The Effects of Culture on Active and Passive Reactions to Financial Offenders; the Moderating Role of Type of Harm

ABSTRACT

In the present research, we examine how culture influences individuals’ reactions to financial offenders. We hypothesized that horizontal individualists deploy increased active reactions (i.e., punishment-oriented) whereas vertical collectivists deploy increased passive reactions (i.e., condemning beliefs) to financial offenders. Moreover, we hypothesized that horizontal individualists would react stronger to a financial offender when an offense has instrumental (i.e., related to material costs) as opposed to symbolic (i.e., related to one’s self-image) implications for a victim, while vertical collectivists would show the opposite pattern of results. In Studies 1 and 2 we directly compared British (i.e., a horizontal individualist culture) versus Greek (i.e., a vertical collectivist culture) participants. Study 3 aimed to replicate Studies 1 and 2 by measuring cultural values at the individual level. The results obtained in the three studies provided support for most of our hypotheses. Theoretical and practical implications are discussed.

KEYWORDS:

• Individualism-collectivism
• financial offender
• punishment
• beliefs about offenders

Over the last few decades there has been a dramatic increase in financial crime; Approximately $800 billion-$2 trillion in US dollars (that is 2–5% of global GDP) is money laundered globally in one single year (United Nations, 2019); Around half of cryptocurrency (i.e., Bitcoin) transactions, involving $76 billion approximately, were associated with illegal activity in 2019, which is close to the market for illegal drugs in Europe and the US together ($100 billion and €24 billion and per year respectively; Foley et al., 2019); Illicit inflows were estimated at $1.4-$2.5 trillion in 2014 (Global Financial Integrity, 2017). Financial crime is defined as the illicit activity against property to obtain financial benefit and advantage and it may be carried out by an individual or a group of people. The main motive in finanicial crimes is the actor’s economic gain (Europol, 2020). Importantly, financial crime refers to (usually nonviolent) crimes that are frequently committed by politicians, managers, (FBI, 2020) and other people with power and high standing (Gregg, 2015) as those people have access to large financial resources (S. T. Fiske, 2010). Financial crime is often interchangeably used with white-collar crime given the great similarity between the two (see Koller et al., 2014; Sutherland, 1950). However, white collar crime, including any kind of organized, nonviolent crime is a much broader category than financial crime. For instance, white collar crime includes financial crimes such as money embezzlement, corporate and mortgage fraud, bank and financial institution fraud, but also public corruption, election law violations, and health care fraud, among others (FBI, 2020). The current study focuses on financial crime in particular.

Despite the serious societal and economic ramifications of financial crime, there is little research investigating how people respond to financial offenders. For instance, prior research has shown that corruption and organized crime get much social media attention (Levi, 2008) disabling offenders from reintegrating into society (Shepherd et al., 2019). Yet, how third parties (i.e., observers) respond to financial criminals remains largely unexplored. In the current study, we investigate the effects of culture (e.g., horizontal individualism vs. vertical collectivism) on different types of reaction to financial offenders. Moreover, we argue that the type of inflicted harm (e.g., instrumental vs. symbolic) will moderate the effects of culture on people’s reaction.

Effects of culture on reactions to financial offenders

Culture is described as the shared attitudes, values, and goals of a nation, institution or organization, and is an important antecedent of punishment of offenders. Individualism/collectivism (Hofstede, 1980; Triandis, 1990) is, amongst all cultural variations, the most investigated dimension in the field of punishment (Eriksson et al., 2017; Gelfand et al., 2011; Hampden-Turner & Trompenaars, 2000; Leung & Stephan, 2001; Na & Loftus, 1998). A principal element of individualism is the supposition that individuals are self-reliant and independent of one another, while collectivism’s key component is the “assumption that groups bind and mutually obligate individuals” (Oyserman et al., 2002).

Research suggests that because individualists are more likely to attribute blame to internal rather than external causes, they are usually harsher to offenders than their collectivist counterparts (Na & Loftus, 1998). This theorizing is in line with other findings showing that individualists are more likely to punish irresponsible behavior (Hampden-Turner & Trompenaars, 2000; Ouchi, 1981). In contrast, collectivists adopt a more passive stance – and thus are less punitive – as they are more external in their locus of control (Spector et al., 2002) and believe they cannot influence others’ behavior through punitive strategies (Dedahanov, 2014). The literature on the matter is inconsistent, however, as the opposite pattern of results also has been observed (e.g., collectivists are more punitive than individualists; Brauer & Chaurand, 2010; Eriksson et al., 2017; Gelfand et al., 2011; Leung & Stephan, 2001; Liu et al., 2017; Wang & Leung, 2010; Zemba et al., 2006). Most importantly, the effects of culture on punishing financial offenders, who are not common criminals but often have a rather privileged background and a powerful standing, remains largely unexplored.

To address the inconsistent findings, we examine individualism and collectivism (Hofstede, 1980) in combination with another cultural variable namely power distance (for a multidimensional operationalization of culture see also, Gelfand et al., 1996). Power distance is defined as “the extent to which the less powerful members of institutions and organizations expect and accept that power is distributed unequally” (Hofstede, 1985, p. 347). Hence, culture, be it individualistic or collectivistic, can be vertical (high power distance) or horizontal (low power distance; Singelis et al., 1995; Triandis, 1995). Specifically, Horizontal collectivism (i.e., low power distance collectivism; e.g., Israeli kibbutz) is characterized by equality, social cohesion and harmony, cooperation, and interdependence, with individuals feeling connected with the members of the in-group. In vertical collectivism (i.e., high power distance collectivism; e.g., Eastern Europe, India, East Asia), individuals accept social hierarchy and power asymmetries, and are imbued with a sense of obedience, submission, and duty to the powerful. In horizontal individualism (i.e., low power distance individualism; e.g., Australia, Sweden, UK), people value both equality and freedom (see A. P. Fiske, 1992; Rokeach, 1973) and are assumed to be autonomous, unique, and independent, yet more or less equal in status with others. In vertical individualism (i.e., high power distance individualism; e.g., US) people are autonomous and unique, however similar to vertical collectivism, they accept social inequalities and endorse competition for the sake of distinction and personal achievement (Singelis et al., 1995).

Given the special focus of this study on financial offenders, who most often are powerful (Gregg, 2015; Sutherland, 1950), and in order to streamline this work, we focus on the two opposite cultural variations of individualism-collectivism and power distance, namely horizontal individualism versus vertical collectivism. We compare how horizontal individualists versus vertical collectivists react toward (powerful) offenders accused of financial crime. Indeed, prior research suggests that horizontal individualists, driven by a sense of independency and egalitarian motives, are inclined to oppose or bring down those in power (N. Feather, 1998). On the contrary, vertical collectivists, who accept or merely tolerate unequal distributions within a group (“that is the way the system works or the way the world is”), are more likely to accept power differences, and therefore to obey powerful members of a group, institution, or society (Triandis, 1994, see also, Cross et al., 2000; Triandis & Gelfand, 1998).

Accordingly, in our first two studies we directly compare people’s reaction to financial offenders in two culturally opposite countries, specifically Greece (i.e., vertical collectivist country) versus the UK (i.e., horizontal individualist country). Based on Hofstede et al.’s (2010) cultural index which comes from a large database of 70 countries, Greece is toward the bottom (35) of the individualism index and to the higher side of power distance (60). In contrast, the UK is amongst the highest of the individualist scores (89) and it sits in the lower rankings of power distance (35). These scores put Greece and the UK at the two opposite extremes along the cultural dimensions with Greece being a vertical collectivist country and the UK a horizontal individualist country (Hofstede et al., 2010). To validate our UK-Greece comparison, in a third study we measure the underlying cultural variables (horizontal individualism/ vertical collectivism) within a single culture, and test whether or not the within-culture approach yields results that converge with the between-culture comparison. We specifically examine (a) if culture predicts a differential preference for active versus passive responses to crime and (b) whether culture determines how susceptible people are to different forms of harm.

Culture and active versus passive response to offenders

Previous research has mainly focused on punitive responses to offenders. Punishment, as a tool of direct communication of one’s reaction to crime is an active response to offenders which aims at achieving a fast resolution (Worthington & Wade, 1999). In this research, besides punishment intent, we consider the underlying motives and emotions that drive punishment as part of an active response to crime. More specifically, retributive motives for punishment (Kant, 1797) underscore the importance of rebalancing feelings of justice by actively paying an offender back for their misconduct (see Goldberg et al., 1999; see also Van Prooijen, 2018). As such, retributive motives are oriented toward taking action against the offender through punishment. Similarly, moral outrage often directly precedes punishment and involves “hot,” outward (i.e., directed toward others than the self; see Barclay et al., 2005) and action-oriented emotions such as rage, anger, disgust, and contempt (N.T. Feather, 2006; Folger & Skarlicki, 2005; Gromet & Darley, 2009). While retributive motives and emotions initially emerge as internal phenomenological states, they are strongly goal-oriented by directly motivating an active punitive response. Besides being action-oriented, moral outrage involves emotions that are intense and often visible to the target (see Demoulin et al., 2004), further supporting its conception as an active reaction to offense.

Besides such active reactions, people may deploy a passive reaction to offenders (Worthington & Wade, 1999), which includes endorsing private beliefs that condemn a transgression without directly communicating this to the offender or directly motivating an active punitive response (Caprara, 1986; Gottman, 1994; Snyder, 1994). Supporting this distinction between punitive behavior versus negative but passive beliefs, Fousiani and Van Prooijen (2019) showed that although beliefs about offenders (e.g., demonization and conspiracy beliefs) and punitive reactions (e.g., punishment intent and motives for punishment) are strongly related, they are grounded in different psychological processes and can transpire independently from each other. Demonizing refers to the perception of a person as an immoral, non-human being (Leyens et al., 2000; Li et al., 2014) with “evil” personality traits (Baumeister, 1997; Ellard et al., 2002; Van Prooijen & Van de Veer, 2010) who takes pleasure into harming others (see Reicher et al., 2008). Conspiracy beliefs are defined as the assumption that people (especially powerholders) are secretly involved in covert activities with the aim to achieve hidden goals that are perceived as unlawful or malevolent (Douglas et al., 2017; Van Prooijen & Acker, 2015; Van Prooijen & Van Vugt, 2018).

Due to the increased emphasis that horizontal individualists put on personal responsibility and accountability (Hampden-Turner & Trompenaars, 2000), we hypothesize that British (i.e., horizontal individualists), as compared with Greeks (i.e., vertical collectivists) will display: a) more active reactions and will thus be more willing to punish a (powerful) financial offender (Hypothesis 1a), b) stronger retributive punishment motives toward financial offenders (i.e., give them their just deserts) (Hypothesis 1b), c) stronger moral outrage as an outcome of the financial offense (Hypothesis 1c).

Due to norms of obedience to authority and a desire to avoid conflict with the powerful, however, Greeks (vertical collectivists), as opposed to British, will form stronger negative but passive beliefs about financial offenders and see them as evil persons (i.e., demonization) who collude in secret with other key persons (i.e., conspiracy beliefs) (Hypothesis 2). However, they will have a decreased preference for displaying active reactions (e.g., punishment intent, retributive motives for punishment, and moral outrage).

Culture and instrumental versus symbolic harm

Not all types of harm are the same or seen as equally detrimental by everyone. An act can cause instrumental harm, that is, a loss of objective, concrete, tangible, and material outcomes (Lievens & Highhouse, 2003) or, it can cause symbolic harm, that is, detrimental consequences for abstract, subjective, intangible, and immaterial outcomes such as one’s reputation or image (Aaker, 1997, 1999; Highhouse et al., 2007; Ledgerwood et al., 2007, see also De Cremer & Van Vugt, 2002; Scheepers et al., 2003). To the best of our knowledge, no study on punitive reactions distinguished between instrumental or symbolic harm. Based on prior distinctions between instrumental versus symbolic needs, we argue that people from different cultures will display different reactions to offenders depending on whether an offense generates instrumental or symbolic implications. Specifically, De Cremer and Van Vugt (2002) argued that instrumental needs are particularly important when people’s personal identity is salient. Instead, symbolic needs are more important when people’s social identity is salient (Tajfel & Turner, 1986; Ting-Toomey, 1988, 2005).

Building on the above findings, it is likely that the type of harm that a financial offender inflicts will moderate the effect of culture on people’s reactions. Notably, an instrumental harm that violates one’s equal access to valued resources, thereby jeopardizing one’s personal interests, would be experienced particularly negative in horizontally individualist cultures (e.g., the UK). In contrast, a symbolic harm that violates one’s duties and responsibilities toward their group, would be experienced particularly negative in vertically collectivist cultures (e.g., Greece). We therefore propose that British participants will indicate stronger negative reactions – both active and passive – to a financial offender for having caused instrumental as opposed to symbolic harm to a victim. In contrast, Greek participants will display stronger reactions – both active and passive – to a financial offender for having inflicted a symbolic rather than instrumental harm (Hypothesis 3) (see Table 1 for a summary of the Hypotheses).

Table 1. Summary of research hypotheses and findings.

Hypothesis	Study 1	Study 2	Study 3
Horizontal individualists, as compared with vertical collectivists, will display more active reactions to financial offenders and will thus:
-be more willing to punish a (powerful) financial offender (Hypothesis 1a).	Not supported (low scale reliability)	Supported	Supported
-display stronger retributive punishment motives toward financial offenders (Hypothesis 1b).	Supported	Supported	Supported
-experience stronger moral outrage as an outcome of the financial offense (Hypothesis 1c).	Not supported	Supported	Not supported
Vertical collectivists, as compared with horizontal individualists, will display more passive reactions to financial offenders and will thus demonize them more and will also endorse stronger conspiracy beliefs about them (Hypothesis 2).	Supported	Partly supported	Partly supported
Horizontal individualists will indicate stronger negative reactions – both active and passive – to a financial offender for having caused instrumental as opposed to symbolic harm to a victim / Vertical collectivists will display stronger reactions – both active and passive – to a financial offender for having inflicted a symbolic rather than instrumental harm (Hypothesis 3).	Partly supported for moral outrage	Partly supported for moral outrage and conspiracy beliefs	Partly supported for moral outrage and conspiracy beliefs

Overview of the studies

To sum up, in the current study, we investigate the effect of culture on observer’s passive and active reactions to financial offenders and we further investigate the moderating role of type of harm. As active reactions to those offenders we consider punishment-oriented responses (e.g., punitive intent, motives for punishment, and moral outrage) whereas as passive reactions we consider the endorsement of condemning beliefs about offenders (e.g., conspiracy beliefs and demonization).

We conducted three experiments where participants read a vignette about an offender who had inflicted financial harm with either instrumental (but not symbolic) implications or with symbolic (but not instrumental) implications to the victims. Given that financial (i.e., white-collar) offenders are often also powerful (Gregg, 2015; Sutherland, 1950), in all three studies the offender was consistently a powerful person (e.g., top manager in an organization). Studies 1 and 2 operationalized culture at the national level (British vs. Greeks¹) whereas Study 3 operationalized culture at the individual level and measured it.

Study 1

Methods

Participants

A total of 308 British participants (201 females and 106 males, 1 unidentified; M_age = 35.52, SD = 11.67) living in the UK, and a total of 381 Greek participants living in Greece (217 females and 152 males, 12 unknown; M_age = 34.34, SD = 10.07) took part in this study. An a-priori power analysis revealed that 245 participants were required for 95% power to detect a medium effect size (f = .25). British participants were recruited via Prolific academic and were paid £1.00 (€1.14) for their participation. Greek participants were approached via our research assistants’ network. Research assistants were university students (in Greece) and did the data collection for course credits. Greek participants were approached via several methods including personal network, LinkedIn, Facebook, and interception and they were not paid. Most participants were employed at several organizations and only a small percentage (10%) were students. Non-British or non-Greek participants could not take part in the study.

Experimental design and procedure

We manipulated the type of harm in vignettes (see online supplemental material). Participants read a scenario (in their native language) which presented a powerful offender (i.e., a top manager of a large organization) as having embezzled large sums of money from the organization where he has been working. This act had inflicted either a) instrumental implications (“ … the organization lost substantial financial resources. As a consequence of the fraudster’s act, the British/ Greek public organization is deprived of the money that is needed to afford many job positions, which means that many employees are going to lose their job”) but not symbolic implications (“ … Luckily, the whole incident was kept ‘internal’ and no publicity was given to what exactly the fraudster did. Therefore, no reputation harm was inflicted on the British/ Greek organization”) or b) symbolic implications (“ … the incident went public and the organization lost face abroad. As a consequence of the fraudster’s act, the British/ Greek public organization is mocked and ridiculed in the foreign media for weeks, which means that its reputation is severely harmed”) but not instrumental implications (“ … the sums of money that were stolen by the British/ Greek fraudster were refunded in full by the insurance. Therefore, no financial harm was inflicted on the British/ Greek organization”). The offender was presented as a British vs. Greek person and was given a typical British or a typical Greek name. After reading this vignette, participants were requested to fill in the manipulation checks and measures assessing active and passive reactions to the offender. The study was programmed on Qualtrics and respondents participated online.

Measures

Manipulation checks

We checked our type of harm manipulation with two items: The offender’s action caused to the British (Greek) organization: 1) Financial problems (lack of monetary resources); 2) Dignity-related problems (ashaming) (1 = not at all, 7 = to a great extent).

Active responses to financial offenders

Punishment intentions

We assessed observers’ intentions to punish the offender with a 3-item scale (“the offender deserves to be punished,” “the offender should receive a sanction for his act,” “it is right that the offender be sanctioned by the responsible authority”; (1 = absolutely disagree, 7 = absolutely agree; α = .51 for Greeks and .77 for British) (cf., Fousiani & Demoulin, 2019; Fousiani & Van Prooijen 2022; Fousiani & Van Prooijen, 2019).

Retributive motives for punishment

We used the four-item Retributive Motives for Punishment Scale (Fousiani & Demoulin, 2019; Fousiani & Van Prooijen, 2019; Fousiani et al., 2019) to measure retributive motives for punishing offenders (1 = absolutely disagree, 7 = absolutely agree). Cronbach’s alpha was. 60 for Greeks and .87 for British.

Moral outrage

We asked participants to rate the extent to which they felt each of the following emotions about the offender: anger, disgust, contempt (1 = not at all, 7 = to a great extent; α = .68 for Greeks and .68 for British). These emotions were included in the moral outrage scale of Bastian et al. (2013) based on Haidt’s (2003) work on moral outrage emotions.

Passive responses to financial offenders

Conspiracy beliefs

We assessed conspiracy beliefs about the offender with the 5-item scale of Fousiani and Van Prooijen (2019). Sample items include: “The offender did not act alone; he colluded with other key-persons in embezzling public money,” “The offender designed the money embezzlement plan in secret together with other persons who are in “key-positions” (1 = absolutely disagree, 7 = absolutely agree; α = .88 for Greeks and .91 for British).

Demonization

To assess demonization of the offender we used the 5-item demonization scale of Fousiani and Van Prooijen (2019) and Van Prooijen and Van de Veer (2010). Some sample items are “this act was caused entirely by the offender’s evilness,” “the offender seems to enjoy committing fraudulent behaviors” (1 = absolutely disagree, 7 = absolutely agree; α = .69 for Greeks and .82 for British).

All scales were adapted to the specifics of the current study. For the Greek participants, all scales were translated and back-translated to Greek² (see online supplementary material for the complete scales).

Results

To test our hypotheses, we conducted all statistical analyses with SPSS 26. We used the lavaan package in R (Rosseel, 2012) to test measurement invariance. Punishment intention was excluded from further analysis given its low reliability in the Greek sample. Thus, hypothesis 1a could not be tested in this study. Correlations between the study variables, means, and standard deviations are presented in Table 2.

Table 2. Pearson correlations coefficients between study variables, means, and standard deviations (study 1).

	1	2	3	4	M (SD)
1. Conspiracy beliefs	1	.31***	.13**	−.14**	3.68 (1.64)
2. Demonization		1	.41***	.13**	4.54 (1.20)
3. Moral outrage			1	.18***	4.33 (1.17)
4. Retributive motives				1	5.75 (1.07)

* p < .05; ** p < .01; ***p < .001

Measurement invariance analyses

Given that we collected data from two different cultures, we first tested for measurement invariance using the lavaan package in R (Rosseel, 2012). We specified three nested factor models, including all items of the four measured variables (retributive motives, moral outrage, conspiracy beliefs, and demonization) to test for configural, metric, and scalar invariance. As indicators of acceptable model fit for the basic configural model we considered the CFI (> .90), the RMSEA (< .08) and the SRMR (< .08). Metric and scalar invariance were tested using the thresholds of ΔCFI < −.010; ΔRMSEA < .015; and ΔSRMR < .030 (Chen, 2007).

The basic configural model had a good fit (CFI = .909; RMSEA = 0.076, CI_90%[.069; .083]; SRMR = .077; χ²[226, N = 684 = 671.41, p < .001). The metric model (restricting factor loadings to be equal across cultural samples) did not deviate from the configural model according to two out of three indicators (ΔCFI = −.021; ΔRMSEA = .006; ΔSRMR = .011; Δχ²[13, N = 684] = 118,70, p < .001). The scalar model (also restricting intercepts to be equal across cultural samples) deviated substantially from the metric model, however (ΔCFI = −.168; ΔRMSEA = .045; ΔSRMR = .067; Δχ²[13, N = 684] = 837.24, p < .001). These results indicate that our measures have metric, but not scalar, invariance across cultures.

Manipulation checks

We conducted a simple (type of harm: instrumental vs. symbolic) univariate ANOVA with the manipulation check items as the main dependent variables. The type of harm manipulation had a significant effect on the perception of the harm as instrumental F(1, 687) = 529.05, p < .001, η² = .43. As expected, participants in the instrumental harm (M = 6.15, SD = 1.27) as opposed to the symbolic harm condition (M = 3.13, SD = 2.09) perceived the harm as more instrumental. Moreover, the effect of type of harm proved significant on the perception of harm as symbolic F(1, 687) = 265.18, p < .001, η² = .28. Participants in the symbolic (M = 5.91, SD = 1.47) as opposed to the instrumental harm condition (M = 4.74, SD = 2.17) perceived the harm as more symbolic. We conclude that our manipulation worked as intended.

Active responses to financial offenders

Retributive motives for punishment

We conducted a 2 (country: British/Greeks) x 2 (type of harm: instrumental/symbolic) Analysis of Variance (ANOVA) with retributive motives for punishment as the main dependent variable. As predicted, the effect of country on retributive motives proved significant F(1, 678) = 49.82, p < .001, η² = .07. In line with Hypothesis 1b, British displayed stronger retributive punishment against offenders as opposed to Greeks (Means and standard deviations are presented in Table 3). However, the country by type of harm effect did not prove significant (F < 1). These results do not support Hypothesis 3.

Table 3. Means and standard deviations for the study variables for British and Greek participants (study 1).

	British		Greeks
	M	SD	M	SD
Retributive motives	6.06	.94	5.50	1.10
Moral outrage	4.32	1.08	4.34	1.25
Conspiracy beliefs	2.94	1.39	4.29	1.57
Demonization	4.14	1.18	4.87	1.11

All ratings were on 7-point scales ranging from 1 = absolutely disagree to 7 = absolutely agree. Ratings of moral outrage scale ranged from 1 = not at all, 7 = to a great extent).

Moral outrage

The main effect of country on moral outrage was unexpectedly not significant (Fs < 1). Hypothesis 1c was therefore not supported. However, the interaction effect between country and type of harm came out significant F(1, 678) = 4.83, p = .03, η² = .01. As predicted, Greeks experienced stronger outrage when the harm was symbolic (M = 4.47, SD = 1.19) as opposed to instrumental (M = 4.21, SD = 1.29), F(1, 372) = 4.20, p = .04, η² = .01. However, the mean difference between conditions was not significant for British participants F(1, 306) = 1.17 p = .28, η² = .004 (see Figure 1 in the online supplementary material). These findings provide partial support for Hypothesis 3.

Figure 1. Horizontal individualism by type of harm in the prediction of moral outrage against offenders (study 3).

Note. Ratings were on a 7-point scale ranging from 1 = not at all, 7 = to a great extent.

Passive responses to financial offenders

Conspiracy beliefs

The main effect of country was significant F(1, 685) = 137.78, p < .001, η² = .17. In line with Hypothesis 2, results showed that Greeks endorse stronger conspiracy beliefs as compared with British (means and standard deviations in Table 3). However, the country by type of harm interaction was not significant (F < 1). Hypothesis 3 was not supported.

Demonization

Again, the main effect of country was significant F(1, 685) = 67.31, p < .001, η² = .09. Results showed that Greeks demonize the offender more as compared with British (means and standard deviations in Table 3). These findings provide further support for Hypothesis 2. However, the interaction effect between country and type of harm was not significant (F < 1). These results did not support Hypothesis 3.

Discussion

Because of low reliability of the punishment intent scale, we were not able to test Hypothesis 1a. In line with Hypothesis 1b, Study 1 showed that British were more willing to punish a financial offender and give them their just deserts (i.e., retributive motives for punishment). However, the main effect of country on moral outrage (Hypothesis 1c) did not prove significant. Furthermore, we found that Greeks reacted to financial offenders in a more passive way as compared with the British. More specifically, Greeks formed stronger negative beliefs about financial offenders, such as demonizing them and endorsing conspiracy beliefs. These findings supported Hypothesis 2. In line with Hypothesis 3, we found evidence for a moderating effect of type of harm on the relationship between country and reaction to financial offenders. As predicted, Greeks experienced stronger moral outrage when the harm had symbolic as opposed to instrumental implications; however, the type of harm did not influence moral outrage of British. Unexpectedly, the country by type of harm interaction did not prove significant for the rest of our variables (besides moral outrage). More studies are needed in order to conclude with certainty if British and Greeks are more or less sensitive to harms with instrumental or symbolic implications.

Study 2

Study 2 aimed to replicate Study 1 and focused on a different type of financial offense, namely fabrication of the numbers regarding finances by a school principal. Similar to Study 1, our participants came from either the UK or Greece. Interestingly, prior research has shown that punitive responses between observers and victims differ such that victims’ punishments are higher than observers’ punishments (Fehr & Fischbacher, 2004). Therefore, we deemed it important to replicate the findings of Study 1 (participants took the perspective of an observer) while asking participants to evaluate the scenario as a victim (rather than an observer) of the fraudulent case. To facilitate involvement in the scenario, all participants were school teachers in real life.

Methods

Participants

A total of 140 British school teachers (104 females and 33 males, 3 unidentified; M_age = 42.43, SD = 10.72) living in the UK, and a total of 140 Greek school teachers living in Greece (110 females and 22 males, 8 unknown; M_age = 39.58, SD = 10.97) took part in this study. An a priori power analysis revealed that 245 participants were required for 95% power to detect a medium effect size (f = .25). British participants were recruited via Prolific academic and were paid £1.00 (€1.14) for their participation. Greek teachers were approached via our research assistants’ network. Research assistants were school teachers themselves working in various schools in Greece. Research assistants used their professional network (colleagues at school) for the data collection. Greek participants were not paid.

Experimental design and procedure

We manipulated the type of harm in vignettes. Participants read a scenario (in their native language) that presented a powerful offender (i.e., the Principal of a school in the UK vs. Greece) as having fabricated the numbers regarding the financial management of the school. This act had inflicted either a) instrumental but not symbolic implications or b) symbolic but not instrumental implications similarly with Study 1 (see online supplemental material for the vignettes). Finally, we added a control condition where participants read about the offense but were not given any information regarding the implications of the harm. As in Study 1, the offender was presented as a British vs. Greek person and was given a typical British or a typical Greek name. After reading the vignette, participants filled in the same measures as in Study 1. The study was programmed on Qualtrics and respondents participated online.

Measures

Manipulation checks

We checked the manipulation of type of harm with two items: “The harm has consequences on my workload as a teacher,” and “The harm has consequences on my reputation as a teacher.”

Punishment intention, retributive motives for punishment, conspiracy beliefs, and demonization were measured as in Study 1. Moral outrage scale was similar to Study 1 and included one additional item: rage. The scale reliabilities were as follows: Punishment intent: α = .89 for Greeks and .91 for British, retributive motives for punishment: α = .71 for Greeks and .87 for British, and moral outrage: α = .85 for Greeks and .88 for British, conspiracy beliefs: α = .91 for Greeks and .90 for British, demonization: α = .79 for Greeks and .87 for British.

Results

Correlations between the study variables, means, and standard deviations are presented in Table 4.

Table 4. Pearson correlations coefficients between study variables, means, and standard deviations (study 2).

	1	2	3	4	5	M (SD)
1. Conspiracy beliefs	1	−.07	.03	.04	.05	3.85 (1.45)
2. Demonization		1	.11	−.01	−.09	3.84 (1.51)
3. Moral outrage			1	.48***	.42**	4.42 (1.50)
4. Punishment intent				1	.51***	6.14 (1.12)
5. Retributive motives					1	5.55 (1.14)

* p < .05; ** p < .01; ***p < .001

Measurement invariance analyses

We again tested for configural, metric, and scalar invariance of our measures. The basic configural model – including all items of the five measures reported here (punishment intent, retributive motives, moral outrage, conspiracy beliefs, and demonization) – had a good fit (CFI = .930; RMSEA = 0.072, CI_90% [.063; .082]; SRMR = .072; χ²[338, N = 273] = 614.10, p < .001). The metric model did not deviate from the configural model (ΔCFI = −.009; ΔRMSEA = .003; ΔSRMR = .005; Δχ²[16, N = 273] = 49.27, p < .001). The scalar model again did deviate from the metric model, however (ΔCFI = −.111; ΔRMSEA = .039; ΔSRMR = .052; Δχ²[16, N = 273] = 421.95, p < .001). As in Study 1, the measures display metric but not scalar invariance across cultures.

Manipulation checks

We conducted a simple (type of harm: instrumental/symbolic/control) univariate ANOVA with the manipulation check item that indicates perception of the harm as instrumental (workload-related harm) as the main dependent variable. The type of harm manipulation had a significant effect on the perception of the harm as instrumental F(2, 277) = 44.762, p < .001, η² = .24. As expected, participants perceived the harm as more instrumental in the instrumental (M = 6.02, SD = 1.32) as opposed to the symbolic harm condition (M = 3.66, SD = 1.89). The mean for the control condition was (M = 4.24, SD = 1.91). Moreover, we conducted a univariate ANOVA with perception of the harm as symbolic (reputation-related harm) as the main variable. Again, the effect of type of harm proved significant F(1, 280) = 25.01, p < .001, η² = .15. Participants in the symbolic harm (M = 5.73, SD = 1.48), but also in the control condition (M = 5.15, SD = 1.80) perceived the harm as more symbolic than in the instrumental condition (M = 3.83, SD = 2.22). We conclude that our manipulations were effective.

Active responses to financial offenders

Punishment intent

As expected, the main effect of country proved significant F(1, 268) = 6.98, p = .009, η² = .03. British displayed a stronger punishment intent as opposed to Greeks (see Table 5 for means and standard deviations). These results support Hypothesis 1a. However, the interaction effect between country and type of harm did not come out significant (F < 1).

Table 5. Means and standard deviations for the study variables for British and Greek participants (study 2).

	British		Greeks
	M	SD	M	SD
Punishment intent	6.31	.91	5.96	1.27
Retributive motives	5.84	1.05	5.26	1.16
Moral outrage	4.59	1.40	4.27	1.56
Conspiracy beliefs	3.84	1.54	3.87	1.35
Demonization	2.99	1.26	4.67	1.24

All ratings were on 7-point scales ranging from 1 = absolutely disagree to 7 = absolutely agree. Ratings of moral outrage scale ranged from 1 = not at all, 7 = to a great extent).

Retributive motives for punishment

The main effect of country on retributive motives proved significant F(1, 268) = 18.05, p < .001, η² = .063. In line with Hypothesis 1b, British displayed stronger retributive punishments against offenders as opposed to Greeks (Means and standard deviations in Table 5). However, the country by type of harm effect did not prove significant F < 1 (Means and standards deviations in Table 6). These results do not provide support for Hypothesis 3.

Table 6. Means and standard deviations for the study variables across experimental conditions (study 2).

	British						Greeks
	Instrumental harm		Symbolic harm		Control		Instrumental harm		Symbolic harm		Control
	M	SD	M	SD	M	SD	M	SD	M	SD	M	SD
Punishment intent	6.63	.65	6.32	.82	6.00	1.10	5.90	1.15	6.08	1.17	5.90	1.46
Retributive motives	6.09	.81	5.81	1.05	5.64	1.20	5.39	1.29	5.31	1.13	5.09	1.08
Moral outrage	4.90	1.31	4.61	1.24	4.28	1.57	3.94	1.49	4.30	1.49	4.49	1.66
Conspiracy beliefs	4.25	1.61	3.77	1.44	3.52	1.50	3.55	1.41	3.99	1.39	4.00	1.24
Demonization	3.20	1.28	3.04	1.14	2.74	1.34	4.76	1.29	4.68	1.31	4.58	1.15

All ratings were on 7-point scales ranging from 1 = absolutely disagree to 7 = absolutely agree. Ratings of moral outrage scale ranged from 1 = not at all, 7 = to a great extent)

Moral outrage

The main effect of country came out significant F(1, 268) = 6.98, p = .009, η² = .03, and in line with Hypothesis 1c, it showed that British display stronger outrage toward offenders as opposed to Greeks (see Table 5). The country by type of harm interaction effect was also significant F(2, 268) = 3.91, p = .02, η² = .03 and showed that British, as opposed to Greeks, display stronger outrage when the harm is instrumental F(1,82) = 10.85, p = .001, η² = .12. However, moral outrage did not differ for Greeks and British in the symbolic and control conditions Fs < 1. (Means and standard deviations in Table 6; see also Figure 2 in the online supplementary material). These findings are in line with Hypothesis 3.

Figure 2. Vertical collectivism by type of harm in the prediction of conspiracy beliefs about offenders (study 3).

Ratings were on a 7-point scale ranging from 1 = absolutely disagree to 7 = absolutely agree.

Passive responses to financial offenders

Conspiracy beliefs

The main effect of country on conspiracy beliefs was not significant (F < 1) and thus Hypothesis 2 was not supported. However, the country by type of harm interaction effect did prove significant F(2, 273) = 4.64, p = .01, η² = .03. In line with Hypothesis 3, British participants endorsed higher conspiracy beliefs in the instrumental harm condition as opposed to the symbolic harm or the control conditions F(2,137) = 3.33, p = .04, η² = .05. However, the mean difference was not significant for Greeks F(2,136) = 1.54, p = .22, η² = .02 (Means and standard deviations in Table 6; see also Figure 3 in the online supplementary material).

Demonization

The main effect of country proved significant F(1, 273) = 127.42, p < .001, η² = .32, and in line with Hypothesis 2, it showed that Greeks demonize a financial offender more compared to British (see Table 5 for means and standard deviations). However, the interaction effect between country and type of harm was not significant (F < 1).

Discussion

Study 2 supported our hypotheses and partly replicated Study 1. More specifically, results revealed that British were more willing to punish a financial offender (Hypothesis 1a) with the aim to make them pay for what they did (Hypothesis 1b), and also experienced stronger moral outrage toward financial offenders (Hypothesis 1c) (all active reactions to offenders). In contrast, supporting Hypothesis 2, Greeks endorsed stronger conspiracy beliefs about financial offenders and demonized the offender more (i.e., passive reactions to crime).

Furthermore, we found evidence for a moderating effect of type of harm on the relationship between country and reaction to financial offenders for some of the measures. First, we observed a significant interaction effect between country and type of harm on moral outrage. More specifically, British as opposed to Greeks, indicated having experienced stronger moral outrage against a financial offender when the inflicted harm was instrumental. In a similar vein, British endorsed stronger conspiracy beliefs about an offender who caused an instrumental as opposed to symbolic harm. However, the type of harm did not influence the endorsement of conspiracy beliefs nor the experienced outrage of Greeks. These findings provide mixed support for Hypothesis 3.

Study 3

While informative, Studies 1 and 2 are limited by focusing on two specific countries that differ in more ways than the assumed underlying cultural dimensions. Furthermore, while our measures showed metric invariance across cultures, they did not show scalar invariance. We therefore conducted a study within the same country and measured the underlying cultural values at the individual level (see Singelis et al., 1995). Moreover, in Study 3 we kept the identity of the offender neutral given that the study took place in a single country.³

Methods

Participants

A total of 260 British participants (135 females and 124 males, 1 unknown; M_age = 35.37, SD = 9.31) living in the UK took part in this study. An a-priori power analysis revealed that 254 participants were required for 95% power to detect a medium effect size (f = .25). We recruited participants via Prolific academic and we paid them £1.00 (€1.14) for their participation. The majority of participants were working at several organizations, while a very small percentage (2.7%) were students.

Experimental design and procedure

Participants first filled in measures assessing their cultural values. Subsequently, they were assigned to one out of four experimental conditions. We manipulated the type of harm in vignettes (see online supplemental material). Participants read a scenario that presented a powerful offender (i.e., a top manager of an organization) as having been involved in scandals related to money laundering through the organization. This act had inflicted either a) instrumental but not symbolic implications or b) symbolic but not instrumental implications. After reading this vignette, participants filled in the same measures as in Study 1. The study was programmed on Qualtrics and respondents participated online.

Measures

We checked the manipulation of type of harm similarly with Studies 1 and 2. Punishment intention, retributive motives for punishment, conspiracy beliefs, and demonization were measured as in Studies 1 and 2. Moral outrage was measured as in Study 2. We adjusted our scales (whenever needed) to the specifics of this study. Cronbach’s alphas of the scales were as follows: Punishment intent: α = .89, retributive motives for punishment: α = .84, moral outrage: α = .88, conspiracy beliefs: α = .85, demonization: .81.

Cultural value orientation at the individual level

We used the 16-item scale of Triandis and Gelfand (1998) to measure personal cultural value orientation (1 = absolutely disagree, 7 = absolutely agree). The scale includes 4 items that measure horizontal individualism (e.g., “I’d rather depend on myself than others,” α = .71), 4 items that measure vertical individualism (e.g., “It is important that I do my job better than others,” α = .69), 4 items that measure horizontal collectivism (e.g., “If a coworker gets a prize, I would feel proud,” α = .75), and 4 items that measure vertical collectivism (e.g., “Parents and children must stay together as much as possible,” α = .72).⁴

Results

Correlations between the study variables, means, and standard deviations are presented in Table 7.

Table 7. Pearson correlations coefficients between study variables, means, and standard deviations (study 3).

	1	2	3	4	5	6	7	8	9	M (SD)
1. Conspiracy beliefs	1	.28***	.12	.13*	.08	.06	.20**	.03	.05	4.09 (1.28)
2. Demonization		1	.45***	.23***	.12	.07	.20***	.15*	.24***	3.89 (1.17)
3. Moral outrage			1	.45***	.27***	.08	.19**	.17**	.25***	5.16 (1.44)
4. Punishment intent				1	.51***	.14*	.09	.13*	.13*	6.52 (.87)
5. Retributive motives					1	.31***	.04	.13*	.15*	6.06 (.94)
6. Horizontal individualism						1	.30***	.02	.07	5.46 (.90)
7. Vertical individualism							1	.03	.26***	3.88 (1.09)
8. Horizontal collectivism								1	.36***	5.26 (1.00)
9. Vertical collectivism									1	5.15 (1.09)

* p < .05; ** p < .01; ***p < .001

Manipulation checks

We conducted a simple (type of harm: instrumental/symbolic) univariate ANOVA with the manipulation check item that indicates perception of the harm as instrumental (financial harm) as the main dependent variable. The type of harm manipulation had a significant effect on the perception of the harm as instrumental F(1, 258) = 1122.57, p < .001, η² = .81. As expected, participants in the instrumental harm condition (M = 6.72, SD = .87) perceived the harm as more instrumental than in the symbolic condition (M = 1.77, SD = 1.46). Moreover, we conducted a univariate ANOVA with perception of the harm as symbolic (reputation-related harm) as the main variable. The effect of type of harm proved significant F(1, 258) = 253.15, p < .001, η² = .50. Participants in the symbolic (M = 6.71, SD = .82) as opposed to the instrumental harm condition (M = 3.38, SD = 2.21) perceived the harm as more symbolic. We conclude that our manipulations worked as intended.

Hypothesis testing

To test our Hypotheses, we conducted separate hierarchical regression analyses for each of the dependent variables. We first entered centered scores of the four cultural dimensions and then the interaction between each cultural dimension with type of harm. The summary of the effects of horizontal individualism and vertical collectivism on each of the dependent variables of the study is presented in Table 1 in the online supplementary material.

Active responses to financial offenders

Punishment intent

Results revealed a marginally significant effect of horizontal individualism only, b = .12, SE = .06, t = 1.87, p = .06, adjusted R² = .04, F(4, 259) = 2.80, p =. 03. This finding supports Hypothesis 1a and shows that horizontal individualism predicts a stronger willingness to punish an offender. No interaction effects between cultural values and type of harm came out significant. Hypothesis 3 hence was not supported.

Retributive motives for punishment

In line with Hypothesis 1b, only the main effect of horizontal individualism on retributive motives for punishment came out significant, b = .34, SE = .07, t = 5.31, p < .001, adjusted R² = .11, F(4, 259) = 9.19, p <. 001, showing that horizontal individualism is a positive predictor of retributive punishment motives.

Unexpectedly, and contrary to Hypothesis 3, the interaction effect between horizontal individualism and type of harm in the prediction of retributive motives did not come out significant. No other interaction effects were significant.

Moral outrage

Results showed a significant positive effect of vertical individualism, b = .18, SE = .09, t = 2.12, p = .04 and vertical collectivism on moral outrage, b = .23, SE = .09, t = 2.54, p = .01, adjusted R² = .09, F(4, 259) = 6.10, p <. 001. Unexpectedly, the main effect of horizontal individualism was not significant. Hypothesis 1c was not supported. Interestingly, we found a marginally significant interaction between horizontal individualism and type of harm on moral outrage (ΔR² = .013), F (1, 256) = 3.80, p = .05, showing that the positive effect of horizontal individualism on moral outrage is stronger when the harm is instrumental, b = .31, SE = .13, t = 2.50, p = .01 rather than symbolic, b = −.06, SE = .14, t = −.40, p < .05 (see Figure 1). These results supported Hypothesis 3. No other interactions were found to be significant (Fs < 1).

Passive responses to financial offenders

Conspiracy beliefs

Results showed a significant positive effect of vertical individualism, b = .23, SE = .08, t = 2.94, p = .004; (adjusted R² = .04), F (4, 259) = 2.56, p = .04. Unexpectedly, the main effect of vertical collectivism did not come out significant. Hypothesis 2 was therefore not supported. However, in line with Hypothesis 3, the interaction effect between vertical collectivism and type of harm proved significant (ΔR² = .016), F (1, 256) = 4.28, p = .04, showing that the positive effect of vertical collectivism on conspiracy beliefs is stronger when the harm is symbolic, b = .25, SE = .11, t = 2.16, p = .03 rather than instrumental, b = −.06, SE = .09, t = −.62, p < .05 (see Figure 2). No other interactions were found to be significant, (Fs < 1).

Demonization

We found a significant positive effect of vertical individualism, b = .16, SE = .07, t = 2.35, p = .02 and vertical collectivism, b = .18, SE = .07, t = 2.56, p = .01, adjusted R² = .09, F (4, 259) = 5.92, p < .001. These findings provide partial support for Hypothesis 2. No other main effects or interactions were found to be significant (Fs < 1).

Discussion

Study 3 measured cultural values at the individual level. Supporting Hypothesis 1, results revealed that horizontal individualism predicts active rather than passive responses to financial offenders. Specifically, horizontal individualism was associated with a willingness to punish financial offenders (Hypothesis 1a), and with retributive motives for punishment (Hypothesis 1b). Moreover, vertical collectivism particularly predicted passive reactions to financial offenders (e.g., demonization), which was in line with Hypothesis 2. Unexpectedly, vertical individualism predicted moral outrage, conspiracy beliefs, and demonization of financial offenders. A possible explanation for this finding is that vertical individualism, which emphasizes independence and autonomy on the one hand, and hierarchy and power distance on the other (Singelis et al., 1995), might activate both active and passive negative reactions to (powerful) financial offenders when the latter are found to violate social norms and conventions. Indeed, those in hierarchically higher positions are assigned the responsibility to ensure that important group goals are met (Sassenberg et al., 2014). When those people commit a financial crime and jeopardize the collective goals, people with individualistic vertical cultural values might be more likely to take matters into their own hands and react both with punitivity and negative beliefs. Further research is required in order to investigate this effect.

Furthermore, consistent with Hypothesis 3, we found that horizontal individualism is related to stronger moral outrage against financial offenders when a harm has instrumental as opposed to symbolic implications. Moreover, vertical collectivism was related to stronger conspiracy beliefs about a financial offender when harm was symbolic instead of instrumental. These findings replicate and extend Studies 1 and 2, and further support Hypothesis 3.

General discussion

Financial crime is widespread and has serious implications for society and the economy. Yet, no research has investigated people’s reactions to financial offenders. In this research we argue that people’s responses to financial offenders are influenced by culture, because one’s cultural values regulate the degree of appropriateness and social acceptance of behavior (Bond, 2004). More specifically, we investigated how different types of individualism and collectivism along the dimension of power distance (see Gelfand et al., 1996), namely horizontal individualism and vertical collectivism, may affect people’s responses to financial offenders. Moreover, prior cultural research has primarily focused on people’s punishment-oriented responses to crime (i.e., active responses), while neglecting people’s private beliefs about offenders (i.e., passive responses), which are not necessarily translated into punishment. In the present work, taking into account that people do not always feel comfortable with actively confronting offenders who are in a powerful position, we have explicitly tested how cultural values predict active and passive responses to crime (e.g., demonization and conspiracy beliefs about financial offenders). Furthermore, we examined the differential effects that symbolic versus instrumental harm may have on observers from different cultures.

Studies 1 and 2 examined how people from culturally opposite countries (Greece vs. UK) react to financial offenders. Greece is a collectivist country with a high acceptance of power distance, constituting a typical example of a vertical collectivistic country. In contrast, the UK is an individualist country where power distance is low, constituting thus a typical horizontal individualistic country (see Hofstede et al., 2010). In both Studies 1 and 2, we found that British (as opposed to Greeks) displayed more active reactions to offenders (operationalized as punitive intent, retributive motives for punishment, and moral outrage). Specifically, British’ punitive reactions were driven by retributive or “just deserts” motives, that is, punishments that aim at rebalancing feelings of justice by actively paying an offender back for their misconduct (see Goldberg et al., 1999; see also, Van Prooijen, 2018). Moreover, British displayed stronger moral outrage against offenders as measured by outward negative moral emotions (anger, rage, contempt, disgust). In contrast, Greeks displayed more passive responses toward financial offenders (operationalized as demonization and conspiracy beliefs about an offender), in that they formed more negative beliefs about offenders such as demonizing them and seeing them as conspirators who collude with key persons. Study 3 largely replicated these findings. While not all results were equally straightforward, results revealed that horizontal individualism predicted stronger active responses, notably punishment intention and retributive punishment motives. Vertical collectivism predicted stronger passive responses, notably demonization of financial offenders. Together, these results support Hypotheses 1 and 2.

These findings suggest that vertical collectivists avoid conflicts and take a more passive stance to financial crime (see Dedahanov, 2014; Hampden-Turner & Trompenaars, 2000; Ouchi, 1981). Therefore, when confronted with a financial crime which is typically enacted by powerful offenders, vertical collectivists are relatively likely to see the harm as a condition that is out of their control (see also Spector et al., 2002), which they cannot combat via active reactions (such as punishment) and thus deploy more passive reactions that are oriented toward making sense of (powerful) offenders. These results are in line with recent findings showing that Chinese, as compared with Americans, endorse stronger conspiracy beliefs, which was attributable to their vertical collectivistic and power distant cultural orientation (Van Prooijen & Song, 2021). In contrast, horizontal individualists are more likely to deploy active reactions toward offenders and punish irresponsible behavior with the aim to restore justice (see also Hampden-Turner & Trompenaars, 2000; Ouchi, 1981).

Most importantly, our findings suggest that in order to paint a clearer picture of people’s reaction to financial (and thus powerful) offenders in particular, seeing individualism/collectivism as a cultural dimension that varies in terms of power distance is necessary. Differently put, individualism/collectivism alone is not a sufficiently informative dimension about people’s response to criminals as individualism/collectivism can take different shades depending on how much power distance is involved. Finally, it is noteworthy that this research is the first attempt to reconcile the inconsistent literature on the effects of culture on people’s reaction to offenders (Brauer & Chaurand, 2010; Dedahanov, 2014; Eriksson et al., 2017; Gelfand et al., 2011; Hampden-Turner & Trompenaars, 2000; Leung & Stephan, 2001; Liu et al., 2017; Na & Loftus, 1998; Ouchi, 1981; Spector et al., 2002; Wang & Leung, 2010; Zemba et al., 2006) and point out the importance of investigating culture as a multidimensional rather than a bipolar construct.

Moreover, all three studies provided evidence for the moderating role of type of harm (instrumental versus symbolic) in the relationship between culture and reaction to financial offenders. Study 1 showed that Greeks display moral outrage against a financial offender when the inflicted harm is symbolic as opposed to instrumental. Study 2 showed that British display moral outrage and endorse stronger conspiracy beliefs about an offender when harm is instrumental rather than symbolic. In Study 3, where culture was measured at the individual level, we found further support for Hypothesis 3 on moral outrage and conspiracy beliefs. Although these findings provide some support for Hypothesis 3, we did not find the culture by type of harm interaction effect on the rest of our variables.

These findings suggest that not all types of harm are viewed as equally detrimental in all cultures. Instead, people who value independence and egalitarian relationships experience instrumental harm as more detrimental and thus react more negatively to it. On the contrary, people who value interdependency and belonging to a group and at the same time accept social hierarchy experience symbolic harm as more detrimental and display more negative reactions to it. These results are in line with prior research showing that people might differ in the way they experience instrumental versus relational/symbolic needs (see Aaker, 1997, 1999; Highhouse et al., 2007; Ledgerwood et al., 2007, see also De Cremer & Van Vugt, 2002; Scheepers et al., 2003) and culture might be an important antecedent of people’s reaction to offenders who violate those needs. To the best of our knowledge, this is the first research investigating the effect of differential types of harm on people’s reaction to offenders; Further research is required to draw more solid conclusions regarding these results.

Practical implications

Apart from the theoretical implications, these findings have important societal and practical implications. First and foremost, the findings of this research are relevant to law practitioners who are regularly confronted with financial crimes and deal with powerful offenders who inflict either instrumental harm (monetary costs) or symbolic harm (reputation or image-related) to their victims. The notion that different types of harm have different implications for the victims depending on their cultural values might influence the type of punishment that is deemed fair for harm-doers. Law practitioners should be aware of the differential impact of instrumental versus symbolic harm to the victims when administering justice. Furthermore, power asymmetries and hierarchies are an integral element of work in organizations. Organizational practitioners should be aware of the differential reactions that financial, and thus powerful offenders (e.g., managers, CEOs etc.) might evoke to employees based on their cultural values. Moreover, organizational practitioners should be aware of the differential reactions that symbolic or instrumental harm may trigger to victims or observers of immoralities at work. For instance, based on our results, employees in organizations who value independence, uniqueness, and egalitarianism would react more negatively to harm that brings about financial implications (i.e., instrumental) whereas employees who value inclusion to groups and interdependency, and rationalize power asymmetries would deploy more negative reactions to harm that puts their image or reputation in jeopardy (i.e., symbolic harm). Finally, organizations should be aware of people’s differential ways (active versus passive) to respond to harm and immoralities; Although in some cultures people prefer direct and explicit ways to deal with immoral doers (e.g., punishment), in other cultures people form negative private beliefs which are not directly communicated to the involved members. Yet, these beliefs exist and could be translated into action retrospectively. We suggest that both law and organizational practitioners take these findings into consideration and design intervention programs that aim at combatting injustice in society in general and in organizations in particular.

Limitations and future directions

One limitation of our study is the question of how well these effects generalize across cultures and crimes. Specifically, our cross-cultural comparisons of Studies 1 and 2 are limited to participants coming from either the UK (horizontal individualists) or Greece (vertical collectivists) and therefore, we cannot infer that these effects generalize to other countries that differ in horizontal individualism versus vertical collectivism. More specifically, the financial background of Greece is different from that of the UK with the former country having experienced a severe financial crisis in the near past and the latter a Brexit with serious financial ramifications. Moreover, besides the cultural differences between the two countries, Greece (similar to many other south European countries like Italy, Cyprus, Spain) scores higher in corruption and fraud as compared with the UK (and other northern European countries, such as the Netherlands and Scandinavian countries). Although tolerance to fraud and corruption is linked to cultural values (Bierstaker, 2009), we cannot conclude with certainty what underlying features of these country’s cultures are driving the findings of this study. Boehnke et al. (2011) point out this kind of limitations in cross-cultural research and suggest “purposive sampling” (i.e., selection of samples that satisfy criteria that are related to the purpose of a study than the selection of representative samples) as a tool to overcome these limitations. Yet, purposive sampling relies on the judgment of the researcher and is therefore prone to researcher bias (Sharma, 2017).

Another methodological limitation of the current study is that the recruitment strategies we used for the data collection in the two countries differed. Although British participants were recruited online via a crowdsourcing platform (and were paid for their participation), Greek participants were approached via the researchers’ network (and were not paid). It is therefore likely that the different recruitment strategies might have influenced the data that we obtained including the lack of scalar invariance.

Furthermore, although British are considered as horizontal individualists and Greeks as vertical collectivists (Hofstede et al., 2010), Studies 1 and 2 do not provide evidence about the cultural values of Greeks and British as measured by culture scales at the individual level. This latter concern is mitigated by Study 3, of course, by showing consistent effects with individual level-measures of these cultural dimensions. Yet, more evidence is needed to disentangle the complex relationships between culture and (active versus passive) reactions to financial offenders. It is worth mentioning that some prior research on cultural dimensions has not always found consistent results between the individual versus country level (Bobbio & Sarrica, 2009). Such discrepancies (i.e., culture, when operationalized at the individual level produces sometimes different outcomes compared to when it is operationalized at the country level) have been attributed to two reasons: a) The assessment of cultural values across nations is not always feasible because the existing measures fail to accurately capture the intended constructs (Van de Vijver & Hambleton, 1996); and b) Participants’ characteristics always interfere with their cultural values, blurring the role of culture on individuals’ behavior (Bobbio & Sarrica, 2009; Gouveia et al., 2003). This contribution hence needs to be seen as a preliminary step toward a more fine-grained understanding of the relationship between culture and people’s responses to crime. Future research needs to include a broader range of countries, offenses, and measures, in order to better understand the cross-cultural effects on people’s reactions to offenders.

Nevertheless, we deem it a strength of our study to investigate the effects of culture both between cultures and within a culture, yielding largely consistent results. Furthermore, the studies presented here make three novel contributions to further understanding the relationship between culture and citizens’ responses to financial offenders. First, we shed light on the effects of culture as a multidimensional variable (i.e., a combination of individualism-collectivism with power distance) on people’s reaction to financial (and thus powerful) offenders, a topic that has never been investigated before. Second, our findings underline that cultures may differ in their preference for active versus passive responses to financial offenders. Indeed, these differential preferences may be related to broader cultural styles of how to cope with injustice, such as the dilemma between being confrontational versus saving face. Third, our studies shed light on the differential value that people place on different types of harm (i.e., symbolic versus instrumental) depending on their cultural background. This distinction was not recognized before in cultural research, yet it stands to reason that it is important given how cultures differ in their emphasis on individual versus group concerns.

Concluding remarks

Financial crime occurs in all cultures and powerholders usually have a protagonistic role in such crimes (see Gregg, 2015). In this research, to illuminate people’s reactions to financial offenders, we conceived individualism-collectivism not as a bipolar dimension but as a multidimensional variable that, when combined with power distance (Gelfand et al., 1996), can paint a clearer picture of culture’s effects on people’s punishing reactions to financial offenders. Moreover, we distinguished between active versus passive responses to crime, and instrumental versus symbolic types of harm. By showing that both of these distinctions matter, the studies presented here underscore some of the subtle complexities that need to be appreciated for a full understanding of how culture shapes citizens’ responses to white-collar criminals.

Compliance with ethical standards

This research involves human participants. All procedures performed in this study were in accordance with the ethical standards of the institutional research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. For the current research we got approval from the Ethical Committee of the Free University of Amsterdam; Veste Commissie Wetenschap en Ethiek (Scientific and Ethical Review Board), Nr VCWE-2017-178.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Data and Online Supplemental Materials are available from the Open Science Framework at: https://osf.io/hptkc/?view_only=64ebec98af7843d184e5a123ccaf8991

Notes

1. It is noteworthy that the data collection from Greek participants took place at the end of 2019 and therefore it did not overlap with the turbulences of the Greek economic crisis (2008–2018; VoxEU.org, 2020).

2. Apart from the direct comparison between countries (i.e., operationalizing culture at the national level), in both Studies 1 and 2 we also measured culture at the individual level using the scale of Triandis and Gelfand (1998). However, some of the measures consistently yielded very low reliability levels, presumably due to the measurement in two different cultures. For this reason, we do not include the measure of cultural values at the individual level in Studies 1 and 2.

3. Prior research has shown that punitive responses between second (victims) and third (observers) parties differ such that second-party punishments are consistently higher than third-party punishments (Fehr & Fischbacher, 2004). Therefore, we deemed important to include this variable in our design and investigate whether taking the perspective of an observer or a victim would produce different results. Accordingly, in Study 3 we had participants taking either the perspective of an observer or the perspective of a victim of a fraudulent case. We did not observe significant differences between victims and observers. In order to avoid unnecessary complexity of the manuscript, we decided to drop this manipulation from our method and results sections.

4. Although the focus of the study lies on vertical collectivism and horizontal individualism only, we assessed all four cultural dimensions (e.g., vertical individualism and collectivism, and horizontal individualism and collectivism) keeping the scale of Triandis and Gelfand (1998) intact, and report the results for all four cultural variations.