The Importance of Unfair Intentions and Outcome Inequality for Punishment by Third Parties and Victims
Abstract
Abstract. Retributive theories predominantly focus on third party’s motives for punishment, which are rather affected by the offender’s malicious intentions than the actual outcome of the offense. However, victims experience an offense from a different perspective. The value/status approach argues that an offense has two facets that produce different threats: the intentional violation of values and status imbalance between offender and victims. We suggested that third parties and victims punish unfair intentions, whereas victims also punish because of the outcome inequality. In the present study, we orthogonally crossed the factors offender’s intention with the actual outcome and perspective of punisher (third-party versus victim). Results show that victims punish harsher than third parties. However, there are no qualitative differences of third-party punishment and punishment by victims. Rather, both punish malicious intentions and outcome inequality. We discuss how the findings relate to retributivism and other psychological theories of punishment.
Problem
Individuals’ intuitions to punish are generally retributive. They punish offenders in proportionality to the severity of the offense without much further regard to whether the punishment may affect deterrence or rehabilitation of an offender (Carlsmith & Darley, 2008; Gromet & Darley, 2009a; Vidmar, 2001). Considering the severity of the offense as the main determinant of punishment, we may overlook the different perspectives of victims and third parties because it comprises two conceptually different facets of an offense: the offender’s malicious intentions and the actual outcome of their action.
The value/status approach argues that these two facets produce conceptually different threats (Wenzel et al., 2008): First, an offense raises doubts about the validity and importance of the violated norms and values because the offender intentionally violated them. Second, an offense produces a status imbalance between offender and victim because it benefits the offender at the expense of the victim. Considering the value/status approach, we suggest that although third parties (i.e., observers) and victims similarly punish for offenders’ malicious intentions, such as unfair decisions, they differ in their punishment tendencies of the mere outcome inequality produced by the offense. The current experiment disentangles the influence of these two theoretically distinct facets of an offense and compares punishment by third parties and victims.
Existing research focuses on punishment by either victims or third parties, but a systematic comparison between both perspectives is rare (e.g., see Gummerum & Chu, 2014). If victims and third parties follow different intuitions about what to punish, then it would be difficult to create a punishment that is perceived as just by all. This could threaten alternative approaches to conflict resolution, which include victims as well as the community in justice processes.
Review of Relevant Scholarship
Retribution as a Psychological Motive for Punishment
In philosophical research, retributivism has several (normative) conceptualizations (e.g., Cottingham, 1979; Walker, 1999). However, most scholars agree that the crime suffices to justify punishment (e.g., see Moore, 1993). According to Kant (1785/2007), perpetrators have to be punished because of their “internal wickedness,” which is manifested in the crime committed. Other potential effects of punishment are not central to their justification. These include reform of the offender and deterrence of the offender or others of future crimes. Hence, the maliciousness of the offender’s intention and thereby the offender’s responsibility and guilt, determine “just” punishment.
Psychological studies indicate that lay intuitions about punishment correspond to retributivism (Carlsmith et al., 2002; Gromet & Darley, 2009b). According to these studies, punishment is thought to re-establish the moral imbalance produced by the offense (Bies & Tripp, 1996; Carlsmith & Darley, 2008; Darley & Pittman, 2003). A just punishment corresponds to the severity of the offense. Indeed, punishers consider the severity of the offense as a key factor in determining the amount of punishment (for comprehensible overviews see, Carlsmith & Darley, 2008; Gromet & Darley, 2009b). Individuals punish highly severe offenses (e.g., murder) more severely than less severe offenses (e.g., theft; Gromet & Darley, 2006). The perceived severity seems to imply the offender’s intention. More severe offenses require more malicious intentions, such as the intention to kill compared to stealing. Moreover, intentionally produced harm is judged and punished more severely than unintentional harm (Ames & Fiske, 2013; Darley & Huff, 1990). Thus, the intention and the negative outcome of an offense are closely associated, even though they differ conceptually. Their separate effects are not assessed when focusing on the severity of the offense as a determinant of punishment, and participants in these studies are usually in a third-party perspective.
Intention, Value Violation, and Punishment
To understand the role of malicious intentions and the negative outcome produced by the offense in determining punishment, one has to differentiate between the intention to do something and the actual outcome of the offense (Hechler & Kessler, 2018; Malle & Knobe, 1997). This notion is reflected in the value/status approach to punishment (Wenzel et al., 2008). According to this approach, an offense produces at least two conceptually different threats: an intentional value violation and a status imbalance between offender and victim. A value violation implies the offender’s malicious intentions, whereas the negative outcome reflects a status imbalance between offender and victim.
An intentional deviation from “how one ought to behave” given the moral convictions, perceived norms, expectations, and/or codes of conducts within a particular context (here referred to as “values”), threatens the validity of the violated value (Fiske & Tetlock, 1997; Hechler et al., 2016; Mendoza et al., 2014; Okimoto & Wenzel, 2010). Respectively, an offense is seen as a value violation when the offender acts intentionally, namely, when the offender intends to produce a negative outcome for a victim in an act that violates values that are known to the actor (Cushman, 2008; Gray et al., 2012; Mikhail, 2007). Actions that are beyond the control of an individual, by contrast, are seen as accidents rather than value violations. While accidents can also produce negative outcomes (e.g., a car accident with fatalities), they are not considered value violations.
With punishment, people express their disapproval and moral outrage and thereby distance themselves from attempted or actual offenses and offenders (Eidelman & Biernat, 2003; Feinberg, 1965). For example, the punishment of ingroup offenders promotes a positive group identity (Hutchison et al., 2008) and increases the punisher’s moral standing (Hofmann et al., 2018). This is particularly true when punishers assume that offenders know the values, for example, when they share a common identity (Pinto et al., 2010; Shinada et al., 2004). Mitigating factors are the offender being a new member in the group (e.g., not knowing the values) or ambiguous intentions, which elicit less desire to punish (Otten & Gordijn, 2014; van Prooijen, 2006).
For third parties (at least those unrelated to the victims), the perceived value violation is the primary cause for punitive tendencies (Okimoto & Wenzel, 2010; Tyler & Boeckmann, 1997). Accordingly, unambiguous malicious intentions determine moral condemnation, outrage, and punishment in third parties (Gray et al., 2012; Mikhail, 2007; van Prooijen, 2006). This holds for successful actions that produce negative outcomes (e.g., murder) as well as for unsuccessful acts that do not produce such outcomes (e.g., attempted murder; Cushman, 2008; Hechler & Kessler, 2018). In contrast, accidental or non-intentional negative outcomes do not elicit moral condemnation, outrage, or punishment in third parties. Consequently, third parties punish offenders who chose a less equal outcome over a more equal outcome, in contrast to those who chose the more equal distribution, irrespective of their net outcome (Gummerum & Chu, 2014).
Victims also often share a group identity with offenders or believe that offenders should appraise the same values. Accordingly, they punish those they expect to adhere to but intentionally violate values. For example, they reject unfair ingroup resource offers more often than unfair outgroup offers (Mendoza et al., 2014). Also, victims punish identical offers more if the offender intentionally chooses the more unequal outcome over the less unequal outcome (Falk et al., 2003; Gummerum & Chu, 2014). Similarly, the same unequal distribution elicits more punishment when offered by a person (i.e., intentional offender) than by a random device (Blount, 1995; Falk et al., 2008).
Outcome Equality, Status Imbalance, and Punishment
In addition to the threat to values through intentional violations, offenses produce a status imbalance between offenders and the victims (Wenzel et al., 2008). An offense reveals that the victim has less control over the situation compared to the offender (SimanTov-Nachlieli et al., 2013) and may consequently be harmed or left with fewer resources. Note, such status imbalance can also emerge without explicit intentions from the offender, as research on relative deprivation indicates (e.g., Kessler & Mummendey, 2002; Smith et al., 2012).
Status imbalance can elicit willingness to harm or to punish the advantaged, particularly in competitive relations, and when norms of fairness and cooperation are absent (Fehr & Schmidt, 1999; Raihani & Bshary, 2019). When cooperative norms are absent, individuals are even willing to harm cooperating others (so-called antisocial punishment; Herrmann et al., 2008). Status imbalances increase the salience of personal identity and interpersonal differentiation for victims relative to a common social identity, which enhances antagonism and harming of the advantaged (Skitka, 2003; Turner et al., 1994). In intergroup relations, groups perceive an attack by an outgroup rather as a status threat than a value violation (Okimoto & Wenzel, 2010). Thus, victims are likely to perceive their relationship with offenders as antagonistic and competitive.
In competitive relations, being disadvantaged can elicit willingness to harm or to punish the advantaged even if a value violation by an offender is absent (Fehr & Schmidt, 1999; for a review, see Raihani & Bshary, 2019). Victims reduce the offender’s payoff because of the relative inequality between them and the offender rather than the absolute losses they suffered (Raihani & McAuliffe, 2012). Also, victims who obtain negatively interdependent relationships with offenders punish inequality more than independent victims (Marczyk, 2017). However, victims do not necessarily aim at reducing outcome inequality between them and the offender. They punish offenders or reject unequal offers even when it has no effect on the inequality (Bone & Raihani, 2015; Falk et al., 2005). Thus, empirical evidence shows that victims are often motived to punish inequality.
Even though third parties primarily punish offenders for their malicious intentions and not accidents, they sometimes also punish mere status imbalances. This action on behalf of the victims is found when third parties care for them, empathize, or share a common identity with the victims (Batson et al., 2009; Bernhard et al., 2006; Gromet et al., 2012; Lieberman & Linke, 2007; Pfattheicher et al., 2019). Such actions may include punishment of the advantaged as well as compensation of the victims. Although it is not clear whether punishment or compensation dominates (Chavez & Bicchieri, 2013; van Prooijen, 2010), studies have shown that third parties with unspecific anger tend to punish, whereas those primed with empathic anger prefer compensation (Gummerum et al., 2016). Thus, third parties are less concerned about status imbalances between offenders and victims. Only when they have a positive relation to the victims, they act on their behalf.
The Current Study
Hypothesis, Aims, and Objectives
Most offenses imply intentional value violations and produce status imbalance at the same time. These two facets, unfair intention and outcome equality, are differentially important to third parties and victims. Therefore, they must be disentangled carefully to test their specific effects on punishment in a full experimental design. According to the value/status approach, third parties primarily punish value violations that imply offender’s unfair intention, whereas victims also punish mere status imbalance such as outcome inequality. In the present study, participants engaged in an experimental game in which “offenders” decide to distribute resources unequally to their own advantage (unfair intention) or equally between them and the “victims” (fair intention). Independently of the offenders’ intention, a random device puts either an equal or an unequal outcome into effect. Participants take the role of third parties or victims and invest their own resources to punish the offenders.
The present study aims at replicating findings suggesting that third parties punish malicious intentions but not unintentional negative outcomes (Cushman, 2008; Gummerum & Chu, 2014; Hechler & Kessler, 2018). Moreover, it aims at replicating prior studies investigating how victims punish offenders who intentionally distributed resources unfairly or fairly and offenders who had limited influence on the outcome equality (Falk et al., 2008; Gummerum & Chu, 2014).
Specifically, we propose that (H1) offender’s intention (unfair vs. fair) elicits punishment in both third parties and victims, whereas (H2) actual outcome (inequality vs. equality) elicits more punishment in victims than in third parties. Hence, we expect a main effect of offender’s intentions on punishment, and a two-way interaction of punisher’s perspective (victim vs. third party) and outcome inequality on punishment, but no two-way interaction of perspective and intentions on punishment. These hypotheses are presented in Figure 1.
Method
Design
The following experiment orthogonally manipulates offender intention, outcome, and participants’ perspective and measures punishment of the offender as the dependent variable. Accordingly, the experiment employs a 2 × 2 × 2 factorial design with intention (unfair/fair) and outcome (unequal/equal) as within-subject factors and perspective (third party/victim) as a between-subject factor.
Materials
The study was preregistered on PsychArchives (http://dx.doi.org/10.23668/psycharchives.4273) as part of this topical issue. Data collection was conducted online via Amazon Mechanical Turk (MTurk) by the ZPID – Leibniz Institute for Psychology, Trier, Germany, in December 2020. Participants received compensation of 1$ plus individual incentives from the experimental game ranging from 1.25$ to 2.75$ (see Procedure). Participation in the study took M = 10.17 min (SD = 3.44). The study was programmed on the online platform Soscisurvey (Leiner, 2018), which participants entered using a link provided in the invitation on MTurk. Each participant was randomly assigned to one perspective condition (third party or victim) using a full randomization procedure. Instructions in each condition merely differed according to perspective.
Ethical approval was obtained from the local ethic-committee at the University of Jena and informed consent of participants. All study materials, measures, (fully anonymized) datasets, and analysis R-scripts are accessible on Open Science Framework (https://osf.io/z495t/).
Sample Size and Power
Our hypotheses were to be tested using a three-way factorial ANOVA. Accordingly, we calculated the sample size needed to detect main effects (H1) and two-way interactions (H2) ANOVA with high (95%) power, using G*Power (Faul et al., 2007). The desired effect size was estimated based on Gummerum and Chu (2014). They investigated victims and third-party punishment for an offender’s decision to implement the more unequal outcome (e.g., choice between 5/5 and 8/2) or the less equal outcome (e.g., choice between 10/0 and 8/2) to the victim’s disadvantage. In adults, they found a small effect of the decision on victims (η2 = .08), as well as on third parties (η2 = .06) using a 1 × 4 design for each group. We use this as the expected effect size to approach H1. Power analysis with η2 = .06 indicated that we would need a total sample size of 53. No prior studies – to our knowledge – directly compared how much victims and third parties punish inequality. Marczyk (2017) shows that participants punished a decision for inequality more when it was at the expense of the participant (interdependence, similar to victim perspective; 75%) than when the outcomes of the players were independent (similar to third party perspective; 18%). A decision for an equal outcome was still punished more in the interdependent (47%) than the independent condition (14%). Different from our study, the dependent variable was a dichotomous choice and the interdependent conditions involved unfair intention, whereas the independent did not. Nevertheless, based on these considerations, we assume a small effect of the perspective × outcome interaction on punishment (H2; f = .10, η2 = .01). This also provides the opportunity to detect small effects of other two-way interactions. We calculated the a priori sample size using a mixed ANOVA with two groups and two measurements. The data of 328 participants allows to detect a small effect with a power of 1 − β = .95 and at a significance level of 5%. We preregistered to exclude participants who would fail an attention check (see Procedure for a detailed description). MTurkers have been found to fail lure questions with a probability of 7% when the question is raised in the beginning (see Hauser & Schwarz, 2016, Study 1). Thus, we aimed at collecting data of 353 participants to compensate for the exclusion of participants.
Participant Characteristics
The ZPID recruited 353 US-American participants via MTurk. In the preregistration, we stated that the sample would consist of participants from the ZPID’s PsychLab participant pool who speak German as their native language. Due to the differential incentives based on the study outcome, the ZPID collected the data via MTurk, which allows for incentivizing participants individually. The MTurk participant pool mainly consists of English-speaking US-American participants. Therefore, the study materials were translated into English while closely following the preregistered instructions. Three hundred forty-six participants indicated that English was their native language. All other participants characteristics conformed with the preregistration. Participants took part in the study anonymously for a participation fee of 1$ plus their incentives based on the experimental interactions. 14.20% of the current sample failed to answer the attention check correctly and were thus excluded from further analyses.
The final sample consisted of 302 participants. Half of them were female (151; two indicated their gender as diverse), and their mean age was M = 39.26 years (SD = 10.81, range = 18–74). Most of the participants were employees (206) or freelancers (53). The phenomenon of costly punishment has been observed across various populations (e.g., Henrich et al., 2006). Thus, heterogeneity within the sample should not affect the results.
Procedure
Participants were invited via e-mail to participate in an online study on resource allocation behavior. Upon clicking on a link, participants read that they will interact with others in allocation tasks. After signing an informed consent form, they read the task instructions. In the following interactions, they were able to collect monetary units (MUs). Each MU had a value of 5 cents. Participants received their total payoff at the end of the study.
They further read that they would be grouped together with allegedly two other participants in each interaction. Participants were informed of three possible roles in the game: the decision-maker – Person A, the recipient (i.e., victim) – Person B, and the third party (i.e., observer) – Person C. In actuality, participants were randomly assigned to one of two roles, the recipient (Person B) or the third party (Person C) and maintained it throughout the experiment. In each round, 10 monetary units (MU) were distributed between person A and B. Person C received 5 MUs. Person A decided how many of the 10 MUs she would keep and how many she would give to Person B. Simultaneously, a computerized random device distributed the 10 MUs between Person A and Person B. A dice roll put one of the distributions into place: Person A’s decision or the random distribution. Person A and Person B received the MUs based on the dice roll (see Figure 2). Person C observed the interaction. Thus, the decision-maker suggested a distribution on which the victim had no influence. The outcomes of the decision-maker and the victim were negatively interdependent, whereas the third party’s outcome was independent of the others’.
Participants engaged in various rounds with different persons in one-shot games. We explicitly informed them that all persons stay anonymous at all times and that sometimes interferences could alter the distribution of the MUs. We further stated that no one was aware of their total gain until all tasks were finished and that interferences with their outcomes were not described to any participant during the experiment. This procedure ensured that the punishment (i.e., reducing MUs from Person A) is not used to communicate with the other participants since we clearly stated that Person A would not be aware of the punishment (see Crockett et al., 2014; Nadelhoffer et al., 2013).
Three test trials assured that participants understood the game and simultaneously measured participants’ allocation preference (see Appendix for detailed description). After completing the test trials, participants answered the lure question as an attention check: “Before the experiment starts, you and the other players do not own any MUs. The MUs are only distributed during the game. As an attention check, please indicate that you own 7 MUs in the following question, even though you actually own 0 MUs at the moment.” followed by the question on the next page: “How many MUs do you own currently in total?” Then participants indicated how much they were concerned about Person B’s outcome as a moderator variable in the secondary analyses.
Subsequently, participants engaged in the interaction trials, in which they observed Person A’s decision, the random distribution, and the implementation of one distribution (see Figure 2). In the first five trials, Person A decided to distribute 3–5 MUs to Person B, and Person A and Person B received a similar amount of MUs, underlining equality as fairness as a common value (Camerer, 2003; Engel, 2011). After these five trials, the critical trials started. These were second- or third-party punishment games (Fehr & Fischbacher, 2004). The participants were informed that they now had the opportunity to reduce points from Person A by investing in their own MUs. Each 1 MU invested reduced 3 MUs from Person A. Participants in the victim perspective received an additional 5 MUs to reduce points from Person A or to keep for themselves. Participants in the third-party perspective could invest the 5 MUs they received at the beginning of each interaction. Third parties and victims learned that no one was aware of their option to punish. The following four interactions were presented in random order.
After each critical trial, we assessed punishment as the investment of the player’s MUs for reducing Person A’s MUs. After the four critical trials, we measured the manipulation checks, as described below. Finally, participants answered demographic questions, including gender, age, occupation, and native language. At the end of the study, participants were debriefed and received their incentives shortly after.
Variables
Participants assumed either the perspective of the victim or the perspective of a third party while completing the study. Participants were informed that Player A decided for a distribution (implying fair or unfair intentions), and then either Player A’s decision or a computerized random distribution would be implemented with equal probability (implying equal or unequal outcome). This allows crossing the factor’s intention and outcome orthogonally. The within-subject factor intention of Player A was either fair or unfair intentions, and the factor outcome was either equal or unequal. The four within-participant conditions were represented in four critical trials in random order:
Unfair Intention/Unequal Outcome
Person A distributes 10 MUs to herself and 0 MUs to Person B, the random distribution distributes 10 to Person A and 0 MUs to Person B. The decision of Person A is implemented;
Unfair Intention/Equal Outcome
Person A distributes 10 MUs to herself and 0 MUs to Person B, the random distribution distributes 5 to Person A and 5 MUs to Person B. The random distribution is implemented;
Fair Intention/Unequal Outcome
Person A distributes 5 MUs to herself and 5 MUs to Person B, the random distribution distributes 10 to Person A and 0 MUs to Person B. The random distribution is implemented (see example illustrated in Figure 2);
Fair Intention/Equal Outcome
Person A distributes 5 MUs to herself and 5 MUs to Person B, the random distribution distributes 5 to Person A and 5 MUs to Person B. The decision of Person A is implemented.
The main dependent variable was punishment, operationalized as the investment of own monetary units to reduce the outcome of the decision-maker on a 6-point scale (0–5 MUs; 1 MU investment equals 3 MUs reduction).
We also assessed participants’ preference for equality and concern for the victims’ outcome (4 items, α = .89) as potential moderators before the critical trials. In the end, we again presented the four critical trials to assess manipulation checks for the perception of Person A’s intentionality, perceived equality of the distribution, and perceived competition with Person A. The measurements of all variables are included in the Appendix.
Results
Pre-Processing
Pre-processing and data analysis were conducted in R (R Core Team, 2017). Data pre-processing and analyses were performed as indicated in the registered report.
We excluded participants from data analysis who failed the attention check (see Procedure for detailed description). Then, we recoded the moderator variable preference for equality on a scale from 0 to 5 according to the deviations from equality, such that 0 represents the preference for inequality and 5 is the preference for equality.
Manipulation Checks
Figure 3 displays means and confidence intervals of the three manipulation checks (i.e., perceived intentions, perceived outcome, and perceived competition) in each condition. ANOVA results are displayed and described in detail in the Appendix.
First, we tested how intention, outcome, perspective, and their interactions affect the perceived intentionality of the offender’s decision. The goal of this analysis was to examine whether they assign responsibility to Person A for their decisions and not for the fairness of their intentions or the actual outcome.1 Therefore, we asked participants how much they thought that Person A is responsible for their decision as preregistered in the Appendix of the registered report. In the preregistered test section, however, we stated that “… we test whether unfair intentions of the offender are perceived as more unfair than fair intentions.” While preparing the study we conformed to the preregistered measurement in the Appendix. There were no effects on the perceived intentionality. Thus, across conditions, participants perceived the decision-maker to be responsible for their own decision.
Second, we were interested in how equal participants perceived the actual distribution. Participants perceived the equal outcome to be more equal than the unequal outcome. Perceived equality was also higher when intentions were fair compared to unfair, and the outcome was equal compared to unequal.
Third, we examined the effects of intention, outcome, and perspective on participants’ perceived competition with the decision-maker. Victims reported experiencing more competition with the decision-maker compared to third parties. This was qualified by a three-way interaction. In the victim perspective, the perceived competition was higher with unfair intentions compared to fair intentions and unequal outcomes compared to equal outcomes.
Overall, as intended, participants perceived the offender’s decision to depict their intentions to a similar degree in each condition. They perceived the equal outcomes to be more equal than the unequal outcomes. Finally, victims felt more competition with the decision-maker than third parties, especially when the outcome was unequal and/or the decision-maker had unfair intentions.
Main Analysis
To test our main hypotheses, we conducted a three-way mixed-design ANOVA on punishment. It revealed the main effects of all three independent variables on punishment but no interaction effects (see Table 1). Specifically, unfair intentions (M = 0.90, SD = 1.47) elicited more punishment than fair intentions (M = 0.71, SD = 1.34), outcome inequality (M = 1.04, SD = 1.52) elicited more punishment than outcome equality (M = 0.58, SD = 1.25), and victims (M = 1.05, SD = 1.60) punished more than third parties (M = 0.58, SD = 1.16; see also Figure 4). Since there were no interaction effects on punishment, we refrained from follow-up analyses. The results support H1, victims and third parties similarly punish malicious intentions. Although victims were punished more than third parties, this was not qualified by outcome equality, which does not support H2.
Secondary Analyses
As secondary analyses, we examined the additional influence of concern for the victim (M = 4.47, SD = 1.37) and preference for equality (M = 3.94, SD = 1.79) on punishment in two separate analyses. There was a significant correlation between concern for the victim and preference for equality, t(300) = 4.88, p < .001, r = .27, 95% CI [.16; .37].
Since we included a measured moderator in each analysis, we calculated mixed linear models on punishment (mean-centered), as preregistered. The predictors were intention (0 = fair/1 = unfair), outcome (0 = equal/1 = unequal), perspective (0 = third party/1 = victim), concern for the victim/preference for equality (mean-centered), and all two- and three-way interactions. Descriptive statistics and results of the mixed models are displayed in the Appendix available from Hechler and Kessler (2021) at PsychArchives at https://dx.doi.org/10.23668/psycharchives.5055.
Using concern for the victim as an additional predictor, we expected a two-way interaction indicating that more concern for the victim and outcome inequality compared to equality increases punishment in both, third parties and victims. The mixed model showed that outcome inequality was punished more than outcome equality, t(298) = 3.69, p = < .001, b = 0.37, 95% CI [0.17; 0.24], and victims punished more than third parties, t(298) = 4.11, p = < .001, b = 0.59, 95% CI [0.31; 0.88]. However, there was no significant effect of intention, t(298) = 1.43, p = .153, b = 0.10, 95% CI [−0.04; 0.24], and no significant effect of victim concern on punishment, t(298) = 1.09, p = .276, b = 0.07, 95% CI [−0.20; 0.06]. Concern for the victim moderated the effect of perspective on punishment, t(298) = −2.48, p = .014, b = −0.27, 95% CI [−0.49; −0.06]. Victims who were more concerned about their outcome punished less, whereas concern for the victim’s outcome does not affect third party punishment to a similar degree (see Figure 5). There was no interaction effect of outcome and concern for the victim, t(894) = 0.25, p = .779, b = 0.02, 95% CI [−0.11; 0.15]. The model did not show any further effects on punishment.
Using preference for equality as an additional predictor, we expected a two-way interaction indicating that a higher preference for equality and unfair intentions compared to fair intentions increases punishment in third parties and victims. People with a higher preference for equality should perceive unfair intentions as more malicious than people with a lower preference for equality. The analysis show that outcome inequality was punished more than outcome equality, t(298) = 3.79, p < .001, b = 0.37, 95% CI [0.18; 0.55], and victims punished more than third parties, t(298) = 2.90, p < .001, b = 0.40, 95% CI [0.13; 0.67]. The mixed model showed no significant effect of intention on punishment and no main effect of preference for equality, t(298) ≥ 1.44, p ≤ .152, b ≤ 0.10. In contrast to our assumption, there was no interaction effect of intention and preference for equality on punishment, t(894) = −0.04, p = .971, b = −0.001, 95% CI [−0.07; 0.07]. There were no further effects. This indicates that preference for equality increases the punishment of unfair intentions with outcome inequality, which accounts more for victims than third parties.
In conclusion, concern for victims moderated the effect of perspective on punishment. Victims who were more concerned about their outcome punished less than victims who were less concerned about their outcome. In contrast, concern for the victim’s outcome did not affect third-party punishment. Moreover, preference for equality did not moderate any of the effects on the punishment that were found in the main analysis.
Discussion
The current study tested whether two conceptually different facets of an offense – intention, and outcome – elicit differential punishment by victims and third parties. Therefore, we conducted an experimental game that disentangled the offender’s intentions from the actual outcome of the interaction and measured punishment by victims and by third parties. These two facets are both relevant for the severity of an offense, which has been shown to determine the deservingness of punishment (e.g., Carlsmith & Darley, 2008; Gromet & Darley, 2009b). Participants either obtained the role of the victim or a third-party observer. In order to punish, participants invested their own resources to reduce the offenders’ payoffs.
In line with our first hypothesis, unfair intentions lead to enhanced punishment in victims and third parties. In contrast to our second hypothesis, equality of outcome crucially enhanced punishment by both, victims and third parties. Moreover, victims generally punished harsher than third parties. However, there were no significant interaction effects. Even though the final sample comprised only 302 participants instead of a priori planned 328 participants, interaction effects were negligible. The effect sizes of the interactions were smaller than ηp2 = .004 and their confidence intervals included zero.
Our findings support the notion that the two facets of an offense are conceptually different, as unfair intentions and unequal outcomes elicited punishment independently of one another. The two facets imply different threats to others: unfair intentions threaten the validity of common values, whereas unequal outcome threatens the status balance between offender and victim (Wenzel et al., 2008). The study was designed to replicate findings showing that malicious intentions and outcome inequality determine retribution. Prior studies had shown that victims punish offenders who intentionally distributed unfairly (Falk et al., 2008; Gummerum & Chu, 2014) and that they punish the advantaged, even when they did not produce the inequality (Batson et al., 2000; Bone & Raihani, 2015; Marczyk, 2017). Other findings indicated that the actual outcome is less relevant for the punishment of third parties than unfair intentions (Cushman, 2008; Gummerum & Chu, 2014; Hechler & Kessler, 2018). This was not the case in the current study. The results show that outcome inequality elicited harsher punishment than unfair intentions. Even though we suggested that victims punish inequality more than third parties, we did not observe such a difference. The strong effect of inequality may be due to the within-subject design. Participants experienced equal as well as unequal outcomes and could thus compare both directly. Only extremely unequal outcomes (0/10) were presented in the critical trials. This may have turned the unequal outcome into something unexpected and thus frustrating and may have affected the punishment reaction (Aina et al., 2020).
The present study extends findings on how punishment by victims and third-party punishment differ. The results give no indication for differential motives to punish by victims and observers. Victims generally punished substantially harsher than third parties. Victims have been found to be more likely to punish and punish harsher than third parties (Crockett et al., 2014). The divergent perspectives can lead to different evaluations of offenses and possibly different amounts of punishment (Mummendey et al., 1984). This may cause mutual dissatisfaction with punishment. Victims may evaluate third-party punishment as too lenient, whereas third parties may evaluate punishment of victims as too harsh punishers.
Additionally, we tested how the concern for the victim’s outcome and a general preference for equality affect the effects of intention and outcome on punishment. Concern for their own outcome decreased punishment by victims, but concern for the victim’s outcome had no significant effect on third-party punishment. In other words, victims decreased their own incentives by investing in punishment when they were not concerned much about their payoff. Other than expected, preference for equality did not moderate any effects on punishment.
Retribution in Victims and Third Parties
Prior studies on retribution suggest that the severity of the deed, revealed by the maliciousness of intentions and by the harm done (i.e., outcome inequality), is a crucial determinant of punishment (e.g., Carlsmith et al., 2002; Gromet & Darley, 2009b). The current results support this notion, as participants punished mere unfair intentions. This is also in line with Kant’s idea of punishment for “internal wickedness” (Kant, 1785/2007). In contrast to this idea, victims and third parties also punished the advantaged irrespective of their malicious intentions. Although an offense is judged more severely when it involves malicious intentions (Ames & Fiske, 2013; Darley & Huff, 1990), there was no interaction effect of intention and outcome on punishment. Nevertheless, participants evaluated the unequal distributions as more unequal when they were produced intentionally compared to accidentally.
Moreover, even randomly produced inequality between offender and victim elicited punishment in victims and third parties. In the present study, the actual distribution of MUs between offender and victim became less unequal when punishers reduced points from the advantaged. Even though the punishment did not positively affect the victims’ outcome, 57% of the participants in the victim role and 44% in the observer role invested at least 1MU to punish in the current study. This might have resulted from negative emotions produced by malicious intentions or outcome inequality (e.g., see Hechler & Kessler, 2018; Montada & Schneider, 1989; Nelissen & Zeelenberg, 2009).
Recent studies suggest that people punish to teach a lesson to offenders, thus obtaining utilitarian goals next to retributivist goals (Funk et al., 2014; Sarin et al., 2020). Others showed that victims and third parties even punish secretly (Crockett et al., 2014; Nadelhoffer et al., 2013). That is, they even punish when the offender was not informed about the punishment but only received unspecified damage. In the current study, participants were informed that offenders would not know about the punishment. All participants only would see their final resources at the end of the study. Supposedly, an additional utilitarian consequence, such as teaching a lesson, might have led to even more punishment for malicious intentions. However, it cannot explain the current effect of intention on punishment.
Intention and Punishment by Victims and Third Parties
The current study carefully disentangled the facets “intention” and “outcome” of an offense. The findings show that malicious intentions were punished even when offender and victim coincidentally received equal outcomes. Such intentions have been suggested to crucially determine moral judgments of an offense and willingness to punish (Gray et al., 2012; Mikhail, 2007; van Prooijen, 2006). In line with the current findings, prior studies demonstrated that unfair intentions trigger punishment in victims and third parties (Blount, 1995; Falk et al., 2003, 2008; Gummerum & Chu, 2014). Whereas all of these studies used the strategy method to assess punishment (i.e., participants react to all hypothetical outcomes), we employed various trials of incentivized one-shot games (i.e., four critical trials, whose outcome directly affected the total outcome of participants). Either method leads participants to punish malicious intentions. This is important for victims and observers, as they feel moral outrage (Hechler & Kessler, 2018), psychologically distance themselves from offenders and their decisions (e.g., Eidelman & Biernat, 2003; Feinberg, 1965), and demonstrate their personal and their group’s moral standards (e.g., Hofmann et al., 2018; Hutchison et al., 2008).
Offenders are held responsible for a negative outcome when they intentionally produce it and are consequently punished (Cushman, 2008; Gray et al., 2012; Mikhail, 2007). In contrast, unintentional negative outcomes are seen as accidents. In the current study, third parties punished unequal outcomes, even though it could not be attributed to the offender’s decision. This indicates that, besides malicious intentions, other motives may drive “punishment” of the advantaged.
Outcome Equality and Punishment by Victims and Third Parties
We suggested that victims would punish outcome inequality more than third parties since they perceive the relatively advantaged as antagonists (e.g., Okimoto & Wenzel, 2010, Turner et al., 1994). In competitive relationships, people harm others even when they do not violate norms (e.g., Fehr & Schmidt, 1999; Raihani & Bshary, 2019). The current study victims perceived their relation to the decision-maker as more competitive than third parties did. Accordingly, victims were willing to invest their own points in order to punish the offender. Like in prior studies, the competition was characterized by a negatively interdependent relationship. Such competitive relations have been found to predict punishment better than focusing on losses without competition (Marczyk, 2017; Raihani & McAuliffe, 2012).
Previous studies also found that victims reduce the outcome of advantaged others irrespective of their responsibility for the inequality (Batson et al., 2000). In this study, the offender first took MUs from participants. Then participants were able to reallocate the outcome between themselves and the other. In contrast, participants in the current study could only gain incentives (but not lose any) and invest their own outcome in reducing others’ outcomes. Nevertheless, both studies indicate that participants punished advantaged even when they did not produce the inequality.
Third parties perceived much less competition with the offenders in the current study. However, they also punished outcome inequality, even though their investment did not benefit the victims’ outcome. In fact, the main effect of the outcome on punishment was even stronger than the main effects of perspective and intention. Several mechanisms might have contributed to third-party punishment of accidental inequality. First, empathy might have triggered third-party punishment since there was no other option to compensate the victim (e.g., Chavez & Bicchieri, 2013; Pfattheicher et al., 2019). In the current study, there was no effect of concern for the victim’s outcome on third-party punishment. Nevertheless, third parties may have related to disadvantaged victims, which may have elicited anger and subsequently punishment (Batson et al., 2009; Gummerum et al., 2016).
Second, outcome inequality may have been perceived as a violation of (descriptive) norms or inequality aversion (e.g., Blanz et al., 1997; Fehr & Schmidt, 1999). As mentioned before, the disruption of norms is an important trigger of punishment (Wenzel et al., 2008; Tyler & Boeckmann, 1997). Accordingly, victims reduce the outcome of advantaged others to establish distributive justice (Batson et al., 2000). In the current study, personal preference for equality could be seen as a proxy of whether an equal distribution would be seen as just. However, it did not moderate punishment of unequal versus equal outcomes for third parties in the current study.
Third, as mentioned before, the repeated trials of the within-subject design of the present study allowed for a direct comparison of the outcomes. The experience of equal outcomes may elicit frustration about the unequal outcome (Aina et al., 2020). In other studies, participants accepted distributions when payoff disparities were randomly produced compared to intentionally (Blount, 1995). In the current study, the various trials may have established the expectation that outcomes would be equal. Reducing points from the advantaged decreased the inequality between victim and decision-maker – on the cost of the punisher and the punished. Future studies may test whether the strong effect of outcome inequality on punishment also emerges in between-subject designs or studies employing a strategy method.
Conclusion
Retributive punishment implies that punishment is proportional to the severity of the deed (e.g., Carlsmith & Darley, 2008; Vidmar, 2001). The severity of an offense includes two threats: intentional violations threaten values, and victim’s disadvantage compared to the offender threatening the victim’s status (Wenzel et al., 2008). These conceptually different facets trigger punishment among both victims and observers. However, victims as the affected party are punished more harshly than third parties. Assuming that retribution seeks to re-establish a moral balance (e.g., Bies & Tripp, 1996; Carlsmith & Darley, 2008), the current findings suggest that malicious intentions, as well as social inequality, disrupt this balance. Thus, victims and third parties follow similar intuitions about what to punish but differ substantially in how much.
The research reported in this article was preregistered at https://dx.doi.org/10.23668/psycharchives.4273. We deviated from our preregistration in our data collection by choosing a US sample instead of a German sample. We explain and justify this deviation transparently in our methods section.
We thank Julia Elad-Strenger, Dana Schneider, and two anonymous reviewers for comments on an earlier draft.
1In the preregistration, the variable that measured Person A’s “perceived intentionality” was defined as: “How much do you think that Person A is responsible for her/his decision?” (see preregistered Appendix). In the preregistered test section, however, we stated that “… we test whether unfair intentions of the offender are perceived as more unfair than fair intentions.” While preparing the study we conformed to the preregistered measurement in the Appendix. Thus, the variable measured to what extend Person A was perceived as responsible for their decision irrespective of the outcome, and not whether unfair intentions were perceived as more unfair than fair intentions.
References
(2020). Frustration and anger in the Ultimatum Game: An experiment. Games and Economic Behavior, 122, 150–167. https://doi.org/10.1016/j.geb.2020.04.006
(2013). Intentional harms are worse, even when they’re not. Psychological Science, 24, 1755–1762. https://doi.org/10.1177/0956797613480507
(2000). Does personal morality exacerbate or restrain retaliation after being harmed? Personality and Social Psychology Bulletin, 26, 35–45. https://doi.org/10.1177/0146167200261004
(2009). Pursuing moral outrage: Anger at torture. Journal of Experimental Social Psychology, 45, 155–160. https://doi.org/10.1016/j.jesp.2008.07.017
(2006). Parochial altruism in humans. Nature, 442, 912–915. https://doi.org/10.1038/nature04981
(1996).
Beyond distrust: “Getting even” and the need for revenge . In R. M. KramerT. R. TylerEds., Trust in organizations: Frontiers of theory and research (pp. 246–260). Sage.(1997). Normative evaluations and frequency expectations regarding positive versus negative outcome allocations between groups. European Journal of Social Psychology, 27, 165–176. https://doi.org/10.1002/(SICI)1099-0992(199703)27:2<165::AID-EJSP812>3.0.CO;2-3
(1995). When social outcomes aren’t fair: The effect of causal attributions on preferences. Organizational Behavior and Human Decision Processes, 63, 131–144. https://doi.org/10.1006/obhd.1995.1068
(2015). Human punishment is motivated by both a desire for revenge and a desire for equality. Evolution and Human Behavior, 36, 323–330. https://doi.org/10.1016/j.evolhumbehav.2015.02.002
(2003). Strategizing in the brain. Science, 300, 1673. https://doi.org/10.1126/science.1086215
(2008).
Psychological aspects of retributive justice . In M. P. ZannaEd., Advances in experimental social psychology (Vol. 40) (pp. 193–236). Elsevier Academic Press.(2002). Why do we punish? Deterrence and just deserts as motives for punishment. Journal of Personality and Social Psychology, 83, 284–299. https://doi.org/10.1037/0022-3514.83.2.284
(2013). Third-party sanctioning and compensation behavior: Findings from the ultimatum game. Journal of Economic Psychology, 39, 268–277. https://doi.org/10.1016/j.joep.2013.09.004
(1979). Varieties of retribution. Philosophical Quarterly, 29, 238–246. https://doi.org/10.2307/2218820
(2014). The value of vengeance and the demand for deterrence. Journal of Experimental Psychology: General, 143, 2279–2286. https://doi.org/10.1037/xge0000018, https://doi.org/10.1037/xge0000018.supp (Supplemental)
(2008). Crime and punishment: Distinguishing the roles of causal and intentional analyses in moral judgment. Cognition, 108, 353–380. https://doi.org/10.1016/j.cognition.2008.03.006
(1990). Heightened damage assessment as a result of the intentionality of the damage-causing act. British Journal of Social Psychology, 29, 181–188. https://doi.org/10.1111/j.2044-8309.1990.tb00898.x
(2003). The psychology of compensatory and retributive justice. Personality and Social Psychology Review, 7, 324–336. https://doi.org/10.1207/S15327957PSPR0704_05
(2003). Derogating black sheep: Individual or group protection? Journal of Experimental Social Psychology, 39, 602–609. https://doi.org/10.1016/S0022-1031(03)00042-8
(2011). Dictator games: A meta study. Experimental Economics, 14, 583–610. https://doi.org/10.1007/s10683-011-9283-7
(2003). On the nature of fair behavior. Economic Inquiry, 41, 20–26. https://doi.org/10.1093/ei/41.1.20
(2005). Driving forces behind informal sanctions. Econometrica, 73, 2017–2030. https://doi.org/10.1111/j.1468-0262.2005.00644.x
(2008). Testing theories of fairness – Intentions matter. Games and Economic Behavior, 62, 287–303. https://doi.org/10.1016/j.geb.2007.06.001
(2007). G*Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behavior Research Methods, 39, 175–191. https://doi.org/10.3758/BF03193146
(2004). Third-party punishment and social norms. Evolution and Human Behavior, 25, 63–87. https://doi.org/10.1016/S1090-5138(04)00005-4
(1999). A theory of fairness, competition, and cooperation. The Quarterly Journal of Economics, 114, 817–868. https://doi.org/10.1162/003355399556151
(1965). The expressive function of punishment. The Monist, 49, 397–423. https://doi.org/10.5840/monist196549326
(1997). Taboo trade-offs: Reactions to transactions that transgress the spheres of justice. Political Psychology, 18, 255–297. https://doi.org/10.1111/0162-895X.00058
(2014). Get the message: Punishment is satisfying if the transgressor responds to its communicative intent. Personality and Social Psychology Bulletin, 40, 986–997. https://doi.org/10.1177/0146167214533130
(2012). Mind perception is the essence of morality. Psychological Inquiry, 23, 101–124. https://doi.org/10.1080/1047840X.2012.651387
(2006). Restoration and retribution: How including retributive components affects the acceptability of restorative justice procedures. Social Justice Research, 19, 395–432. https://doi.org/10.1007/s11211-006-0023-7
(2009a). Punishment and beyond: Achieving justice through the satisfaction of multiple goals. Law & Society Review, 43, 1–38. https://doi.org/10.1111/j.1540-5893.2009.00365.x
(2009b). Retributive and restorative justice: Importance of crime severity and shared identity in people’s justice responses. Australian Journal of Psychology, 61, 50–57. https://doi.org/10.1080/00049530802607662
(2012). A victim-centered approach to justice? Victim satisfaction effects on third-party punishments. Law and Human Behavior, 36, 375–389. https://doi.org/10.1037/h0093922
(2014). Outcomes and intentions in children’s, adolescents’, and adults’ second- and third-party punishment behavior. Cognition, 133, 97–103. https://doi.org/10.1016/j.cognition.2014.06.001
(2016). Costly third-party interventions: The role of incidental anger and attention focus in punishment of the perpetrator and compensation of the victim. Journal of Experimental Social Psychology, 65, 94–104. https://doi.org/10.1016/j.jesp.2016.04.004
(2016). Attentive turkers: MTurk participants perform better on online attention checks than do subject pool participants. Behavior Research Methods, 48, 400–407. https://doi.org/10.3758/s13428-015-0578-z
(2018). On the difference between moral outrage and empathic anger: Anger about wrongful deeds or harmful consequences. Journal of Experimental Social Psychology, 76, 270–282. https://doi.org/10.1016/j.jesp.2018.03.005
(2021). Supplemental materials to “The importance of unfair intentions and outcome inequality for punishment by third parties and victims”. https://dx.doi.org/10.23668/psycharchives.5055
(2016). The infamous among us: Enhanced reputational memory for uncooperative ingroup members. Cognition, 157, 1–13. https://doi.org/10.1016/j.cognition.2016.08.001
(2006). Costly punishment across human societies. Science, 312, 1767–1770. https://doi.org/10.1126/science.1127333
(2008). Antisocial punishment across societies. Science, 319, 1362–1367. https://doi.org/10.1126/science.1153808
(2018). Moral punishment in everyday life. Personality and Social Psychology Bulletin, 44, 1697–1711. https://doi.org/10.1177/0146167218775075
(2008). Getting rid of the bad ones: The relationship between group identification, deviant derogation, and identity maintenance. Journal of Experimental Social Psychology, 44, 874–881. https://doi.org/10.1016/j.jesp.2007.09.001
(1785/2007). Grundlegung zur Metaphysik der Sitten
[Critique of Metaphysics] . Suhrkamp.(2002). Sequential or parallel processes? A longitudinal field study concerning determinants of identity-management strategies. Journal of Personality and Social Psychology, 82, 75–88. https://doi.org/10.1037/0022-3514.82.1.75
(2018). Sosci Survey (Version 2.5.00–i1142). https://www.soscisurvey.de/
(2007). The effect of social category on third party punishment. Evolutionary Psychology, 5, 289–305. https://doi.org/10.1177/147470490700500203
(1997). The folk concept of intentionality. Journal of Experimental Social Psychology, 33, 101–121. https://doi.org/10.1006/jesp.1996.1314
(2017). Human punishment is not primarily motivated by inequality. PLoS One, 12, 1–11. https://doi.org/10.1371/journal.pone.0171298
(2014). For members only: Ingroup punishment of fairness norm violations in the Ultimatum Game. Social Psychological and Personality Science, 5, 662–670. https://doi.org/10.1177/1948550614527115
(2007). Universal moral grammar: Theory, evidence and the future. Trends in Cognitive Sciences, 11, 143–152. https://doi.org/10.1016/j.tics.2006.12.007
(1989). Justice and emotional reactions to the disadvantaged. Social Justice Research, 3, 313–344. https://doi.org/10.1007/BF01048081
(1993). Justifying retributivism. Israel Law Review, 27, 15–49. https://doi.org/10.1017/S0021223700016836
(1984). Actor or victim of aggression: Divergent perspectives – divergent evaluations. European Journal of Social Psychology, 14, 297–311. https://doi.org/10.1002/ejsp.2420140305
(2013). Folk retributivism and the communication confound. Economics & Philosophy, 29, 235–261. https://doi.org/10.1017/S0266267113000217
(2009). Moral emotions as determinants of third-party punishment: Anger, guilt, and the functions of altruistic sanctions. Judgment and Decision Making, 4, 543–553. http://journal.sjdm.org/91001/jdm91001.pdf
(2010). The symbolic identity implications of inter and intra‐group transgressions. European Journal of Social Psychology, 40, 552–562. https://doi.org/10.1002/ejsp.704
(2014). Was it one of us? How people cope with misconduct by fellow ingroup members. Social and Personality Psychology Compass, 8, 165–177. https://doi.org/10.1111/spc3.12098
(2019). Compassion magnifies third-party punishment. Journal of Personality and Social Psychology, 117, 124–141. https://doi.org/10.1037/pspi0000165
(2010). Membership status and subjective group dynamics: Who triggers the black sheep effect? Journal of Personality and Social Psychology, 99, 107–119. https://doi.org/10.1037/a0018187
. (2017). R: A language and environment for statistical computing. https://www.R-project.org/
(2019). Punishment: one tool, many uses. Evolutionary Human Sciences, 1, Article e12. https://doi.org/10.1017/ehs.2019.12
(2012). Human punishment is motivated by inequity aversion, not a desire for reciprocity. Biology Letters, 8, 802–804. https://doi.org/10.1098/rsbl.2012.0470
(2020). Punishment is organized around principles of communicative inference. Cognition, 208, Article 104544. https://doi.org/10.1016/j.cognition.2020.104544
(2004). False friends are worse than bitter enemies: “Altruistic” punishment of in-group members. Evolution and Human Behavior, 25, 379–393.
(2013). Individuals’ and groups’ motivation to restore their impaired identity dimensions following conflicts: Evidence and implications. Social Psychology, 44, 129–137. https://doi.org/10.1027/1864-9335/a000148
(2003). Of different minds: an accessible identity model of justice reasoning. Personality and Social Psychology Review, 7, 286–297. https://doi.org/10.1207/S15327957PSPR0704_02
(2012). Relative deprivation: A theoretical and meta-analytic review. Personality and Social Psychology Review, 16, 203–232. https://doi.org/10.1177/1088868311430825
(1994). Self and collective: Cognition and social context. Personality and Social Psychology Bulletin, 20, 454–463. https://doi.org/10.1177/0146167294205002
(1997). Three strikes and you are out, but why? The psychology of public support for punishing rule breakers. Law & Society Review, 31, 237–265. https://doi.org/10.2307/3053926
(2006). Retributive Reactions to Suspected Offenders: The Importance of Social Categorizations and Guilt Probability. Personality and Social Psychology Bulletin, 32, 715–726. https://doi.org/10.1177/0146167205284964
(2010). Retributive versus compensatory justice: Observers preference for punishing in response to criminal offenses. European Journal of Social Psychology, 40, 72–85. https://doi.org/10.1002/ejsp.611
(2001).
Retribution and revenge . In J. H. SandersV. LeeEds., Handbook of justice research in law (pp. 31–63). Kluwer Academic.(1999). Even more varieties of retribution. Philosophy, 74, 595–605. https://doi.org/10.1017/s0031819199000704
(2008). Retributive and restorative justice. Law and Human Behavior, 32, 375–389. https://doi.org/10.1007/s10979-007-9116-6
