Everyday sadism in the business area

Purpose – The purpose of this study is to identify sociodemographic factors that are predictive of the level of everyday sadism (SAD) in the business area. Design/methodology/approach – A survey was conducted on 424 graduate and postgraduate students from business schools in Brazil and the USA. SAD was quantified by the assessment of sadistic personality proposed by Plouffe Saklofske and Smith (2017). The variables included age, gender, managing experience, education and nationality. Findings – The average level of SAD was low. SAD was negatively associated with gender, age and nationality and positively associated with managing experience and education. Practical implications – As individuals ascend professionally and academically, they display higher levels of everyday sadism. Depending on the context, dark personalities can cause either benefit or harm to the company’s business and to society. However, the literature shows that seeking pleasure and dominance with no regard for consequences affects the business area directly or indirectly. Originality/value – Very few studies have addressed everyday sadism in the business area, let alone evaluated predictive factors and discussed possible implications.


Introduction
Firms possess tangible and intangible assets used and managed by human capital (staff and managers) to achieve specific goals. Human capital is unique and will influence the entire organization in unique ways. According to Hambrick and Mason (1984), corporate results reflect the cognitive biases and values of managers, who inevitably make decisions from a position of bounded rationality. and processes (Epstein & Ramamoorti, 2016) and mitigate negative effects. To help them do so, we attempt to answer the question: Q1. What individual traits are predictive of the level of everyday sadism in the business area? Hambrick and Mason (1984) claimed managers' behavior was primarily defined by gender, age, managing experience, education and nationality. These aspects may also explain differences in the intensity of individual traits (G ois, 2017; Meere & Egan, 2017;Plouffe, Saklofske, & Smith, 2017). Thus, the main objective of this study was to identify factors potentially determining the level of everyday sadism in organizations.
Our study is relevant, in that it addresses a topic which has received little attention in the corporate sphere while using recently developed metrics to quantify sadism. Our search of the literature yielded no previous studies on the impact of everyday sadism on business administration, let alone studies considering a range of cultural contexts. Most research on sadism is in the field of psychology and focuses on sexual disorders, whereas our study was designed to explore the social effects of everyday sadism on business school students and the association between sociodemographic factors and levels of sadism.
Moreover, as pointed out by Trémolière and Djeriouat (2016), research in this field is important owing to the questionable social usefulness of sadismarguably the only socially aversive trait which offers no advantage in terms of accomplishing goals. Unlike narcissistic behaviors (which are driven by a desire for superiority, power or fame), sadistic actions reflect no specific objectives. In fact, sadistic individuals may seek pleasure and dominance to the point of completely disregarding the consequences of their acts, directly or indirectly compromising corporate performance. Cattell (1973) distinguished two basic components in personality: common traits and unique traits. As shown in the literature, common personality traits fall into five categories or dimensions referred to as "the Big Five" (Goldberg, 1990;Judge, Higgins, Thoresen, & Barrick, 1999), namely, extraversion, neuroticism/emotional stability, agreeableness, conscientiousness and openness. On the other hand, unique personality traits are peculiar to specific individuals and are scarcely distributed in the general population. Among the unique traits, in this study, we focus on the "dark traits" which underlie "Dark Tetrad" personalities, characterized by callousness, impulsivity, manipulation, criminality, grandiosity, enjoyment of cruelty and misconduct (Paulhus, 2014;Paulhus & Jones, 2014;Southard et al., 2015).
Machiavellians are calculating and manipulative. Strongly Machiavellian individuals may have a cynical worldview, lack morality and be risk-averse (Jones, 2013;Paulhus, 2014;Southard et al., 2015). They are not typically impulsive or aggressive, but prefer to engage in long-term thinking and in the building of alliances to protect their reputation (Jones, 2013;Paulhus & Jones, 2014).
Sadists enjoy inflicting pain on others or witnessing cruelty and humiliation, whether the abuse is physical, sexual or psychological. Such behaviors make sadists feel empowered and are a source of morbid pleasure (Buckels, Jones, & Paulhus, 2013;Meere & Egan, 2017;O'Meara et al., 2011).

Sadism
Sadism is as malevolous as psychopathy. The enjoyment of cruelty can take many forms, but criminal and sexual expressions generally receive the most attention (Paulhus & Dutton, 2016).
The word "sadism" has its origin in the name of the Marquis de Sade (Donatien Alphonse François), a revolutionary and libertine novelist who depicted sexual fantasies with emphasis on pain and humiliation (Brame, 2015;Paulhus & Dutton, 2016).
The DSM-III-R (Diagnostic and Statistical Manual of Mental Disorders; American Psychiatric Association, 1987) introduced "sadistic personality disorder" as a category with eight criteria focusing on dominance and subjugation (Nell, 2006), but in the following edition, the term was replaced with the more specific term "sexual sadism disorder." O'Meara et al. (2011, p. 523) define sadists as individuals who intentionally humiliate others, act in a consistently cruel and demeaning manner or inflict physical, sexual or psychological suffering on others affirm their position or simply for the sake of enjoyment. Pinker (2011) described a milder and ubiquitous form of sadism, a subclinical personality trait referred to as "everyday sadism" (Paulhus & Dutton, 2016). Everyday sadists derive self-affirmation or pleasure from cruel, humiliating or aggressive behaviors (Myers et al., 2006;O'Meara et al., 2011), such as may be observed in the workplace.
Everyday sadism may be direct or vicarious, depending on the sadist's role in the interaction (Paulhus & Dutton, 2016). Thus, a direct sadist enjoys inflicting pain personally, while vicarious sadists prefer to watch cruelty being committed. Moreover, according to Buckels et al. (2013), direct sadism may be either physical or verbal. The harm caused by the latter can be more severe and long-lasting than that caused by the former.
Earnings management and fraud are common in firms led by "Dark Tetrad" personalities, leading to poor earnings quality (G ois, 2017). Auditors with sadistic traits are inclined to make distorted judgments and unethical auditing decisions (G ois, Lima, & De Luca, 2018).

Methodology
The study population consisted of 424 undergraduate and graduate students from Brazilian and US business schools who agreed to answer an on-line survey. In comparison, many international studies have used samples of undergraduate and graduate students with results very similar to those of studies based on samples of professionals (Lakey, Rose, Campbell, & Goodie, 2008;Majors, 2016;Paulhus & Jones, 2014;Paulhus & Williams, 2002).
On the other hand, Fréchette (2015) believes studies based on students and professionals can yield discrepant results, depending on the study objective. In fact, the use of student pools can introduce selection bias, with loss of external validity. Arguably, the results of carefully designed studies based on students may be extrapolated not only to the universe of students (Smith, 2003), but in some cases to professionals as well (Fréchette, 2015).
The survey included three parts: an informed written consent form, questions about demographics and questions about everyday sadism. The second part collected information on gender (male/female), age (years), managing experience (years), level of education (undergraduate vs graduate) and nationality (Brazil vs other countries).
The assessment of sadistic personality (ASP) proposed by Plouffe et al. (2017) was used to quantify subclinical everyday sadism. The questionnaire contains nine statements illustrating behaviors related to subjugation, pleasure-seeking and lack of empathy (Table II).
The participants were asked to assign a score from 0 to 10 to each statement (0 = completely disagree, 10 = completely agree). The internal consistency was evaluated by calculating Cronbach's alpha for each item. The average alpha value (0.8257) was relatively close to 1, indicating a good level of consistency (Hair, Anderson, Tatham & Black, 2005). The participant's overall level of everyday sadism (SAD) corresponded to the mean score assigned to the nine statements, as shown in equation (1): To meet the study objectives, we submitted our data to descriptive statistics, multiple correspondence analysis, Pearson and Spearman correlation analysis and multiple linear regression with robust errors. Descriptive statistics show the behavior of each variable through minimum, maximum, mean and standard deviation. Multiple correspondence analysis uses graphs to visualize associations between study variables. Pearson and Spearman correlation analysis reveals potential bivariate relationships between variables, indicating the first results of the study. Finally, multiple linear regression with robust errors allows identifying factors with a significant influence on the outcome variable (in this study, everyday sadism in the business area). According to G ois (2017), Hambrick and Mason (1984), Kagel and McGee (2014), Meere and Egan (2017) and Plouffe et al. (2017), demographic variables have an influence behavior and, consequently, the personality of the manager. For the purposes of this study, we used equation (2) to express the relationship between everyday sadism (SAD), gender (GEN), age (AGE), managing experience (EXP), education (EDU) and nationality (NAT): Before proceeding, the assumptions (normality, homoscedasticity and multicollinearity) for cross-sectional regressions were addressed (Fávero & Belfiori, 2017). To avoid homoscedasticity, we used White's correlation matrix with robust errors, as described by Fávero & Belfiori (2017), but we also used White's test (Table VI). Multicollinearity was tested by correlation analysis (Table V) and the variance inflation factor (VIF) analysis. The correlation between independent variables was similar to that between dependent and independent variables (mean VIF = 2.14) (Table VI), indicating the absence of multicollinearity. Studies in the social sciences often fail the normality test (Fávero & Belfiori, 2017). According to the law of large numbers (Wooldridge, 2002), samples with over 30 observations are more likely to pass the normality test. In other words, the larger the sample, the more normal the distribution. With 424 observations, our study was expected to have normally distributed residuals. However, even after performing a Box-Cox transformation of the variables, the residuals did not display normality. Table III is a summary of the data used to meet the study objectives, along with their measurement.
In the multiple correspondence analysis, SAD and AGE were converted into categorical variables using percentiles in three ranges. Thus, SAD was classified as either low, moderate or high, while AGE was divided into the ranges 17-23 years, 24-27 years and 28-62 years.

Analysis of results
The results are initially presented in the form of descriptive statistics (Table IV), followed by multiple correspondence analysis (Figure 1).
Despite the presence of highly sadistic individuals in the sample (maximum value: 98.3 per cent), the mean level of everyday SAD was low (18.5 per cent). This is compatible with the results of Plouffe et al. (2017), Plouffe, Smith, and Saklofske (2018) and G ois (2017), who reported mean levels of everyday sadism between 15.41 per cent and 29.30 per cent. When the SAD scores were distributed according to intensity, the mean percentages were 7.7 per cent (low), 14.9 per cent (moderate) and 33.8 per cent (high).
The participants were of age 26.64 years on average, with a slight predominance of the female gender (52.12 per cent). Over half were Brazilians (56.13 per cent), while those attending school in the USA were from the USA (13.21 per cent), China (13.21 per cent), India (7.08 per cent) or other countries (10.38 per cent). The mean time of managing experience was 1.15 years, but the range was wide (0-25 years).
The results of the descriptive statistics are complemented by the results of the multiple correspondence analysis (Figure 1). When the relationship between SAD and the sociodemographic variables was analyzed with the chi-square test, AGE, NAT and EDU  (2014) Everyday sadism in the business area were significant at the level of 1 per cent, GEN was significant at the level of 5 per cent and EXP was significant at the level of 10 per cent. Thus, most of the sociodemographic variables had a significant influence on the level of everyday sadism. Figure 1 shows how everyday sadism relates to the study variables. In our sample, low levels of sadism were associated with the female gender, little or no managing experience, youth (17-23 years), undergraduate status and Brazilian nationality. Craker & March (2016) found dark personality traits to be less common among women and young people, the latter in part justifying the association with lower levels of education and managing experience. As for nationality, countries with a predominantly collectivist and empathic culture, such as Brazil, may be expected to display lower levels of dark personality traits (Hofstede, 2011). Table V shows the associations observed between everyday sadism and demographic variables using Pearson and Spearman correlation analysis.
Most of the variables were associated with SAD in one way or another. The exceptions were EXP and, among the control variables, NAT and AGE.
The association between SAD and GEN was negative, meaning that the male gender displayed higher levels of sadism. This is supported by Plouffe et al. (2017) and G ois (2017), but not by Meere and Egan (2017) who found no association.
AGE was negatively associated with SAD, meaning that the younger the individual, the more sadistic he/she is. This matches the results of G ois (2017) but differs from the results of Meere and Egan (2017) who observed no association.
EDU was positively associated with SAD. Thus, the level of sadism increases as the individual ascends academically. As for the variable NAT, confirming the results of the multiple correspondence analysis (Figure 1), Brazilians were less sadistic than non-Brazilians. Table VI shows the relationship between everyday sadism and the demographic variables, with an indication of the statistical significance of each association.
The model is significant down to the level of 1 per cent (F test), with the study variables representing 20.1 per cent of the variance of the level of sadism. This indicates the influence of other factors on SAD.  Notes: *** = significant at the level of 1%; ** = significant at the level of 5%; * = significant at the level of 10%; a = Pearson correlation coefficients; b = Spearman correlation coefficients Notes: *** = Significant at the level of 1%; ** = significant at the level of 5% Everyday sadism in the business area EXP was the only variable not significantly predictive of everyday sadism. Nevertheless, other studies have found a positive and significant association between sadism and the amount of managing experience, favoring the emergence of destructive corporate leaders inclined to engage in corruption and fraud to the detriment of organizations, the capital market and society (D'Souza and Lima, 2015;G ois, 2017;Krasikova et al., 2013;Umphress & Bingham, 2011).
Confirming the results of the Pearson correlation analysis (Table V), EDU was positively associated with SAD. In other words, participants with higher academic degrees were generally more sadistic. This may be evidence of the effect of mental and emotional stress associated with highly taxing postgraduate activities. According to several authors, the prevalence of mental disorders (especially depression and suicidal ideation) is high among postgraduate students owing to the psychologically toxic university environment (Evans, Bira, Gastelum, Weiss, & Vanderford, 2018;Levecque, Anseel, De Beuckelaer, Van der Heyden, & Gisle, 2017). It is therefore not unreasonable to assume that this type of experience predisposes towards the manifestation of dark personality traits.
GEN was not significantly associated with SAD, matching the results of Meere and Egan (2017) who likewise found no correlation.
As in the Pearson correlation analysis (Table V), AGE was negatively associated with SAD. Thus, the younger the participant, the higher the level of everyday sadism. This is supported by G ois (2017), but not by Meere and Egan (2017) who observed no correlation.
The observed negative association between NAT and SAD shows that the Brazilian participants were less sadistic than other countries' participants. This may be explained by cultural factors. According to Hofstede Insights (2018), Brazilians are predominantly collectivist people, that is, by nature, Brazilians are more empathic and less anti-social than many other nationalities.
To test the results for robustness, we estimated a model less sensitive to the assumptions of multiple linear regression (quantile regression) and one in which the dependent variable was truncated (Tobit regression). We also estimated a probit model and a logistic model by transforming the dependent variable into a binary variable, where 1 corresponded to high sadism (above the third quartile) and 0 otherwise. Table VII shows the four regression models used to test for robustness (1 = quantile regression; 2 = Tobit regression; 3 = logistic regression; 4 = probit regression).
The tests showed our results to be consistent, except GEN, which was non-significant in the previous analysis but was significant in all four models used to test for robustness. Regardless of this change, it should be pointed out that GEN was negatively (though not significantly) associated with SAD in the regression in Table VI, suggesting females are less prone to sadism than males, as observed by Plouffe et al. (2017) and G ois (2017), but contradicting Meere and Egan (2017), according to whom sadism displays no gender preference. Table VIII synthesizes the results of our study. In other words, sociodemographic characteristics (gender, age, education, nationality) influenced the level of everyday sadism expressed by the participants of this study. Interestingly, the association observed between the level of higher education and sadism suggests that individuals with higher academic degrees are likely to be more sadistic. Because sadism goes hand in hand with aggressiveness, impulsivity and bullying (Bates et al., 1991;Ferris and Grisso, 1996), high levels of sadism tend to produce destructive leaders who are likely harm the business strategy.

Final considerations
In this study, we looked at factors potentially influencing the level of everyday sadism in the business area. To do so, we surveyed 424 undergraduate and post-graduate students attending business schools in Brazil and the USA. Everyday sadism was quantified with the assessment of sadistic personality proposed by Plouffe et al. (2017).
Despite the presence of highly sadistic individuals in our sample, the average level of everyday sadism was relatively low, approaching the levels reported in the psychology literature for different populations (adolescents, adults, military service members, etc.).
Nationality, age and education were the most significant explanatory variables. Thus, being Brazilian, over 28 years of age and/or undergraduate were predictive of lower levels of sadism. The robustness testing showed gender to have an influence as well, with sadism being less intense among females.
Thus, everyday sadism, defined as taking pleasure in inflicting pain and humiliation, was observed in the business area, modulated by a range of sociodemographic factors. This finding has implications for the corporate setting as the detection and understanding of sadistic traits among staff can help organizations develop internal controls capable of Notes: *** = Significant at the level of 1%; ** = significant at the level of 5%; * = significant at the level of 10% Table VIII.

Everyday sadism Result
Gender -Age --Managing experience Education þþ Nationality --Notes: A single sign (À or þ) indicates that the result was observed in at least one model, but not in all models. A double sign (À or þþ) indicates that the result was observed in all models mitigating the impact of sadistic behaviors, especially as everyday sadism is subclinical and therefore present to a greater or lesser degree in all individuals. Internal controls should focus on corporate ethics and contribute to the development of a sound corporate culture capable of inhibiting abusive behaviors, earnings management and fraud (G ois, 2017;G ois et al., 2018;Karandikar et al., 2018;Trémolière & Djeriouat, 2016).
Given the above, organizations are advised to include assessments of everyday sadism in their recruitment processes and to monitor this trait over time. When highly sadistic individuals gain access to leadership positions, they are prone to create a toxic work environment, resulting in damage to the organization and staff (Krasikova et al., 2013). In the long run, this is deleterious for organizations which prioritize corporate efficiency, profitability and business perpetuation.
On the other hand, dark personality traits have also been associated with benefits in certain scenarios (Zeigler-Hill & Marcus, 2016). For example, highly sadistic executives are by some considered vigorous leaders who will efficiently protect the organization in times of crisis, even if this has adverse effects on coworkers and other stakeholders.
Our study is one of very few addressing everyday sadism in the business area. Little is known about the factors which determine the level of sadism among professionals, making our findings a relevant contribution to this area of study.
The study is limited mainly by the use of business school students to build the sample. While some may question this choice, the literature defends the use of students (Cohen, Pant, & Sharp, 2001) in this type of survey. In fact, student-based investigations have been shown to yield results compatible with those of studies based on samples of professionals (Lakey et al., 2008;Majors, 2016;Paulhus & Jones, 2014;Paulhus & Williams, 2002).
Considering the relevance of personality traits (especially everyday sadism) to corporate culture and performance, and the scarcity of studies on sadism among corporate executives (Spain, Harms, & Wood, 2016), we suggest investing in research evaluating the association between sadism and destructive leadership, toxic work environments and earnings management. Such investigations, we believe, are likely to confirm the arguments and results presented in this paper.