between the emotional facet and utilitarian decision-making.
This could also explain why we did not find that utilitarian decision making and Factor 1 psychopathy were more strongly associated for personal dilemmas. Utilitarian decision making is only associated with the interpersonal facet, and not directly with the affective facet. This could mean that harm aversion does not direct individuals highly in psychopathy traits to utilitarian decisions. Therefore you would also not expect this association to be stronger for personal dilemmas.
To conclude, our study not only found that utilitarian decisions were associated with psychopathy traits, but also that there was no association between utilitarian decisions on moral dilemmas and utilitarian behavior. Therefore, our study provides more evidence to question the use of sacrificial moral dilemmas as tools for investigating moral reasoning.
Literature
1. Bartels DM, Pizarro DA (2011) The mismeasure of morals: Antisocial personality traits predict utilitarian responses to moral dilemmas. Cognition 121(1):154–161.
2. Białek M, De Neys W (2016) Conflict detection during moral decision making: Evidence for deontic reasoners’ utilitarian sensitivity. Journal of Cognitive Psychology 28(5):631–639. 3. Cima M, Tonnaer F, Hauser MD (2010) Psychopaths know right from wrong but don’t care.
Social Cognitive and Affective Neuroscience 5(1):59–67. 4. Cohen J (1992) A power primer. Psychological Bulletin (1):155–159.
5. Cornell DG, et al. (1996) Psychopathy in instrumental and reactive violent offenders. Journal of Consulting and Clinical Psychology 64(4):783–790.
6. Crockett MJ, Clark L, Hauser MD, Robbins TW (2010) Serotonin selectively influences moral judgment and behavior through effects on harm aversion. Proceedings of the National Academy of Sciences of the United States of America 107(40):17433–17438.
7. Croes EAJ, Antheunis ML, Schouten AP, Krahmer EJ (2016) Teasing apart the effect of vis-ibility and physical co-presence to examine the effect of CMC on interpersonal attraction. Computers in Human Behavior 55:468–476.
8. Cushman F, Gray K, Gaffey A, Mendes WB (2012) Simulating murder: The aversion to harm-ful action. Emotion (Washington, D.C.) 12(1):2–7.
9. Evans JSB (2008) Dual-processing accounts of reasoning, judgment,and social cognition. Annual Review of Psychology 59(1):255–278.
10. Gao Y, Tang S (2013) Psychopathic personality and utilitarian moral judgment in college students. Journal of Criminal Justice 41(5):342–349.
11. Gleichgerrcht E, Young L (2013) Low levels of empathic concern predict utilitarian moral judgment. PLoS One 8(4):e60418.
12. Glenn AL, Koleva S, Iyer R, Graham J, Ditto PH (2010) Moral identity in psychopathy. Judg-ment and Decision Making 5(7):497–505.
13. Glenn AL, Raine A, Schug RA (2009) The neural correlates of moral decision-making in psychopathy. Molecular Psychiatry 14(1):5–6.
14. Greene J (2003) From neural ‘is’ to moral ‘ought’: What are the moral implications of neuro-scientific moral psychology? Nature Reviews. Neuroscience 4(10):846–850.
15. Greene JD (2009) The cognitive neuroscience of moral judgment. Cognitive Neuroscience 4:1–48.
16. Greene J, Haidt J (2002) How (and where) does moral judgment work? Trends in Cognitive Sciences 6(12):517–523.
17. Greene JD, Nystrom LE, Engell AD, Darley JM, Cohen JD (2004) The neural bases of cogni-tive conflict and control in moral judgment. Neuron 44(2):389–400.
18. Greene JD, Sommerville RB, Nystrom LE, Darley JM, Cohen JD (2001) An fMRI investigation of emotional engagement in moral judgment. Science 293(5537):2105–2108.
19. Haidt J (2001) The emotional dog and its rational tail: A social I ntuitionist approach to moral judgment. Psychological Review 108(4):814–834.
20. Hare RD (2003) The Psychopathy Checklist–Revised.
21. Hauser M, Cushman F, Young L, Kang-Xing Jin R, Mikhail J (2007) A dissociation between moral judgments and justifications. Mind & Language 22(1):1–21.
22. Jakiela P (2013) Equity vs. efficiency vs. self-interest: On the use of dictator games to mea-sure distributional preferences. Experimental Economics 16(2):208–221.
23. Jones DN (2014) Risk in the face of retribution: Psychopathic individuals persist in financial misbehavior among the Dark Triad. Personality and Individual Differences 67:109–113. 24. Kahane G, Everett JA, Earp BD, Farias M, Savulescu J (2015) ‘Utilitarian’ judgments in
sacrifi-cial moral dilemmas do not reflect impartial concern for the greater good. Cognition 134:193– 209.
25. Kahneman D (2011) Thinking, fast and slow. (Farrar, Strauss, Giroux, New York). 26. Koenigs M, Kruepke M, Zeier J, Newman JP (2012) Utilitarian moral judgment in psychopathy.
Social Cognitive and Affective Neuroscience 7(6):708–714.
27. Lykken DT (1955) A study of anxiety in the sociopathic personality. Journal of Abnormal and Clinical Psychology 55(1):6–10.
28. McCroskey LL, McCroskey JC, Richmond VP (2006) Analysis and improvement of the mea-surement of interpersonal attraction and homophily. Communication Quarterly 54(1):1–31. 29. Neal TM, Sellbom M (2012) Examining the factor structure of the Hare Self-report
Psychopa-thy Scale. Journal of Personality Assessment 94(3):244–253.
30. Paulhus DL, Williams KM (2002) The dark triad of personality: Narcissism, Machiavellianism, and psychopathy. Journal of Research in Personality 36(6):556–563.
31. Pujol J, et al. (2012) Breakdown in the brain network subserving moral judgment in criminal psychopathy. Social Cognitive and Affective Neuroscience 7(8):917–923.
32. Sidgwick H (1907) The methods of ethics. (Hackett Publishing).
33. Sun R (2013) Moral judgment, human motivation, and neural networks. Cognitive Computa-tion 5(4):566–579.
34. Singer P (1979) Killing humans and killing animals. Inquiry 22(1-4):145–156.
35. Swann, Jr WB, Gómez Á, Dovidio JF, Hart S, Jetten J (2010) Dying and killing for one’s group: Identity fusion moderates responses to intergroup versions of the trolley problem. Psychological Science 21(8):1176–1183.
36. Valdesolo P, DeSteno D (2006) Manipulations of emotional context shape moral judgment. Psychological Science 17(6):476–477.
37. Vitacco MJ, Neumann CS, Caldwell MF, Leistico AM, Van Rybroek GJ (2006) Testing factor models of the Psychopathy Checklist: Youth Version and their association with instrumental aggression. Journal of Personality Assessment 87(1):74–83.
38. Welker KM, Lozoya E, Campbell JA, Neumann CS, Carré JM (2014) Testosterone, cortisol, and psychopathic traits in men and women. Physiology & Behavior 129:230–236. 39. Williams KM, Paulhus DL, Hare RD (2007) Capturing the four-factor structure of psychopathy
Appendix A: Moral dilemmas
Impersonal.
1. Fumes You are the late-night watchman in a hospital. Due
to an accident in the building next door, there are deadly fumes rising up through the hospital’s ventilation system. In a certain room of the hospital are four patients. In another room there is a single patient. If you do nothing the fumes will rise up into the room containing the four patients and cause their deaths. The only way to avoid the deaths of these patients is to hit a switch, which will cause the fumes to bypass the room containing the four patients. As a result of doing this the fumes will enter the room containing the single patient, caus-ing his death. How acceptable is it to hit the switch in order to save the four patients? Completely unacceptable Unac-ceptable Slightly unacUnac-ceptable Slightly acUnac-ceptable AcUnac-ceptable Completely acceptable
2. MineworkersYou are working in the control room of a
mine. An accident has occurred and the mine is flooding. Five mineworkers are mining in a passage that is in the path of the flooding water and there is no way for them to escape: they will drown when the water reaches them. One mineworker is in a separate passage that is closed off by a watertight door and therefore safe from the water. The only way to save the five mineworkers is to open the watertight door from your control-room. If you do so, the water will be diverted from the passage containing the five miners. The one mineworker in that room will drown, but the water won’t flow into the passage with the five mineworkers and they will live. How acceptable is it to press open the door in order to save the five mineworkers? Completely unacceptable Unacceptable Slightly unacceptable Slightly acceptable Acceptable Completely acceptable
3. TrolleyYou are at the wheel of a runaway trolley quickly
approaching a fork in the tracks. On the tracks extending to the left is a group of six railway workmen. On the tracks extending to the right is a single railway workman. If you do nothing the trolley will proceed to the left, causing the deaths of the five workmen. The only way to avoid the deaths of these workmen is to press a button on your dashboard that will cause the trolley to proceed to the right, causing the death of the single workman. How acceptable is it to press the button in order to save the six workmen? Completely unacceptable Unacceptable Slightly unacceptable Slightly acceptable Ac-ceptable Completely acAc-ceptable
Personal.
1. Motorboat You are the driver of a motorboat. You notice
four swimmers in severe trouble in the distance. If you do not drive towards them at top speed then you will not arrive on time, and all four will certainly die. You have a passenger on the boat, making the boat too heavy to drive at top speed. The only way to save the swimmers is to quickly push the passenger off the boat. The passenger cannot swim and will drown, but the boat will be light enough to drive at top speed and save all four of the drowning swimmers. How acceptable is it to push the passenger off the boat in order to save the four swimmers? Completely unacceptable Unacceptable Slightly unacceptable Slightly acceptable Acceptable Completely acceptable
2. VitaminsYou are a doctor flying with a group on a travel
expedition. Your group has had to do an emergency landing in the desert. You are several days away from any means of help. The group includes you (the only person with medical
knowledge and abilities), the pilot, and a party of five people with a rare genetic disease: they have a vital vitamin deficiency and must receive the vitamin through supplements to their diet or they will die. There is plenty of water for everyone to survive, but without the vitamin the party of five will die before you can reach help. Normal human kidneys contain large amounts of the vitamin they need. You cannot remove your own kidney or teach the others to do it to you, and the kidneys of the people with the disease do not contain the vitamin. The only way to save the lives of the five people is to remove one of the pilot’s kidneys and use it as a source of the vitamin. He will die if you do this, but the other five people will be saved. How acceptable is it remove the pilot’s kidney in order to save the five people? Completely unacceptable Unacceptable Slightly unacceptable Slightly acceptable Acceptable Completely acceptable
3. EarthquakeYou are a doctor in the intensive care unit
of a hospital. An earthquake strikes, and you are trapped in a room with seven critically ill patients. One patient is on a life-support machine and can live for several days. The other six patients will die without immediate care from rescue crews. The only way for to signal for immediate help from rescue crews is to kill the patient on the life support machine, which will trigger an alarm system into “code red.” A “code red” alarm will attract the rescue teams immediately, and the other six patients will be able to live. If you do not kill the patient, the one patient on the life-support machine will live, but rescue crews will not arrive in time and the other six patients will die. How acceptable is it to take the life of the patient on life support in order to save the other six patients? Completely unacceptable Unacceptable Slightly unacceptable Slightly acceptable Acceptable Completely acceptable
Appendix B: Moral Dilemma Criteria
We selected these moral dilemmas based on the following criteria.
1. A utilitarian decision could not mean that participants would also save themselves. Otherwise, a reason for choos-ing that option could be self-preservation rather than maximizing the greater good.
2. The person to be sacrificed was not going to die anyway. 3. The person to be sacrificed was without bad intent (e.g.,
not planning to kill everyone).
Appendix C: SRP III - SF
Please rate the degree to which you agree with the following statements. You can be honest because your name will be detached from the answers as soon as they are submitted.
1. I’m a rebellious person.
2. I have never been involved in delinquent gang activity. 3. Most people are wimps.
4. I’ve often done something dangerous just for the thrill of it.
5. I have tricked someone into giving me money.
7. I have pretended to be someone else in order to get some-thing.
8. I like to see fist-fights.
9. I would get a kick out of ‘scamming’ someone.
10. It’s fun to see how far you can push people before they get upset.
11. I enjoy doing wild things.
12. I have broken into a building or vehicle in order to steal something or vandalize.
13. I don’t bother to keep in touch with my family any more. 14. I rarely follow the rules.
15. You should take advantage of other people before they do it to you.
16. People sometimes say that I’m cold-hearted. 17. I like to have sex with people I barely know. 18. I love violent sports and movies.
19. Sometimes you have to pretend you like people to get something out of them.
20. I was convicted of a serious crime.
21. I keep getting in trouble for the same things over and over.
22. Every now and then I carry a weapon (knife or gun) for protection.
23. You can get what you want by telling people what they want to hear.
24. I never feel guilty over hurting others.
25. I have threatened people into giving me money, clothes, or makeup.
26. A lot of people are “suckers” and can easily be fooled. 27. I admit that I often “mouth off” without thinking. 28. I sometimes dump friends that I don’t need any more. 29. I purposely tried to hit someone with the vehicle I was
driving.
Appendix D: Likeability and Social Attractiveness Scale
In Study 2, we included the LSAS to assess the likeability and social attractiveness of recipient participants as perceived by the participant dividing the money. The LSAS comprises 8 items on a 5-point Likert-scale (1 “Strongly disagree”, 5 “Agree strongly”) and has a score range from 8 to 40 (see items below). A higher score indicates that the participant was rated as more likeable and socially attractive. Seven of the items were taken from the measurement of interpersonal attraction by McCroksey, McCroskey and Richmond (2006). The eighth item was taken from Croes, Antheunis, Schouten, & Krahmer (2016).
Please rate the degree to which you agree with the following statements.
1. I think the other participant could be a friend of mine 2. I think the other participant and me could establish a
friendship with each other
3. I think the other participant and me could have a friendly chat
4. The other participant wouldn’t fit into my circle of friends 5. The other participant would be pleasant to be with 6. The other participant is unpleasant to be around 7. The other participant and me could become good friends 8. I would like to meet up again with the other participant
