Do (gendered) ageism and ethnic minorities explain workplace bullying?

Abstract

Purpose

This study aims to answer two research questions: first, to what extent can workplace bullying be explained by ageism? And second, does the likelihood of workplace bullying increase when age interacts with gender and ethnic minority?

Design/methodology/approach

The authors report results from a survey carried out in 11 organizations in Estonia (N = 1,614) using the Negative Acts Questionnaire-Revised (Einarsen et al., 2009).

Findings

The results show that ageism does not explain bullying in Estonia. As in some earlier studies, older age correlates negatively with negative acts, and women report less work-related bullying than men. These findings were unexpected because Estonia's post-socialist background and the highest gender wage gap in Europe suggested otherwise. However, there is gendered ageism in work-related bullying such that older women report more negative acts in their workplace. Respondents from ethnic minority groups do not experience more bullying in general, nor in combination with age. Surprisingly, managers reported both person- and work-related bullying more than employees with no subordinates.

Originality/value

The study contributes to intersectionality literature with a view to workplace bullying in post-socialist study context.

Citation

Jaakson, K. and Dedova, M. (2023), "Do (gendered) ageism and ethnic minorities explain workplace bullying?", International Journal of Manpower, Vol. 44 No. 9, pp. 199-215. https://doi.org/10.1108/IJM-10-2022-0492

Introduction

Extension of working life and high employment rates among older employees are inevitable consequences of population aging. However, one obstacle to the prolongation of fulfilling working lives may be workplace ageism, defined as “systematic stereotyping and discrimination against people because they are old” (Butler, 2008, p. 25). The United Nations report on ageism claims that every second person in the world has moderate or highly ageist attitudes (WHO, 2021). This paper focuses on a specific form of discriminatory behavior, a serious interpersonal deviance, workplace bullying (Mackey et al., 2021).

Ageism and age-related bullying often stem from negative stereotypes related to age. Older employees are considered to have lower motivation, productivity, adaptability, flexibility, willingness to accept change, and ability to train and learn (Posthuma et al., 2012; Harris et al., 2018). Younger employees experience ageism because they are considered less experienced, less responsible, and less devoted (Jyrkinen and McKie, 2012).

Intersectionality literature (Crenshaw, 1989; Jyrkinen and McKie, 2012; McBride et al., 2015) postulates that people embody multiple social characteristics simultaneously. Combinations of those characteristics – gender, age, ethnicity, social class, etc. – constitute different subgroups with unique lenses to experience social reality. Working class black women's lived experience cannot be attributed to black community in general, let alone middle-class white men (Holman and Walker, 2021). Thus, ageism may be fueled by other generic characteristics signaling a person's vulnerability. Studies have investigated how gender, race, ethnicity, or disabilities interact with age at a workplace (Jyrkinen and McKie, 2012; Potter et al., 2019; Thomas et al., 2014), but in general, empirical work on intersectionality is scarce (McBride et al., 2015; Potter et al., 2019) and often limited by specific geographies such as the USA or the UK Age, in particular, has largely been neglected in intersectionality literature (Holman and Walker, 2021). Furthermore, there is relatively little research on workplace bullying using the intersectionality lens – Gardner et al. (2020) and Patel et al. (2022) studies are rare exceptions. This paper fills this gap and looks at gender and ethnic minorities in tandem with age. Focusing on intersectionality is probably more helpful in figuring out how to design prevention efforts to safeguard vulnerable groups of the workforce. Hence, we aim to answer the following research questions: a) whether and to what extent can workplace bullying be explained by ageism, and b) does the likelihood of workplace bullying increase when age interacts with other individual generic characteristics, namely, gender and ethnicity?

Empirical studies have reported regional variations in workplace ageism dissemination (e.g. Jyrkinen and McKie, 2012; Rippon et al., 2015; Trusinová, 2014). Reports convey more positive attitudes toward older employees in Nordic countries than in southern and Eastern Europe (Salomon, 2012). We use survey data from Estonia, an exciting study site because, on the one hand, the country is gender-egalitarian and age-friendly when looking at employment rates. On the other hand, it has the highest gender wage gap in Europe, discriminatory practices towards minorities, and, as in other post-socialist countries, may exhibit ageist attitudes in society. To our knowledge, Estonia has never been in the focus of intersectional literature in relation to workplace bullying.

This paper is organized as follows: first, we define workplace bullying and then provide a literature overview of the age dimension and its interactive effects on workplace bullying to develop hypotheses. Subsequently, we describe the data and methods, followed by the results. Discussion, limitations, and suggestions for further research are presented.

Workplace bullying

Bullying is defined as “situations where a worker or supervisor is systematically mistreated and victimized by fellow workers or supervisors through repeated negative acts. To be a victim of such bullying, one must also feel inferiority in defending oneself in the actual situation” (Einarsen and Skogstad, 1996. p.185). Unfortunately, bullying is an understudied yet widespread phenomenon (Léon-Perez et al., 2021).

Bullying may take different forms: social exclusion, humiliation, or verbal abuse. The typical characteristic is that the treatment is unpleasant, offensive, and humiliating for the victim, it is not a one-off act, and there is a power imbalance between the parties involved (Einarsen et al., 2009; Branch et al., 2013; Gardner et al., 2020). There are several ways to categorize workplace bullying: direct and indirect, psychological and physical (Escartin et al., 2011b). One can also observe particular behaviors: isolation, control and manipulation of information, abusive working conditions, emotional abuse, professional discredit and denigration, and devaluation of the professional role (Escartin et al., 2011a). In this paper, we follow Einarsen et al. (2009), who distinguish between personal bullying, work-related bullying, and physically intimidating forms of bullying. Because bullying and workplace discrimination have a large amount of overlap, we occasionally use concepts from studies on workplace discrimination to inform our hypotheses (Striebing et al., 2023). The fundamental distinctions between the two are that bullying is a persistent behavior whereas discrimination can occur only occasionally, and that discrimination is based on primary identity characteristics while bullying victims might be diverse. Our focus on gender, age, and ethnicity, however, is also prominent in discrimination literature.

Age-related bullying

Age represents an important form of inequality in organizations (Meliou and Mallett, 2022), but organizational studies have long ignored it (Thomas et al., 2014). As a vulnerable group in the labor market, older workers may experience bullying because of health-related constraints, obsolescence of skills and underestimation of experience (Taylor et al., 2016). For example, the mean age of bullied employees was two years higher than that of non-bullied employees in a Swedish study (Hansen et al., 2006). However, Duncan and Loretto (2004) point out that as every individual is of an age, he or she can be both a victim and perpetrator and can experience hostile behavior by other people of the same age.

The first scholar to introduce the term “ageism,” Robert Butler, argued that younger and older age groups become the target of discrimination because they are more dependent on middle-aged groups (Butler, 1969). Nevertheless, there is also a notion of middle ageism (Gullette, 1998) – ageism directed at those in their middle years. The most vulnerable age group to workplace bullying is 40 years and older (Powell, 2010), yet co-occurrence studies on age-related bullying are scarce and inconclusive. The US-based research by Harnois (2015) reveals that it is the youngest and oldest cohort of employees who equally perceive age-based mistreatment at work. In Finland and Scotland, both youth and seniority were perceived as a hurdle in the labor market (Jyrkinen and McKie, 2012). In contrast, Potter et al. (2019) found that perceived discrimination decreased with age. Similarly, in the hospitality industry, the youngest cohort of employees systematically reported more mistreatment than older colleagues (Ariza-Montes et al., 2017). Thus, empirical studies show that the expression of workplace ageism finds different objects and largely depends on local context and industry.

Estonia is a fascinating study context for this kind of research. First, according to OECD statistics on age group employment, almost 74% of the Estonian population aged 55–64 years were employed in 2022 according to OECD. This figure is the fourth-highest rate in Europe (after Iceland, Sweden, and Norway). Since 2018, this indicator has increased by 3%. There is no compulsory retirement age in most fields of activity, but the state pension age increases gradually from 63 to 65 years by 2026. Average state pension without individual's voluntary contribution is about half the average salary, which partly explains high elderly employment rates. Indeed, Estonia shows exceptional employment rates for the age group 65–74: 28% for men and 25% for women.

The only hints about workplace ageism in Estonia stem from comparative analyses (Trusinová, 2014; Zhang and Gibney, 2019). Results are mixed. For example, Trusinová (2014) suggests that countries with long democratic traditions, such as Western and Northern European countries, are more sensitive to manifestations of ageism. Her findings reveal that ageism toward older people is the highest in the Czech Republic and Russia. The lowest figures are in Norway, Switzerland, and the UK. Estonia performs moderately, occupying the eighth position out of the 16 countries. According to the Special Eurobarometer 2015, age-based discrimination was even more prevalent in the UK than in Estonia; unfortunately, Norway and Switzerland were not in the sample. In a more recent study, Zhang and Gibney (2019) studied how workplace ageism affects employees' perceived job sustainability in 28 EU member states. Job sustainability refers to the perspectives of employees to keep their current job or a similar role as they get older. The authors classify Estonia as a post-socialist liberalist country with Bulgaria, Romania, Lithuania, and Latvia. The characteristics of this group are a flexible labor market and a small number of policies aimed at employment protection. Surprisingly, respondents in the post-socialist liberal regime group and social-democratic regime countries (Denmark, Finland, the Netherlands, and Sweden) reported very similar experienced ageism, 4.2% and 4.4%, respectively, and these were the highest shares. As expected, ageism negatively influences perceived job sustainability.

Given the evidence above, the question arises whether ageism plays a role in workplace bullying and what age groups are subject to mistreatment. We propose the following hypothesis.

H1.

Both the youngest and oldest age groups perceive greater exposure to bullying behaviors than the middle-aged group.

Gender dimension of ageism

The linkage between gender dimension and bullying stems from two theoretical foundations: stereotypes (e.g. Harnois, 2015) and organizational power (e.g. Gardner et al., 2020; Simpson and Cohen, 2004). Access to organizational power structures is traditionally more granted to men, which is why fewer women are found in leadership positions or in high-paid jobs. Masculine bias of organizational power prevents women's career progression (Hallward and Bekdash-Muellers, 2019; Jyrkinen and McKie, 2012). As bullying assumes power inequality, women tend to be the targets of bullying and mistreatment more likely than men (Gardner et al., 2020; Harnois, 2015; Moreno-Jimenéz et al., 2008; Simpson and Cohen, 2004), or at least they perceive certain behaviors as threatening or unwelcome (Escartin et al., 2011a; Simpson and Cohen, 2004). However, according to studies in the USA and UK, men face higher levels of discrimination than women (Rippon et al., 2015; Potter et al., 2019; Patel et al., 2022) and a meta-analysis of 66 studies on interpersonal deviance by Mackey et al. (2021) marginally produced the same result.

For women, older age has even more severe implications. “There is much to suggest that the discourse of ageism has a worse effect on midlife and older women, who are particularly stigmatised due to other forms of discrimination in work organisation” (Thomas et al., 2014, p. 1573). “Gendered ageism” was coined by Itzin and Phillipson (1995). Gendered ageism is defined as double jeopardy, which highlights a combination of patriarchal norm prevalence and preoccupation with youth, leading to a quicker depreciation of older women's status than men (Jyrkinen and McKie, 2012). When Duncan and Loretto (2004) studied perceptions of ageist attitudes in the USA, they discovered that until the age of 45, the responses of men and women were similar, but in the category of 45+, significantly more women reported being victims of discriminative attitudes. However, Meliou and Mallett (2022) call for more studies on age and gender because ageism at work is rarely gender-neutral.

Gender and age are associated with discrimination, but various studies have revealed different directions and strengths of the associations. In the EU, older women experience more significant discrimination than men. In the USA, these findings are not supported (Potter et al., 2019). Lössbroek and Radl (2019) explored gender differences in older employees' training participation in nine European countries. They found that Eastern Europe (Hungary, Bulgaria) is susceptible to gendered ageist stereotypes. Moreover, the study results show that older women more often become the target of managerial ageism by excluding female employees from the available training opportunities compared to male employees. It is relevant to note that female-caretaker stereotype still prevails (Haškova and Dudova, 2020; Jyrkinen and McKie, 2012), but it is not neutral with respect to age. Younger women are not equally valued in the workplace because they are expected to take care of children. Older women, in turn, often need to take care of their parents, because nursing homes are inaccessible to many. In New Zealand, while female employees more frequently self-categorized as having been bullied, no multiplicative effect of age was found (Gardner et al., 2020).

There is little evidence of gender-related workplace bullying in Estonia. Pörhöla et al. (2020) studied university students' perceptions of bullying in four countries, including Estonia. According to their results, female students were bullied 7% more than men. In the workplace context, despite Estonia's high female employment rate, it had the most significant gender wage gap (21.7%) in Europe in 2019, according to Eurostat (2019). Work-family institutions, on the other hand, are the most generous amongst the OECD countries–total paid maternity leave in weeks is about three times longer than in Germany or Canada (Masso et al., 2022), reflecting pronatalist policy professed by Eastern European governments in general (Haškova and Dudova, 2020). Low union presence and weak collective bargaining practices stipulate the dominant influence of within-firm bargaining and other firm-level factors on the gender wage gap (Masso et al., 2022). Inequality in wage policy may cause or result from other malpractices at work. Based on these findings, we put forward the following hypotheses:

H2a.

Women perceive greater exposure to bullying behaviors compared to men.

H2b.

Gender moderates the relationship between age and workplace bullying, such that older women perceive more bullying behaviors than older men.

Ethnic minority dimension of ageism

As suggested by the overview by Branch et al. (2013), minority groups are more likely to be bullied. There are various theoretical explanations why this happens, including social identity theory, self-categorization theory, similarity-attraction paradigm, cultural distance hypothesis, and extension of social stratification to workplace setting (Bergbom et al., 2015; Patel et al., 2022). Again, organizational power comes into play: minorities rarely occupy powerful structural positions and thus bear greater risk of bullying (Branch et al., 2013). By adding minority status to ageism, we respond to Thomas et al. (2014) call to examine the interactive effects of belonging to multiple vulnerable groups.

Potter et al. (2019) showed that black individuals reported more mistreatment than white participants, but the interaction between age and race was insignificant in predicting reported discrimination scores. Fekedulegn et al. (2019) disagreed: in their study, middle-aged and older black employees experienced highest workplace discrimination and mistreatment. In Finland, immigrant status generally predicted higher reports of bullying. However, the study's authors also note that the youngest cohort of immigrant employees experienced more bullying than other age groups (Bergbom et al., 2015).

Following the above findings, we suggest that:

H3a.

Belonging to the ethnic minority group positively relates to perception of exposure to workplace bullying behaviors.

H3b.

Belonging to ethnic minority groups moderates the relationship between age and workplace bullying, such that older minority employees perceive greater exposure to bullying behaviors than older majority employees.

We test the following framework for the study based on the hypotheses above (see Figure 1).

Methodology

Sample

The data were collected in Estonia between March 2018 and November 2019 as part of a larger project. At first, 88 organizations were selected to take part in the study to represent different sectors, firm sizes, and regional locations. In each firm (Human Resource, HR) managers were contacted with the invitation to participate. 14 organizations agreed and the surveys were disseminated via an electronic platform to ensure the confidentiality of respondents. Two reminders were sent by the (HR) managers to their employees, but no incentives were offered and participation was voluntary for the employees. However, in three organizations, some important control variables were missing, so our final sample is 11 organizations. The organizations were both public and private, manufacturing, and service. The data comprised 1,614 observations with a minimum of 21 to a maximum of 520 responses per organization. Table 1 presents the characteristics of the samples.

Measures

We used the Negative Acts Questionnaire-Revised (NAQ-R) by Einarsen et al. (2009) to measure bullying. This instrument explores self-perception of bullying in the workplace by asking the respondents on a 5-point Likert scale from 1 (“never” to 5 “every day”) how often certain negative acts have taken place during the past six months. The questionnaire contained 22 negative behaviors. Example items were “Being ordered to do work below your level of competence” and “Having your opinions ignored”. We obtained Estonian and Russian versions of NAQ-R from Tambur and Vadi (2009) via direct contact with the authors. Both versions had been translated and back-translated using an English questionnaire.

NAQ-R originally divided bullying into three dimensions: work-related, person-related, and physically-intimidating. Similar to Einarsen et al.’s (2009) results, Cronbach's alpha for 22 items was very high (0.92), indicating internal consistency and co-occurrence of negative acts. Kaiser-Meyer-Olkin Measure (KMO) was 0.951, which clearly suggests reducing the data to factors. Eigenvalues of Principal Component Analysis (PCA) revealed three components as proposed by the Einarsen et al. (2009) model. However, the parallel analysis discards the third component as the third randomly generated Eigenvalue (1.15) exceeded our PCA Eigenvalue (1.12). Earlier studies have shown that physical component in workplace bullying in Europe is not significant (Escartin et al., 2011b), and this was also the case on Estonian data. Additionally, the three items for physical intimidation showed rather low Cronbach's alpha (0.60) and we omitted them from the study.

Given the nature of the phenomenon we assume that the two components are correlated and therefore used promax rotation method to obtain factor loadings (see Appendix). These can be named work-related bullying (six items), and person-related bullying (10 items).

We operationalized ageism with three dummy variables. First, a Younger age dummy was created such that it took the value “1” when employees were up to 30 years (17% of the sample) and “0” if older. Second, an Older age dummy took value “1” if employees were 56 years or older (18% of the sample) and “0” if younger. The choice of 56 years benchmark can be justified by the Soviet legacy in Estonia, whereby for decades, women retirement age was 55. Although the retirement age has been gradually raised since 1994, for many people the cognition of retirement age is anchored in the Soviet era. Ages 31–55 formed middle category (65% of the sample).

We defined the Minority dummy by assigning value “1” to employees whose childhood language at home was other than Estonian, “0” for Estonian. Stemming from the country's Soviet history, Russians made up most of the ethnic minority in the sample: 306 out of 320, and altogether the minority formed a little less than 20% of the sample.

We included both individual-level, organizational level, and sector level control variables. First, employee position (manager “1”, employee “0”) was included as a control variable. The level of education was significantly correlated with the position (managers are more educated) and gender (females are more educated); therefore, we do not include education in our models. Previous studies have confirmed that the social climate of the work environment plays a decisive role in bullying (Hansen et al., 2006; Ariza-Montes et al., 2017; Mackey et al., 2021). Therefore, an assessment of supportive work environment was included. We measured the supportive work environment of the organization using four items (Cronbach's alpha = 0.87). These were “In our organization, we communicate politely,” “Relationships between colleagues in our organization are constructive,” “Our organization has a positive internal climate,” and “Colleagues help each other in problematic work situations.” These statements were rated on a Likert scale ranging from 1 (“completely disagree”) to 5 (“completely agree”).

Third, we created five dummy variables to represent our sample organizations' industries following EU Nomenclature of Economic Activities (NACE) categorization: public administration and defense, construction, financial and insurance activities, manufacturing, and electricity, gas, steam and air conditioning supply. In all the models, the public administration is the reference category as five organizations out of 11 belonged to that group. For data analysis, we used STATA/IC 15.1.

Results

Table 2 presents the means and correlations of the study variables in relevant employee groups. The table provides the Spearman correlation coefficient for dummy variables and the Pearson correlation coefficient for continuous variables.

Table 2 shows that workplace bullying is generally not prevalent in the sample, but slightly more work- than person-related bullying (p = 0.000) exists. Correlations tend to be very small. Neither work-related or person-related negative acts are normally distributed, we therefore log transformed these variables.

To test our hypotheses, we first constructed linear regressions for work-related bullying. In all regressions, we cluster standard errors by organizations to eliminate organization-specific effects (see Table 3).

In Model 1, we test only the direct effects. As can be seen, there is no ageism in hypothesized direction for work-related bullying. Compared to employees between 31–55 years, both younger and older employees perceive slightly less bullying. Thus, H1 is not confirmed. Hypothesis H2a, which suggests that women report more bullying is clearly rejected, because all models produced a negative and significant coefficient. As for double jeopardy of age and gender (H2b) – an interaction in Model 3 is positive and significant for older women, confirming H2b in the work-related domain. Concerning the hypothesis on ethnic minorities (H3a) we can see that direct effects are insignificant and so are the interactions in Models 4 and 5 between age and minorities (H3b), so this hypothesis cannot be confirmed for work-related bullying.

Robustness tests

In Model 3, older age and female categories produce significant negative results, but their interaction is positive and significant at p < 0.05 level, indicating that female employees over 55 perceive more work-related bullying. When we relaxed the older age category to 46 years, direct effects remained basically the same in case of female β = −0.073 (p = 0.000), and age β = −0.062 (p = 0.005). The interaction effect weakened to a negligible extent from 0.08 to 0.076 and remained significant (p = 0.000). Hence, gendered (middle-)ageism is present in bullying female employees with work-related matters.

We also tested whether the unexpected negative coefficient for female may be due to our female-dominated sample. Previous studies inform us that being a gender minority at work significantly predicts workplace bullying (Eriksen and Einarsen, 2004; Gardner et al., 2020). We therefore singled out two organizations (n = 505), where the proportion of women was not more than one-third of the workforce and regressed gender variable (with all control variables) on work-related negative acts. Being female still reduced work-related bullying (β = −0.033; p = 0.03). Similar to the main result, being an older female employee produced positive and significant coefficient (β = 0.082; p = 0.05) for work-related bullying. In sum, gendered ageism is present in both female-dominated and male-dominated organizations.

It is somewhat unexpected that managers report significantly more work-related bullying. Some previous studies have found the opposite effect or reported similar levels of bullying (e.g. Moreno-Jimenéz et al., 2008; Gardner et al., 2020), while others recognize bottom-up bullying (Escartin et al., 2011b). As we had line and middle managers in our sample, these positions are probably more vulnerable due to mistreatment from subordinates (bottom-up bullying), peers (horizontal bullying), and superiors (top-down bullying). Sector controls show that compared to the public sector there is consistently more work-related bullying in the construction sector, financial sector, and to a lesser degree energy sector.

Next, we repeated the same procedure for the person-related bullying. Table 4 presents the results of this study.

Looking at the direct effects in Model 1, it appears again that no hypothesized ageism is present for person-related bullying. The youngest employees experienced even less person-related bullying compared to middle-aged cohort. Gender did not make a difference as well as belonging to ethnic minorities. Gendered ageism and age and minority interactions are insignificant at p < 0.05 level.

To sum up, we refute H1 predicting ageism playing a role in workplace bullying: there is no evidence that the oldest or youngest cohorts of employees experience either work- or person-related bullying more in Estonian organizations. On the contrary, especially for the youngest employees, even less person-related bullying exists compared to middle-aged employees. The gender effect (H2a) hypothesis is not confirmed. The ethnic minority effect (H3a) is also disconfirmed. Gendered ageism was evident for work-related bullying, whereby older women report more negative acts; hence we partly confirm H2b. Ageism for ethnic minorities was not present and H3b was not confirmed.

Additional analysis

Perception of negative acts might not necessarily be bullying in victim's view. We test this with the question “Have you been bullied at work?” where 1 – no, never; 2 – Yes, but rarely; 3 – Yes, sometimes; 4 – Yes, several times a week; 5 – Yes, almost every day. Almost 90% of the sample chose “No, never”. Interestingly, the correlation between reporting of bullying and work-related negative acts is moderate 0.37. After log transforming the bullying variable we run four models with interactions between age/gender and age/minority – none of the models gave significant results. Thus, negative acts concerning work-related matters is therefore not viewed as dramatic, and employees who are subjected to it, i.e. older women, do not believe that it constitutes actual bullying. We summarize the results to our hypotheses in Table 5.

Discussion

Based on our data from 11 large organizations, this study does not find evidence of a positive association between employee age and perceived exposure to negative acts at work. This result is surprising given previous knowledge about older age-related attitudes in Eastern Europe. One explanation for the lack of ageism may be that negative acts are not perceived as particularly “negative” by the older employees, and they, therefore, underestimate it. We suggest three possible reasons for it: lower awareness, stigma, and positivity effect. In Estonia, workplace bullying emerged in public discourse only in the mid-2000s (Tambur and Vadi, 2009). Today, employees who are close to 50 and older were also active in the labor market when workplace bullying was not widely condemnable in society. Furthermore, negative acts are not directly associated with bullying as demonstrated by our additional analysis. The awareness issue was also brought out by Rippon et al. (2015) when they tried to explain why there is seemingly more age discrimination in the UK than in the USA. Some older employees may feel ashamed to admit that they have been a victim of bullying or negative behavior, as they may think of it as a sign of weakness. This could be due to the stigma associated with such incidents and could cause them to feel similar to the way they feel when discussing mental health issues. As a result, older employees may be more likely to underreport incidents of negative behavior.

The third explanation is the more universal positivity effect. The positivity effect (Charles et al., 2003) stipulates that older people tend to recall more positive information compared to younger people. According to Carstensen and DeLiema (2018) many studies have shown by now that older people experience relatively low levels of negative emotions in everyday life. Hence, older employees may be less attentive about negative acts.

We detected gendered ageism as double jeopardy for work-related bullying only, whereas women's age under risk started already from mid-forties. This finding partly supports previous findings (Harnois, 2015; Jyrkinen and McKie, 2012; Pörhöla et al., 2020) and echoes the surprising finding by Escartin et al. (2011a). They assumed that women would more often emphasize person-related bullying, whereas the opposite came to light: women were more concerned about professional discredit instead. Intriguingly, being female was associated with less perceived bullying, although with a negligible difference, resembling some US findings (Rippon et al., 2015; Potter et al., 2019). This result points to an essential difference between wage discrimination and workplace bullying – both are mistreatments, but their root causes differ. Estonia's most enormous gender wage gap in Europe is attributed to individual bargaining (Masso et al., 2022) – lack of collective agreements, trade unions, and treating salaries as confidential information makes women disadvantaged in salary negotiations. Employers use the situation legitimately. Workplace bullying, however, does not occur because of weak formal institutions; it is a much more idiosyncratic phenomenon. In organizations where bullying is a problem, all employees suffer from it, and women no more than men.

Our findings on minorities are both similar to and different from Bergbom et al. (2015) study. In the Finnish case, culturally close immigrants to Finns, that is, Estonians and Swedes, did not report more bullying than ethnic Finns. The same seems to hold in Estonia, meaning that Russian minority groups, as culturally close to Estonians, did not report significantly more bullying. While the younger ethnic minority was more bullied in Finland, this did not occur in Estonia, where age did not make significant difference. This result may be related to our sample organizations. “Young work” is often in the service and hospitality industry, which is susceptible to hosting mistreatment (Ariza-Montes et al., 2017). However, the current study's service organizations belonged to the public and financial sector.

Practical implications

Bullying as a form of workplace deviance refers to deficiencies in the ethical climate. Thus, attaining a bullying-free environment is not only about hindering bullying but also about promoting an ethical climate with respective systems, procedures, and policies (Gardner et al., 2020). Also, a diversity-supportive climate impedes bullying. Supportive practices for older female employees would include celebrating those employees' achievements, providing mentoring opportunities for them, giving them visible and meaningful roles as well as training them. With mindful diversity practices in place, bullying is less likely to occur.

There are direct measures to mitigate bullying, including a code of conduct, regulations, and rules, such as anti-age discrimination policies. While these do not hurt, they are, however, not enough as “… formal written policies and regulations should be accompanied by training and development initiatives that are integral to addressing workplace bullying” (Branch et al., 2013, p. 289). Thus, a supportive environment is equally formal and informal, especially colleague support, in reducing the likelihood of bullying. As bullying takes advantage of power inequality, support from colleagues empowers the victim to confront mistreatment.

Contribution, limitations, and suggestions for future research

This study makes several contributions to the literature. First, we applied an intersectionality lens to study workplace bullying, which has to date been rare, but is much more revealing. Instead of discovering the extent and motivation of bullying, identifying hazardous employee groups may better assist in designing mitigation measures. In line with intersectionality research, which suggests that different experience may occur within the intersections of the distinct social categories (Crenshaw, 1989; McBride et al., 2015; Holman and Walker, 2021), we find that work-related bullying is more reported by older female employees, whereas age and gender variables alone did not suggest it. Our findings highlight the need for simultaneous consideration of certain social characteristics when studying workplace bullying as subgroup inequalities clearly exist. In addition to gender, age, and ethnic minority, future studies may include other social categories, for example religion, single parenting, belonging to LGBT community. In quantitative studies, including interaction terms is the easiest method (Holman and Walker, 2021).

Second, our results indicate that workplace bullying should be approached in a more nuanced manner: for some employee groups (women), victimization occurs only in work-related domains, but not in person-related issues. Studying bullying in general may mask important differences: perceiving the extent of work-related bullying versus person related bullying differently was also revealed by Patel et al. (2022).

Third, although the employee's position in the current study was a control variable, we still note that managers reported more bullying than regular employees. The reason for this is unclear at this stage and deserves further research. We suggest that the distinction between formal and informal power and vertical and horizontal workplace bullying as discussed by Branch et al. (2013) and Escartin et al. (2011b), respectively, are relevant for understanding the complexities of bullying.

Our study has some limitations. The workplace bullying questionnaire was initially developed in a different national setting; the problem of redundant or unrecognizable statements and recall bias may affect the results (Moreno-Jiménez et al., 2008) as we could only rely on self-reporting. Our sample is limited to only 11 organizations, and controlling for various types (public/private, industry/service) was not applicable. Male-dominated fields and female-dominated sectors likely have different predictors of bullying (Moreno-Jiménez et al., 2008), as has been shown for the occurrence of age discrimination (Rippon et al., 2015). Also, supportive work environment items we used do not form a valid scale and given its important role in alleviating bullying, future studies might use established scales, e.g. psychosocial safety climate scale. Finally, the cross-sectional nature of our data does not allow us to draw causal conclusions.

Given the effect of pandemics and the massive spread of remote work, some facets of bullying, like physical intimidation, will likely lose ground, while others, such as cyberbullying, may emerge. The imbalance in technological expertise of perpetrators and victims in the situation of cyberbullying (Vranjes et al., 2018) raises the question of vulnerability of less advanced users of communication technologies and existence of both work-related and person-related bullying in the virtual world. Remote work requires digitally-savvy employees, and older cohorts are disadvantaged in this respect.

In addition, perpetrators' characteristics including age, gender, or ethnicity were not the focus of the present research, but relevant studies can contribute to the understating of hostile attitudes at work. In our view, future research should be also focused on new avenues for age discrimination (Rudolph et al., 2021) and ageist causes of workplace bullying. It is possible that bullying becomes even more topical because of the weakening social support helping to cope with stressful situations and reduce power inequality (Branch et al., 2013).

Hence, future studies in the field of workplace bullying should contribute to the debate in such questions as: (1) how organizations can establish supportive practices for creating ethical climate both in physical and cyber spaces and propose their content; (2) how organizations can raise awareness and educate employees of various positions to recognize and fight workplace bullying towards various subgroups comprising not only gender, age, and ethnic minority, but also religion, marital status, disability and other characteristics. It would be valuable to gain evidence from a larger number of organizations comprising male-dominated, female-dominated sectors and those without specific attachment. In the national context longitudinal qualitative studies with exploratory aims for perceived workplace bullying would be of use. However, for obtaining more comprehensive vision, cross-national quantitative studies would be highly appreciated.

Figures

Figure 1

Study framework

Table 1

Sample characteristics

Category	Sub-category	N	%
Gender	Male	563	34.9
	Female	1,050	65.1
	Not specified	1	0.0
Ethnicity	Majority: native Estonian speaker	1,288	79.8
	Minority: non-native speaker	320	19.8
	Not specified	6	0.4
Age	=<30	276	17.1
	31–45	694	43.1
	46–55	352	21.8
	56=<	290	18.0
	Not specified	2	0.0
Education	Primary or basic education	25	1.5
	High-school or college education	234	14.5
	Vocational education	301	18.7
	Higher education	1,051	65.1
	Not specified	3	0.2
Position	Employee	1,341	83.1
	Manager with subordinates	267	16.5
	Not specified	6	0.4

Source(s): Compiled by the authors

Table 2

Means and pairwise correlations of study variables

Variables	Work-related bullying	Person-related bullying	Work-related bullying	Person-related bullying
	Means		Correlations
Work-related bullying	1.515
Person-oriented bullying		1.236	0.6433**
Up to 30 years	1.454	1.196	−0.069**	−0.051*
Years 31–55	1.554	1.25	0.123**	0.068**
56+ years	1.422	1.213	−0.086**	−0.036
Native Estonians	1.498	1.225	−0.032	−0.046
Minority	1.584	1.28	0.032	0.046
Female	1.492	1.23	−0.043	−0.007
Male	1.559	1.248	0.043	0.007

Note(s): **p < 0.01; *p < 0.05

Source(s): Compiled by the authors

Table 3

Regressions for work-related bullying as the dependent variable (unstandardized coefficients)

Variables	Model 1	Model 2	Model 3	Model 4	Model 5
Reference: 31–55 year old employees
Younger age	−0.067***	−0.068**		−0.055***
Older age	−0.041**		−0.074***		−0.035
Reference: male employees
Female	−0.047***	−0.049**	−0.06***
Reference: native Estonians
Minority	0.031			0.036	0.022
Interactions
Female × younger age		0.016
Female × older age			0.08**
Minority × younger age				−0.01
Minority × older age					0.04
Control variables
Supportive work environment	−0.217***	−0.218***	−0.215***	−0.217***	−0.214***
Reference: employees with no subordinates
Position: manager	0.068***	0.07***	0.073***	0.075***	0.08***
Reference: Public sector
Construction	0.149***	0.174***	0.179***	0.142***	0.15***
Financial	0.125***	0.135***	0.127***	0.132***	0.121***
Manufacturing	0.000	0.018	0.031	0.022	0.031
Energy	0.063*	0.062*	0.069**	0.084**	0.087**
Constant	1.212**	1.207***	1.20***	1.159***	1.149***
R2	0.275	0.212	0.271	0.270	0.267
N	1,608	1,608	1,608	1,608	1,608

Note(s): *p < 0.10, **p < 0.05, ***p < 0.01

Source(s): Compiled by the authors

Table 4

Regressions for person-related bullying as the dependent variable (unstandardized coefficients)

Variables	Model 1	Model 2	Model 3	Model 4	Model 5
Reference: 31–55 year old employees
Younger age	−0.031**	−0.01		−0.033*
Older age	−0.008		−0.015		−0.011
Reference: male employees
Female	−0.016	−0.011	−0.019
Reference: native Estonians
Minority	0.023			0.019	0.014
Interactions
Female × younger age		−0.026*
Female × older age			0.021
Minority × younger age				0.021
Minority × older age					0.046*
Control variables
Supportive work environment	−0.186***	−0.185***	−0.185***	−0.185***	−0.185***
Reference: employees with no subordinates
Position: manager	0.034**	0.034**	0.036**	0.036***	0.039***
Reference: Public sector
Construction	0.055*	0.071**	0.075**	0.056*	0.057*
Financial	0.003	0.006	0.004	0.005	0.001
Manufacturing	−0.024	−0.008	−0.005	−0.013	−0.015
Energy	−0.022	−0.021	−0.02	−0.014	−0.014
Constant	0.945***	0.938***	0.936***	0.926***	0.923***
R2	0.324	0.324	0.322	0.324	0.323
N	1,608	1,608	1,608	1,608	1,608

Note(s): *p < 0.10, **p < 0.05, ***p < 0.01

Source(s): Compiled by the authors

Table 5

Summary of hypotheses

Hypothesis	Result	Comment
H1: Both the youngest and oldest age groups perceive greater exposure to bullying behaviors than the middle-aged group	Not confirmed	Younger and older employees tend to report lower exposure to bullying behaviors compared to middle-aged group
H2a: Women perceive greater exposure to bullying behaviors compared to men	Not confirmed	Female employees reported lower exposure to work-related bullying behaviors compared to men
H2b: Gender moderates the relationship between age and workplace bullying, such that older women perceive more bullying behaviors than older men	Partly confirmed	Female employees older than 46 years reported greater exposure to work-related, but not to person-related bullying behaviors
H3a: Belonging to the ethnic minority group positively relates to perception of exposure to workplace bullying behaviors	Not confirmed	Ethnic minority group did not report greater exposure to workplace bullying behaviors
H3b: Belonging to ethnic minority groups moderates the relationship between age and workplace bullying, such that older minority employees perceive greater exposure to bullying behaviors than older majority employees	Not confirmed	Older minority employees did not report greater exposure to bullying behaviors compared to older majority employees

Source(s): Compiled by the authors

Table A1

Factor loadings of workplace bullying items (promax rotation)

Items	Person related bullying	Work related bullying	Uniqueness
Someone withholding information which affects your performance (dropped)	0.1643	0.3722	0.7575
Being humiliated or ridiculed in connection with your work	0.5615	0.1689	0.5367
Being ordered to do work below your level of competence	−0.0602	0.6438	0.6307
Having key areas of responsibility removed or replaced with more trivial or unpleasant tasks	0.0447	0.6571	0.5292
Spreading gossip and rumors about you	0.7211	0.0400	0.4421
Being ignored, excluded, or “sent to Coventry”	0.4510	0.2546	0.5872
Having insulting or offensive remarks made about your person (i.e. habits and background), your attitudes or private life	0.8057	−0.0887	0.4329
Hints or signals from others that you should quit your job	0.5228	0.1276	0.6264
Repeated reminders of your errors or mistakes	0.5489	0.2091	0.5105
Being ignored or facing a hostile reaction when you approach (dropped)	0.3574	0.4340	0.4887
Having your efforts constantly criticized	0.5308	0.2491	0.4898
Having your opinions ignored	0.2035	0.5807	0.4725
Practical jokes carried out by people you don't get on with	0.6955	−0.0200	0.5335
Being given tasks with unreasonable deadlines	−0.0345	0.7442	0.4773
Having allegations made against you	0.5681	0.1884	0.5071
Excessive monitoring of your work	0.1902	0.4224	0.6842
Pressure not to claim something which by right you are entitled (sick leave, holiday, entitlement, travel expenses) (dropped)	0.3772	0.2427	0.6836
Being the subject of excessive teasing and sarcasm	0.7761	−0.1341	0.5106
Being exposed to unmanageable workload	−0.0184	0.6434	0.6006

Source(s): Compiled by the authors