Abstract
The dynamics of employment relations in small and medium-sized enterprises (SMEs) have attracted academic interest since the 1970s. To date, research debates have converged around two competing perspectives extolling either the opportunities, or the exploitation caused by informal working practices in smaller sized firms. Responding to calls for a more balanced and nuanced view, we analyse n = 1764 responses from a nationally representative study of workplace relations in Ireland, specifically focusing on negative behaviours in SMEs. We contribute to bullying and SME literatures by disaggregating the SME label and showing that certain employee groups in medium-sized firms are likely to report higher incidences of ill-treatment than their counterparts in smaller and larger firms. We conclude by making recommendations on how managers, owners and HRM practitioners can use our study’s findings to improve employee experiences and tackle bullying, harassment and other types of ill-treatment in their respective working environments.
Keywords
Introduction
From unreasonable treatment and incivility/disrespect to threats and physical violence, the detrimental impact of bullying on workplaces has been widely discussed in the literature (Cioni and Savioli, 2016; Einarsen et al., 2011; Fevre et al., 2011). Furthermore, it is well-established that organisational context, including the work environment, culture, reward systems and staff relations, is a critical factor influencing bullying behaviours (Einarsen et al., 2011; Nielsen et al., 2011; Salin, 2003). However, the role of workplace size and associated internal dynamics is less clear. This is especially so in the case of small and medium-sized enterprises (SMEs), where the correlation between size and workplace experiences has been the subject of long-standing critique (Henderson and Johnson, 1974; Marlow, 1992; Wilkinson, 1999), yet, the evidence is inconclusive. More recent studies (Kitching and Smallbone, 2010; Mallett et al., 2019) have also questioned the tendency to view small and medium-sized firms as an SME ‘aggregate’, issuing a call for further research into their internal dynamics (Kitching et al., 2015a, 2015b). However, a ‘small is beautiful’ perspective can be identified in the literature suggesting that workplace bullying is more common in larger organisations (Dignity at Work, 2001; O’Connell et al., 2007; Parent-Thirion et al., 2007). Less encumbered by the bureaucratic apparatus of larger organisations, SMEs seem to enjoy reduced management-worker distance and seem better equipped to resolve worker problems locally (Edwards et al., 2003; Grimshaw and Carroll, 2006). However, a contrasting, ‘bleak house’ (Rainnie, 1991) view suggests that SME managers and owner-managers have significant scope to interpret and adapt regulations (Gilman and Edwards, 2008; Mallett et al., 2019). This has historically limited employee abilities to raise issues regarding working conditions (Wilkinson, 1999) whereby size alone may not afford protection (cf. Mallett and Wapshott, 2017 for an extended discussion).
This article, therefore, responds to calls to disaggregate the SME ‘black-box’ (Baldacchino et al., 2015; Kitching and Smallbone, 2010) and consider the complex dynamics of medium-sized firms. We analyse data from n = 1764 face-to-face interviews conducted as part of the Irish Workplace Behaviour Study (IWBS – Hodgins et al., 2018) over the 2015–2016 period in Ireland. The study uses questions drawn from the IWBS but based on the British equivalent, British Workplace Behaviour Study (BWBS – Fevre et al., 2011) and its modified version of the Negative Acts Questionnaire-Revised (NAQ-R; Einarsen et al., 2009). This not only permits the disaggregation of data across different-sized organisations but enables us to extend understanding of SME dynamics in an Irish context through both exploring the possible impact of variation within SMEs and exploring their particular organisational context. Given the central role that SMEs play to the Irish economy (Harney and Nolan, 2014), understanding how adverse social behaviours in the general work environment are enacted in Irish SMEs and how organisational size mediates or mitigates work environment factors and negative behaviours, is critical. The article is structured as follows. We provide an overview of the bullying and ill-treatment literature, which we situate in the specific context of SME organisations. We then report our findings into the experiences of workers in medium-sized firms and conclude by offering recommendations for managers and human resource practitioners.
Bullying and ill-treatment in context
An extensive literature exists regarding the antecedent causes and resultant outcomes of negative behaviours in the workplace (Fevre et al., 2011; Salin, 2018). Bullying is a special case of such negative behaviours that can be considered an extreme social stressor that is bounded by persistency over a long-time period and where the target finds that they are, in some way, unable to stop or escape from the situation (Ägotnes et al., 2018; Einarsen et al., 2011).). Workplace bullying is frequently based on power inequality, for instance managers withholding information, undermining, or setting impossible deadlines to subordinate workers (Einarsen et al., 2011; Houshmand et al., 2012), or co-workers who control group dynamics by means of gossip, inappropriate humour or through in-group/out-group controls (Lewis et al., 2020). The overarching causes of bullying are usually dichotomised in the literature as either individual or organisational (Ågotnes et al., 2018; Einarsen et al., 2011). Individual causes include personality traits in the target, perpetrator or both, while organisational causes include job design, employee relations, organisational culture, leadership and reward systems (Salin and Hoel, 2011). While there is evidence that individual traits have a contributing role, the balance of evidence is substantially in favour of factors that reside within the work environment (Feijo et al., 2019; O’Connell et al., 2007; Task Force on the Prevention of Workplace Bullying, 2001; Trepanier et al., 2016).
Studies on the prevalence of bullying behaviours typically measure both personally directed aggression, as well as a wider range of workplace ‘ill-treatments’, an umbrella term for unreasonable work treatment and work-related negative acts. Examples of such negative acts may be incivility, that is, verbally abusive behaviour of lower intensity that may or may not intend to cause harm to the target (Anderssen and Pearson, 1999; Hodgins et al., 2018), or disrespectful behaviours such as social isolation or being ignored (Fevre et al., 2011; Hodgins et al., 2018). Thus, BWBS (Fevre et al., 2011) found that just over half of respondents (54%) experienced ill-treatment (at least one item) in one of these forms. In this instance, unreasonable treatment, mostly meted out by managers, was the most common item (47%), followed by incivility/disrespect (40%; Fevre et al., 2011). IWBS (Hodgins et al., 2018), a replication of the BWBS, found just under half (43%) experienced some form of ill-treatment in their workplace. Like the British study, unreasonable treatment by managers/supervisors was the most commonly reported item (37%), followed by incivility/disrespect items (31%). Regardless of what specific terms are employed, it is unequivocal that these experiences have a negative impact on health and well-being for individuals who witness and live through them, which negatively affects absenteeism, employee turnover and lost productivity (Hodgins et al., 2018; Parent-Thirion et al., 2007; Schilpzand et al., 2016).
The case of SMEs – ‘small is beautiful’ vs ‘bleak house’
Debates on how working environments in large and small firms differ in respect of employee relations (Barrett and Rainnie, 2002; Ram et al., 2007; Yaw and Mmieh, 2009) can be extended to include the respective environmental impact on workplace ill-treatment. The ‘small is beautiful’ (Schumacher, 1973) view of small businesses, has historically presented them as free of cumbersome bureaucratic structures, reduced management-worker distance, flexibility and ease-of-communication (Mallett and Wapshott, 2017; Marlow et al., 2010). Small firms are thought to be more transparent, and responsive, better equipped to resolve worker problems without recourse to formal and disruptive human resource interventions or trade union (TU) disputes (Atkinson et al., 2016; Bischoff and Wood, 2013). There is some evidence to support this position. Inferences from large datasets such as Eurostat or the European Working Conditions Surveys (Eurofound, 2017) for Europe, and the Workplace Employment Relations Survey (WERS) in the United Kingdom (Van Wanrooy et al., 2013), are that employees in SMEs appear more likely than employees in large firms to indicate that their views are heard and acted on by managers. Two large-scale surveys in Ireland identified a positive correlation between increasing exposure to bullying and organisational size. Specifically, O’Connell et al. (2007) reported bullying levels of 5% for organisations with 1–4 workers, which doubled to 10.9% in organisations with over 100 staff (O’Connell et al., 2007; Task Force on the Prevention of Workplace Bullying, 2001). A similar pattern was observed in the Fourth European Working Conditions Survey, which, employing standard organisational size measures, found bullying and harassment were experienced by just over 4% in micro companies and 7.5% in large enterprises (Parent-Thirion et al., 2007). O’Connell et al. (2007) also found that the strength of consultation and perceived quality of working relationships both decreased as organisation size increased.
It should be noted that the study of bullying in larger organisations usually relies on responses to a direct question about whether the participant experienced bullying or not. Referred to as ‘the self-labelling method’, a participant is offered a definition of bullying and is directly asked whether or not she or he has experienced this (Nielsen et al., 2011). However, this measure is confounded by the possible impact of defence mechanisms (where the participant is unwilling to be seen as a victim; Notelaers et al., 2006) or because of subjective interpretations in reading the definition provided (Illing et al., 2016). Our study followed BWBS (Fevre et al., 2011) and thus, focused upon the behaviours that underpin constructs such as bullying and ill-treatment. Furthermore, large organisations are more likely to recognise TUs (Eurofound, 2006) and anti-bullying policies (O’Connell et al., 2007), suggesting a potential danger to employee experience in SMEs where safeguards may be reduced or absent, although a dearth of current and up-to-date studies in the literature make conclusions problematic (Mallett and Wapshott, 2017; Monat, 2018). This alternative, ‘bleak house’ (Rainnie, 1991; Wilkinson, 1999) perspective of SMEs suggests that the selfsame flexibility and scope for faster, localised decision-making as well as proximity between employees and owner-managers (Nadin and Cassell, 2007) could also present a new raft of challenges and opportunities for employee exploitation. There are ongoing problems with the fair treatment of employees and the absence of voice channels in SMEs across the Agriculture, Construction, Hospitality and Commerce industries (Eurofound, 2017), which is exacerbated in family businesses although, again, studies are scarce (Botero and Litchfield, 2013; Cruz et al., 2014). In turn, the majority of such firms are micro or small with 98% having less than 50 workers and unlikely to have TU presence (Central Statistics Office, 2005). In those instances, the informality of working contexts makes the experience of employees in SMEs individualised and contingent upon personal relationships (for instance, with managers) rather than framed by standardised employment conditions (Atkinson et al., 2016; Mallett and Wapshott, 2017; Ram and Edwards, 2003). This further endorses concerns that a lack of formal structures such as human resource (HR) functions and TU representation will have a negative impact on the employment relationship in SMEs (Lewis et al., 2017; Matsuura and Noda, 2017). This is, however, a largely unexplored area (Elsetouhi et al., 2018) and especially so, given the shortage of employment relations and human resource management (HRM) data for Ireland (Harney and Dundon, 2006; Wilkinson, 1999). Thus, in order to test the impact of TUs, and we pose that, in the context of Irish SMEs:
Hypothesis 1: Trade Union participation will limit SME employees’ exposure to ill-treatment.
The paradox of SMEs
This article addresses calls to better understand workplace dynamics in SMEs (Lai et al., 2017; Paauwe, 2009; Ram and Edwards, 2003) and thus, unpack the SME paradox identified earlier. The literature on SME dynamics increasingly adopts a nuanced view (Atkinson et al., 2016) and seeks to go beyond assumptions that the complexity of interactions in SMEs are polarised into large or small, are solely and deterministically driven by size (Curran, 2006; Rainnie, 1991) or formality or informality of HRM processes (Kitching et al., 2015a; Nadin and Cassell, 2007). So, although a range of situational factors such as the specific industry segment, available resources, labour supply and so on are considered (Mallett and Wapshott, 2017), it is also the case that size itself is a complex variable. This leads us to propose that it is not only variation between SMEs and larger firms that need to be considered when studying bullying, but differences within the SME categories themselves (Nadin and Cassell, 2007; Wilkinson, 1999). Specifically, we argue that the SME-employee relations paradox can be addressed with specific reference to medium-sized firms. We acknowledge evidence that SME employees are at times better protected from exposure to bullying, harassment and work ill-treatment (Lai et al., 2017). Yet, we hypothesise that work environment factors, as the primary drivers of workplace bullying (Einarsen et al., 2011; Lewis et al., 2017), play a particular role in medium-sized firms. We propose that the ‘protective’ factors in SMEs viz. transparency, flexibility and informality may not work in the same way in such firms.
This is consistent with the literature regarding possible causes or drivers of workplace bullying and ill-treatment. Specific aspects of the work environment such as high levels of conformity and rigid hierarchies (Ashforth, 1994; Salin, 2003), workload and job autonomy (Baillien et al., 2011a), role conflict and role ambiguity (Salin and Hoel, 2011), poorly communicated or managed organisational change, social climate and communications climate (Baillien et al., 2011b) reward systems (Salin and Hoel, 2011) and organisational culture (Salin, 2003) are all associated with, and act as antecedents for bullying in workplaces. Given that these aspects of the working environment could be mediated by firm size and allowing for medium-sized firms potentially facing a ‘double-whammy’ of reduced personalisation in staff-manager relationships as well as less fulsome policy coverage afforded by large organisations (Lai et al., 2017; Nadin and Cassell, 2007), we offer the second hypothesis:
Hypothesis 2: Workers in medium-sized organisations are particularly likely to report higher levels of ill-treatment.
Central to countering the impacts of poor working environments are the support of managers and colleagues through social interactions in the workplace and maintenance of appropriate behaviours and challenging of inappropriate ones (Baillien et al., 2011b; Lewis et al., 2017). Irish studies on bullying such as Cullinan et al. (2019) demonstrate that employees who reported poor relationships between management and employees and who further perceived (a) a lack of social support at work and (b) the existence of negative workplace relationships, creates significant stressors at work. Contradictory expectations and demands or breaches of the psychological contract are often a feature of family-run firms which are almost exclusively micro or small businesses (Nadin and Cassell, 2007). It is an assumption that family businesses operate like one big happy family, and it is very possible that family politics and workplace politics become enmeshed, to the detriment of some employees. Indeed, SMEs that are family businesses are more likely to experience bullying (Baillien et al., 2011b), while Dundon et al.’s study of a medium-sized family-run business found a complex picture in relation to employee satisfaction, and high levels of tension between employees and managers (Dundon et al., 1999).
Leadership has also been prominent in research into bullying with researchers noting the importance of leadership style in countering negative employee outcomes, such as abusive leadership styles in tyrannical (Einarsen et al., 2007), or laissez-faire ways (Ägotnes et al., 2018). Similarly, levels of manager support in failing leadership organisations are likely to be low/absent and thus, control of work situations becomes challenging for employees (Cullinan et al., 2019) leading to negative employee outcomes. In the face of poor levels of leadership/management support, it is not unreasonable for employees in SMEs to seek co-worker support, particularly in the smallest of firms where employees and owner-managers work in close proximity. While Cullinan et al. (2019) found no evidence in Ireland for the effects of co-worker support on stress levels, others such as Lewis et al. (2017) reported their importance in buffering the effects of bullying and harassment in SMEs. Fevre et al. (2012) capture these concepts using the FARE acronym representing fairness and respect (see also Walker and Fincham, 2011). Fevre et al. (2012) argued that the FARE items and particularly, those related to not being treated as an individual and having to compromise one’s principles, were greater predictors of workplaces that might be troubled by the sorts of ill-treatment under which bullying falls, than more traditional measures associated with stress or job satisfaction. Their findings of strong correlations between FARE items and measures of incivility and disrespect and with problems associated with employment rights, make this prime territory for further exploration. In turn, the FARE instrument (see Table 2 below) also reflects tensions around job demands and access to resources through, respectively, low and high questionnaire scores, and we formulate our third hypothesis accordingly:
Hypothesis 3: FARE scores in medium-sized organisations are positively correlated with experiences of ill-treatment.
Methodology
Data and sample
Data used in this study were derived from IWBS (Hodgins et al., 2018). The probability sample included individuals who had been in employment for at least two years prior to the study. Face-to-face interviews were completed at n = 1764 addresses across Ireland with one participant selected for interview per household, based on a random sample of eligible individuals within visited houses. The obtained sample of n = 1500 was supplemented by 200 non-Irish nationals and 64 people with a disability to ensure sufficient numbers for sub-group analyses, providing the overall sample of 1764. Data were weighted by age, gender, ethnicity, disability and social class using the National Household Survey Q2 2015 (Central Statistics Office Ireland – CSO.Ie, 2015).
Measures
The main response measure is a 21-item, workplace behaviour scale (WBS), derived from BWBS (Fevre et al., 2011) as per Table 1 below. A higher count of WBS items represents a wider range of the different types ill-treatment were experienced.
WBS 21 item ill-treatment scale.
With regard to the main independent measures, respondents were asked how many employees were there in total (excluding owners and directors) at their place of work. In order to test Hypotheses 1 and 2, response options were grouped according to standard EC categories (EC, 2003) which are as follows: micro (<10), small (10–49), medium (50–249) and large (250+ employees) For testing Hypothesis 1, respondents were asked if they were members of a trades union or staff association with yes/no response options and grouped as trades union/staff association member (reference: not a member of a TU/staff association). In order to test Hypothesis 3, workplace conditions and culture were measured using 10 statements, termed FARE items (Fevre et al., 2011), asking participants to indicate with a yes/no response, the applicability of statements regarding their workplaces (see Table 2 below). In this study, FARE items that had been positively oriented (‘I decide how much work I do or how fast I work’) were subsequently reverse coded in order to indicate the negative condition across all items. In order to test Hypothesis 3, the 10 items were then summed to create a measure where higher numbers of reported items represent more negative workplace conditions and culture.
FARE items.
Items that had been positively oriented when presented to participants and subsequently reverse coded.
Respondent and workplace characteristics
In line with extant literature, a number of individual employee and workplace characteristics identified in previous studies (Baillien et al., 2011b, 2011a; De Cuyper et al., 2009; Lewis et al., 2017; Zapf et al., 2011) were assessed as control variables in our analyses. We took those steps to ensure their effect was accounted for in determining variances across workplace sizes on the number of WBS items experienced. Control variables were gender female (reference: male); age as four categories: 35–44, 45–54 and 55+ (reference: 18–34), ethnicity BAME (reference: White); supervisory role (reference: non-supervisory role); contract permanent (reference: temporary); length of current/last employment less than two years (reference: more than two years); working hours part time (reference: full time); sector public (reference: private, voluntary or other); and workplace is part of a larger organisation as five categories: 250–10,000, 5–249, 10–49, < 10 (reference: Workplace not part of a larger organisation, that is, respondents working in a smaller unit that was part of large statutory organisation such as a small health clinic with service agreement with Health Services Executive Ireland).
Analysis
Unadjusted, population weighted mean counts of the WBS scale, with a potential range of 0–21 items, and Spearman correlations between item counts and the main independent and control variables were determined. To test hypotheses, these were then adjusted for individual and workplace characteristics by modelling counts of WBS items using Poisson and negative binomial distributions. These count-based models provide exponentiated effect estimates as incidence rate ratios. Increases in counts in this study, however, do not indicate increased incidence of ill-treatment but represent a greater number of types of ill-treatment experienced from the full 21 item scale within the previous two years. For Hypotheses 1 and 2, individual and workplace characteristics were entered into the model simultaneously along with interaction terms for each with workplace size. Changing the workplace size reference group for each model enabled estimation of main effects (ratio of estimated means relative to independent variable reference group means); interaction effects (as ratios of main effects relative to main effects in large workplaces); as well as estimated marginal means for each characteristic and ratios of estimated means for micro, small and medium work places relative to estimated means in large organisations. Among the high percentage of response cases that had a zero value none were structural, all have the meaning of ‘no items reported’, and therefore, a zero-inflated model was not theoretically appropriate.
Poisson and negative binomial models had deviance and Pearson values above 1 and this over dispersion was adjusted using a negative binomial distribution with log link, maximum likelihood scaling and robust covariance. Model comparisons between Poisson and negative binomial distributions were tested using the log likelihood ratio test and the Lagrange test was used to assess the fit of scale parameters in the negative binomial model. The significance of exponentiated estimates was determined using log likelihood profile 95% confidence intervals. Probabilities for differences and ratios between estimated means in micro, small and medium, relative to large organisations for individual and workplace characteristics were Bonferroni adjusted to determine significance at p < 0.05. All modelling analyses were conducted using SPSS version 25 using generalised linear estimation.
Results
We provide descriptive statistics in Tables 3 and 4 below. Table 3 shows the unadjusted weighted means and standard deviations for the reported number of WBS items by respondent and workplace characteristics. An overall mean number of items of 2.0 with standard deviation of 3.5 show data were highly left skewed with variances higher than means with a high percentage of zeros (56.8%).
Weighted mean (SD) number of WBS items experienced for demographic and workplace characteristics by workplace size.
BAME: black and minority ethnic; TU: trade union.
Percentages (n) within each workplace size by individual and workplace characteristics and bivariate correlations with number of items experienced.
BAME: Black and Minority Ethnic; SD: standard deviation; TU: trade union.
Spearman’s rho p < 0.05; **p = 0.01.
Table 4 shows percentages for each independent variable by workplace size and correlations between these variables and number of ill-treatment items experienced. Weak significant bivariate correlations were found, with number of items of ill-treatment increasing as workplace size increased and as age decreased. More items were reported by BAME (Black and Minority Ethnic) individuals, those on permanent contracts and those in the public sector.
Given the highly skewed nature of the number of ill-treatment items experienced (Table 3), Poisson followed by negative binomial distributions were used to adjust outcomes for the WBS scale. Compared to the Poisson distribution, the negative binomial was a better fit (log likelihood chi test: p ⩽ 0.001). The negative binomial model scale parameter in the model showed a good fit using the Lagrange test (p > 0.9). From the total sample of n = 1764 (weighted: 1754), 733 (weighted: 385) were excluded (21%) due to missing values providing a model sample of n = 1035 (weighted: 1396).
Table 5 below shows the main and interaction effects of the negative binomial model when the reference group in the interaction terms with workplace size is successively changed in the model. Table 6 shows the estimated means derived from the model and also ratios of the means for medium, small and micro relative to large firms.
Main and interaction effects.
BAME: Black and Minority Ethnic; TU: trade union.
Negative binomial distribution parameter = 2.486 (SE 0.16, 95% CI 2.20–2.81), Lagrange test, p = 0.998; Log Likelihood = −2237.864 (Poisson model −3271.7, lr test p < 0.001); AIC = 4607.7; BIC = 4952.4; Deviance = 1184.9 and Pearson = 1224.7 values by df(1304) = 0.909 and 0.940, respectively.
Significant reduction in items relative to reference category *p < 0.05, **p < 0.01.
Significant increase in items relative to reference category *p < 0.05, **p < 0.01.
Note. Please refer to the online version of the article to view the table in colour.
Estimated mean (SE) number of WBS items experienced and ratios relative to large workplaces.
BAME: Black and Minority Ethnic; SE: standard error; TU: trade union.
Bonferroni adjusted significant differences in estimated means relative to large workplaces p < 0.05.
Note. Please refer to the online version of the article to view the table in colour.
Hypothesis 1 proposed that TU participation limits exposure to negative behaviours. Overall, one-third of respondents were members of trades unions (32%, see Table 4) with SMEs having percentages near to this level (37% and 35%, respectively) but being lower in micro firms (10%) and higher (57%) in larger organisations. Table 5 shows that Hypothesis 1 holds for large organisations with a 63% reduction in number of ill-treatment items experienced by TU members relative to non-members. However, these data do not support Hypothesis 1 in medium, small and micro firms as Table 6 shows; there is a fourfold increase in items experienced in medium compared to large organisations among trades union members. Significant interaction effects emerge where the difference in the mean response for a measure, relative to its reference group, is in the opposite direction for medium, small or micro firms compared to the equivalent difference for large firms, examples of this effect are shown in Figure 1a and b. The significant (p < 0.05) interaction effect shown in Table 5 and depicted in Figure 1b reflects this reversal of the situation for trades union members in medium-sized firms.
Estimated marginal mean number of items experienced (a) by workplace size and ethnicity and (b) by workplace size and trades union membership.
Hypothesis 2 proposed that employees in medium-sized firms are likely to report higher levels of ill-treatment. Adjusted means (Table 6) show that overall, those in medium organisations experienced a significant (p < 0.05) fourfold higher number of items compared to large organisations. Increases relative to large organisations were also found in small (2.3×) and micro (3.2×) organisations but were not significant. Furthermore, as shown in Table 6, those significantly (p < 0.05) more at risk in medium relative to large firms include younger employees (aged 18–43; 4.2×), males (4.7×), those in non-supervisory roles (4.0×), of more than two years in their post, trades union members (6.0×), in the public sector (4.4×) and those whose workplace is part of a larger organisation but were of medium size (7.2×).
As illustrated in Figure 1a, a significant interaction occurs for ethnic group by workplace size. In large organisations there is a lower number of ill-treatment items reported by BAME individuals compared with White, but in medium and micro firms there is a reverse of this situation, with higher numbers of items reported by BAME employees compared with White employees. Other significant interaction effects (Table 5) also occur in micro firms for women (with more favourable experience in micro firms for women), those in micro workplace settings that form part of a medium-sized firms (with relative increase in experience of ill-treatment in this setting) and those in who work alone or in low numbers that are part of a micro firm (with relative reduced experience of ill-treatment in this setting).
Hypothesis 3 proposed that FARE scores in SMEs are positively correlated with experiences of ill-treatment. Table 5 shows that the number of ill-treatment behaviours experienced increased across all workplace sizes by approximately 55%–70% for each increase in the number of FARE items reported thus, supporting this hypothesis.
Discussion
The ongoing incidence of bullying, harassment and ill-treatment in organisations and the negative impact of such behaviours on employee welfare has been noted in the literature (Cioni and Savioli, 2016; Einarsen et al., 2011). However, there are few studies which study the culture and organisational environment in SMEs (Gray et al., 2012) where the correlation between workplace size and internal dynamics (as factors buttressing negative behaviours) is less clear. The literature has tended to coalesce around two key perspectives on SMEs, namely, ‘small is beautiful’ and ‘bleak house’ (Rainnie, 1991; Schumacher, 1973; Wilkinson, 1999). Assuming a much closer relationship between manager-owners and workers, the ‘small is beautiful’ perspective suggests that bullying and harassment are likely to be a feature of larger firms and their bureaucratic and formalised internal HR and policy mechanisms (O’Connell et al., 2007). However, formal conflict-management structures, sanctions and warnings (Einarsen et al., 2017) can be a source of protection from arbitrary and favouritist treatment of staff, which suggests that there are more safeguards for employees in larger organisations (Lewis et al., 2017; Yaw and Mmieh, 2009) than in SME ‘bleak houses’. In line with this apparent paradox, we used data from n = 1764 face-to-face interviews collected between 2015 and 2016 to disaggregate the composite SME label and study bullying, harassment and other ill-treatment behaviours of workers across micro, small, medium and large firms in Ireland (IWBS – Hodgins et al., 2018).
This yielded a number of surprising insights. Our descriptive statistics showed that the nuanced everyday reality of SMEs cannot be reduced to an either/or view of ‘small is beautiful’ or ‘bleak house’ on account of within SME differences. Thus, we found that although, in aggregate, SME employees reported experiencing more ill-treatment behaviours than larger organisations, this was significantly more in medium firms (4.0 times more), than micro (3.2 times more) or small (2.3 times more). The particularly negative environment in medium-sized firms was observed when testing our hypotheses. We found that despite evidence of a positive impact of TU membership on employees in larger organisations (Lewis et al., 2017), this effect was reversed in medium-sized firms where TU members were of greater risk of ill-treatment; Hypothesis 1 was thus rejected. There was partial support for Hypothesis 2 as the working environment in medium-sized firms placed at risk of ill-treatment certain groups of workers, namely, older (55+) and younger (18–43) employees. Female employees were also at greater risk of ill-treatment in medium-sized, as compared to small and micro firms. Finally, we found support for Hypothesis 3 since there was an increase (of approximately 55%–70%) of ill-treatment behaviours (such as bullying) for each increase in the number of reported FARE items.
It has been argued that larger organisations are more likely to have systems and processes in place to tackle bullying and associated ill-treatment (O’Connell et al., 2007). Similarly, while the existence of anti-bullying or dignity at work policies is associated with lower stress levels for employees (Baillien et al., 2011a; Cullinan et al., 2019) questions about the preparedness of SMEs to tackle adverse social behaviours at work remain. Thus, our study is able to contribute to the debate but focusses on medium-sized firms. In seeking to explain our results, it may be useful to recollect that Atkinson et al.’s (2016) study of medium-sized firms shows that external consultants were used to oversee the employment relationship and protect the firm from the complex regulatory environment (see also Saridakis et al., 2013). While relying on external advice is understandable as a firm begins to grow but, perhaps, as yet lacks an in-house HR function, there are inherent risks of failing to take ownership of problems that occur in the employment relationship. As an example, organisational politics and negative behaviours born of external pressures may be visible and easier to address in smaller firms, but less so as the numbers increase. As recognised by Mallett and Wapshott (2017), resource availability is a key factor impacting the internal environment of SMEs and it is possible that resource competition is particularly intense in medium-sized firms. Instruments such as the Management Standards for Work-related Stress in the United Kingdom and Work Positive in Ireland (Mackay et al., 2004) and models proposed by Karasek (1979) based on Job-Demands-Control, have been successfully used by bullying researchers to demonstrate how an absence of support and resources versus high job demands and low control (Baillien et al., 2011b; Lewis et al., 2017) strongly correlate with bullying experiences. This relationship may be advantageous to managers in medium-sized firms who could strategically target workers framed as unable to take the pressure or because they are deemed inefficient and bully them until they leave.
Against such a backdrop it is perhaps unsurprising that members of a TU reported higher levels of ill-treatment (Hypothesis 1), possibly on account of being better informed of their employment rights and through adequate representation (Lewis et al., 2017). Furthermore, medium-sized firms face an additional paradox of having the workforce size to necessitate in principle but in practice lack the policies and process such as alternative dispute resolution (ADR) mechanisms, that Roche and Teague (2012) identified as necessary to address workplace conflicts in Ireland. This supports research showing that when TUs are viewed sceptically (Dundon et al., 1999; Forth et al., 2006), or are bypassed by adoption of HR practices (Harney and Dundon, 2006), or negated by informally placing faith directly in owner-managers (Atkinson et al., 2016), there are inherent risks to not only experiencing but recognising workplace ill-treatment.
The negative experiences of older (55+) and younger (18–43) employees in medium-sized firms (Hypothesis 2) may be attributed to the fluidity of working environments, and particularly, the incidence of role conflict and role ambiguity, likely to be seedbeds of negative behaviours (Reknes et al., 2014). Interestingly, in the United Kingdom, 79% of SME managers have responsibility for employee relations (Van Wanrooy et al., 2013) yet, when managers themselves are bullied, there is a potential for both role conflict and ambiguity to increase (Lewis et al., 2017) so, perpetuating the process. Existing research demonstrates mixed findings relating to age and the types of mistreatment that leads to bullying (Hauge et al., 2009). Some argue that older workers are more experienced and able to control personal emotions and stand up for themselves (Barling et al., 2009) while others report heightened levels of bullying for older workers (Einarsen and Skogstad, 1996). As such, research suggests age is situational and in the context of this study it is possible that both younger and older workers were vulnerable to redundancies, such as those imposed in Ireland as a direct consequence of austerity, or through the types of employment contracts associated with precarious work such as zero-hours contracts (Manolchev et al., 2018).
In turn, the negative experiences of BAME workers in medium-sized firms may be linked to requirements for SMEs to abide by anti-discrimination legislation covered by the creation of the Irish Human Rights and Equality Commission Act 2014. Ireland has experienced significant growth in immigration over the last two decades (Central Statistics Office Population and Migration Estimates April, 2018), with BAME migration showing the highest growth in Ireland’s 2016 Census of Population data. With Pan-EU data showing 80% of respondents stating harassment was related to ethnic origin or immigration status (Second European Union Minorities and Discrimination Survey, 2017), our data supports the wider picture recognised by McGinnity et al. (2017) that being an ethnic minority, or person of colour, places individuals at the forefront of being detrimentally and perhaps even strategically targeted at work.
Aside from age and ethnicity, our results show that male workers are more at risk of encountering ill-treatment in micro and medium-sized firms compared with counterparts working in large firms. Evidence has also previously indicated concerns about limited opportunities for employee voice in SMEs (Lewis et al., 2017), particularly against risks of being classified as undermining social cohesion (Marlow and Patton, 2002). On a wider scale, the precarity of neoliberal working contexts also is likely to lead to less representation and diminished collectivisation (Manolchev et al., 2018) with speaking up against ill-treatment being inherently risky, particularly in an economy faced with continued austerity (SME Market Report, 2017).
Finally, our data support Fevre et al.’s (2012) findings on the importance of the FARE items in understanding troubled workplaces (Hypothesis 3). Our data show that the number of ill-treatment behaviours experienced increases across all workplace sizes by approximately 55%–70% for each increase in the number of FARE items reported. Fevre et al. (2012) stated the FARE items related to a fundamental employee belief that their individual contributions were somehow secondary to organisational ones, where individuality is lost, and principles were compromised. Our finding that these items are so prevalent among employees in medium-sized firms surprised us. With notions of close proximity between employee and owner-manager/supervisor and in the absence of faceless bureaucratic structures, we might have expected employees in medium-sized firms to reject the FARE measures. This is related to the benevolence of their employer in being willing to resolve concerns and listening in a responsive manner (Van Wanrooy et al., 2013). It appears that this is not the case and the direct connections between the full range of ill-treatment behaviours and FARE items suggest this could be fertile ground for predicting workplaces troubled by ill-treatment, bullying and discrimination.
Implications and limitations
Lai et al. (2017) reported positive financial performance in UK SMEs that had adopted high-performance HRM metrics. They also argued that the greater the formality of the human resource mechanism, the weaker this relationship. However, they also reported that this was only effective when SMEs already had a highly satisfied workforce. In our article, employees in medium-sized firms are at considerable risk of exposure to various types of ill-treatment, compared with those in large firms. These employees require greater formality, not an avoidance of it, and under-performing medium-sized firms would benefit from more formalised HRM practices to work towards the attainment of fairness and justice described by Saridakis et al. (2013). The direct connections between the study’s ill-treatment measures and the FARE items present a compelling case for recognising the critical connections between perceptions of fairness and respect on the one hand, and workplace mistreatment on the other. The position of BAME employees in medium-sized firms is particularly troubling, on account of the significantly elevated experiences of ill-treatment encounters. If mainstream medium-sized firm employees are operating with limited voice mechanisms, these are likely to be significantly exacerbated for BAME employees, many of whom will be immigrants to Ireland and who may be attracted to work alongside others from similar ethnic groups.
There are numerous implications arising from this study not least of which is the need for medium-sized firms to recognise the importance of dealing with workplace ill-treatment in-house, and to provide organisational owner-managers and leaders training opportunities in the sensitive management of ill-treatment. As indicated by Lai et al. (2017) and Marlow et al. (2010), the solution for SMEs is not to deploy a one-size-fits-all solution but instead, adopt a nuanced approach that understands the concerns expressed by TU members, younger, older and BAME employees. Voice channels for each group will differ as will their engagement in ADR channels. Expecting mediation to be deployed effectively for all groups is naïve as indeed is perhaps the expectation for mediation to work for severe incidences of ill-treatment that might best be characterised as bullying. Full-blown conflicts that have gone on for considerable time, such as bullying, are unlikely to be remedied by mediation (Deakin, 2014). Nevertheless, when conflicts are minor in nature, or are encountered early in the conflict cycle, mediation can be effective. It must not be seen as a tool of last resort for SMEs (Latreille et al., 2012) but rather a measure of first response.
While existing evidence points to concerns about excessive HR practices stifling SME effectiveness (Saridakis et al., 2013), striking the right balance between performance and benevolence is key. Investing in equality and diversity training is also eminently sensible. Not only is it morally sound, but it makes business sense for an economy with growing migrant populations. It is also likely to reap dividends in retaining quality staff, attracting new employees and keeping the organisation away from damaging legal action for discrimination and mistreatment. Respectful treatment starts and ends with behaviour and this must be role modelled by those who own, manage and lead businesses. If medium-sized firms are to overcome these problems, they must address the resource poverty deficit in leadership development described by Garavan et al. (2016). Investing in good leadership and management practices could be beneficial for productivity, labour turnover and employee retention, all key metrics for workplaces with high incidence of ill-treatment.
The study has, of course, limitations. Although we deployed a national probability sample of n = 1764 persons, the numbers of respondents across the spectrum of ethnicity categories meant it was necessary to combine these into a single BAME category. There are significant costs of purposively sampling hard-to-reach populations (Lewis et al., 2013) and researchers would be wise to fully investigate the costs of accessing such populations in order to capture sufficient numbers of responses. Surveys of the general population in Ireland rely on samples drawn from the GeoDirectory of addresses; with 15% of addresses (at the time of the study) being unoccupied, this places considerable financial pressures on researchers when conducting fieldwork as it is unknown which properties are occupied or not. Finally, face-to-face in-home researcher-participant interactions of the kind conducted for this study can encumber some participants who may not wish to disclose intimate details of their problematic working lives in front of family members, particularly if those family members are working in the same firm which is distinctly possible in rural Ireland. Good practice would suggest researchers have back-up methods such as telephone and postal/on-line survey tools available (Lewis et al., 2013). This could be particularly appropriate for minority groups who may be the sole wage-earner in that household or, are working in family businesses.
In conclusion, our article illustrates the considerable risks of ill-treatment in medium-sized firms across Ireland, as compared to practices encountered by workers in small and large firms. In response to calls for the further analysis of the SME ‘black-box’, we investigate the internal dynamics of medium-sized firms in Ireland, as an underexplored context of study. Consequently, we are able to offer both empirical insights and conceptual contribution to a field of growing interest which has hitherto offered fragmented or combined evidence on SMEs as a totality. In adopting medium-sized firms as our focus, we highlight the utility of disaggregating the SME label, to move beyond the historic ‘beautiful vs bleak’ SME paradox in the literature. In doing so, our study also offers contemporary methodological direction for further research into the nuanced and dynamic environments associated with understanding bullying and HR issues in medium as well as small firms and large organisations across Ireland and elsewhere. By focusing on inappropriate behaviours at work, we have shed light on those which are the most prevalent and upon the individuals who are most at risk of encountering them. As global interest rises in the lived experiences of people of colour, our findings reiterate the importance of understanding the manifestations of discrimination and the ill-treatment behaviours that might buttress them.
Footnotes
Authors’ note
Duncan Lewis is also affiliated with Edith Cowan University, Australia.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by a grant from the Institution of Occupational Safety and Health (IOSH) in the United Kingdom.
