Why and when do less assertive auditors
leave their audit firm? The importance of
work-life balance and organizational
identification.
Student: Johan Karssenberg S3506851
Supervisor: Dr. D.B. Veltrop
University of Groningen Faculty of Economics and Business
23 June 2019
Master Thesis MSc Accountancy (EBM869B20)
Abstract
This paper studies the auditor shortage many audit firms face these days, and the contemporary discussion about work-life balance in the audit sector, which received a lot of attention in recent years by both researchers and practitioners. I posit that auditors leaving the audit firm is a result of auditors’ work-life imbalance. Where previous studies have focussed on busy season and the number of clients as explanatory factors, this study focuses on a crucial auditor trait, namely the assertiveness of the auditors themselves. A study of 131 Dutch auditors shows that assertiveness is indirectly negatively related to turnover intention, mediated by work-life balance. This study shows that more assertive auditors are more likely to have a good work-life balance, and as a result will have less intention to leave their audit firm. In contrast to my expectation I do not find support for a moderating effect on auditors’ organizational identification. All in all, my results indicate that audit firms and auditors should be aware of the importance of work-life balance for auditor turnover.
Keywords: Work-life Balance, Turnover Intention, Assertiveness, Organizational Identification
2
KPMG spokesman Bonekamp confirms the sector wide turnover by stating that ‘It is hard to retain employees after their first years’ (Bonekamp, De Telegraaf, 2018)1
Introduction
The shortage of auditors continues to be a major problem facing audit firms, as the demand for financial specialists like auditors increased by 65% last year (Harskamp, 2018). An important reason for this auditor shortage originates from high turnover among them and the increasing amount of work auditors perform. Audit firms struggle to retain employees for the auditors for the long-term, mainly because their career structure is traditionally ‘pyramid’ (Flynn & Leeth, 1997). According to Flynn and Leeth (1997) the prevailing ‘up or out’ philosophy is part of the high turnover rate. This problem is also recognized in practice; the NBA (The Royal Netherlands Institute of Chartered Accountants) acknowledges a significant higher increase in turnover intention than a decade ago (Harskamp, 2018). This is apparent from a survey recently conducted among certified public accountants (abbreviated: CPA); more than 55% of the respondents stated they would leave their function on short notice (Vries & Herrijgers, 2018). These results are worrying since the financial costs of turnover comprise between 100 and 250 percent of an employee’s annual salary (Mello, 2011; Altman, 2017). Considering the high costs of turnover for employers, companies should endeavour to manage turnover strategically (Azanza, Moriano, Molero, & Mangin, 2015).
A particular feature of audit work is the busy season. Every year, auditors experience the long hours and compressed workloads of the busy season (Jones, Strand, & Wier, 2010). In this particular time of year, the majority of firms close their fiscal year and need their financial statements to be audited by an independent external auditor (PWC, 2017). The workload pressures of busy season result from the tension between limited audit resources and the need to complete a high number of audit engagements within a limited time window (López & Peters, 2012). In addition, throughout the year, auditors generally participate on multiple audit engagements simultaneously (Vries & Herrijgers, 2018). In order to meet deadlines from working on multiple engagements and multiple audit partners, individual auditors usually have to
1 De Telegraaf (2018) Accountants niet aan te slepen,
3
work long hours and spend many weekends and evenings in meetings, oftentimes, conflicting deadlines. Young auditors often even work more since they have to study alongside their engagements to achieve their title as certified public accountant (CPA) (NBA, 2019).
A recent report published by the NBA states that 39% of all CPA’s are dissatisfied with their work-life balance. This is worrying for the audit sector since work-life imbalance is not only important for the individual auditors’ intention to leave the audit firm (Baltes, Briggs, Huff, Wright, & Neuman, 1999; Dembe, 2009), but also because work-life imbalance may negatively impact audit quality since highly qualified auditors seek their happiness elsewhere (Ladva & Andrew, 2014). Work-life balance crucially affects life satisfaction and employee well-being (Allen, Herst, Bruck, & Sutton, 2000). Although family and work are essential roles in an individual’s life, external factors like work performance and overtime burden can affect their personal and professional lives (Rasheed, Iqbal, & Mustafa, 2018). Therefore, if work responsibilities influence family demands, employees may face problems that lead to work-family conflict (Lambert, Qureschi, Frank, Keena, & Hogan, 2016). In order to solve work-life imbalance, employees tend to leave their current job (Ladva & Andrew, 2014). Hence, work-life balance has a direct negative relation to turnover intention. This relation is also found in practice; Vries and Herrijgers (2018) find that work-life balance is by far the most important reason for auditors to leave the accountancy sector. In view of the importance of work-life balance in relation to individual auditors and the audit profession, the common auditing culture of working many hours is concerning (Crompton & Lyonette, 2011; Haworth & Lewis, 2005; Lewis, 2007; Lightbody, 2008). Considering the generally high work pressures and the multiple engagements auditors are engaged in, auditors have multiple and oftentimes conflicting demands. Particularly for auditors that have difficulty saying no to (multiple) audit partners or engagement leaders, it can readily be expected that these auditors work many weekends and evenings. For these unassertive auditors – less likely to say no to their managers – their work may not only seriously undermine their private life, but in so doing also likely increases their intention to leave the audit firm. In this study I postulate that assertive auditors will be better able to say no to the high workload that is prevalent in the audit sector, and hence, are more likely to have a better work-life balance; decreasing their intent to leave the audit firm.
4
In addition to assertiveness impacting auditors’ life balance, auditors’ work-life balance may not uniformly impact auditors’ turnover intention. That is, for some auditors work-life imbalance may not necessarily increase turnover intention, whereas for other it may even lead to stronger turnover intention. It is important to realize in this respect that auditors can feel a deep, self-defining affective and cognitive bond with the audit firm (Edwards & Peccei, 2007), and thus derive their social identity from their employer (Ashforth & Mael, 1989). Ashforth and Mael (1989) argue that this is a specific form of social identification, called organizational identification. Hence, for some auditors, the audit firm plays a very important role in defining who they are. The higher the organizational identification, the more important role of the audit firm in defining the identity of their auditors.
For high-identifying auditors working long hours and weekends may not only seem normal, as they incorporate the organizations characteristics (Dick, et al., 2004), but for these auditors work-life imbalance may not necessarily increase their turnover intention because working as an auditor is a central and defining part of who they are. Thus, high and low identifying auditors may experience the negative effects of work-life imbalance differently. Auditors with high organizational identification are less likely to experience the negative consequences since working for the audit firm defines who they are (Dutton, Dukerich, & Harquail, 1994). This also indicates that auditors with low organizational identification may benefit more from work-life balance, since they are more likely to suffer from work-life imbalance. Hence, I expect a moderating effect of auditor organizational identification on the relation between work-life balance and turnover intention, such that the effect of work-life (im)balance on auditor turnover intentions will be stronger for low-identifying auditors.
In sum, I anticipate that work-life balance mediates the relation between assertiveness and turnover intentions. And that the relation between work-life balance and turnover intention is moderated by organizational identification. To investigate the effect of assertiveness on turnover intentions, the following research question has been formulated:
To what extent does work-life balance mediate the relation between assertiveness and turnover intentions, and how is this relation moderated by organizational
5
Using a dataset of 131 Dutch auditors, I demonstrate that work-life balance mediates the relation between assertiveness and turnover intention. In the remainder of this research paper, the theoretical background and research method will be elucidated followed by an elaboration of the results, discussion and conclusion.
Theoretical background
In this section the relevant theory will be further elaborated on, eventually resulting in the hypothesis of this study.
Auditor turnover intention
‘Professional service firms, such as accounting and consulting firms, can be considered examples of the knowledge-based organisation’ (Morris & Empson, 1998). These organisations are characterized by their highly educated (professional) workforce which form their primary assets (Greenwood, Li, Prakash, & Deephouse, 2005). Given the importance of employees in accounting firms, the results of a recent conducted survey are worrying to say the least. The survey was conducted by the NBA among certified accountants (CPA), and more than 55% of the respondents stated that they would leave their function on short notice. As a result, all audit employers struggle to bind employees for the long term as it is hard to retain employees after their first years (Harskamp, 2018). Given the importance of the employees for accounting firms and the costs of turnover to replace them, it has become a major problem facing audit firms.
A crucial behavioural precursor for auditors leaving the audit firm, is called employee turnover intention (Keng, Nor, & Ching, 2018; Alam, 2012). Tett and Meyer (1993), Ali (2008) and Mobley et al., (1979) describe turnover intention as the willingness of an employee to leave an organization. The intention to leave an organization can arise from multiple causes: Vries and Herrijgers (2018) find among other things the following causes contribute to turnover intention: work-life balance, not interesting work, bad work environment, not enough flexibility, not enough career opportunities and salary. Before making a turnover decision, employees will compare the benefits and costs of their current employment, measure material and psychological investments in the current position and discover the quality of potential alternatives (Rusbult & Farrell, 1983).
In previous research several factors have been identified to cause turnover intention. Most studies, for example Alam (2012) and Keng et al., (2018), have found
6
a relation between job satisfaction and employees’ turnover intention. Similarly, in a study among internal auditors, fulfilling employees’ needs is seen as an important requirement for employers (Muliawan, Green, & Robb, 2009). Consistently, other studies demonstrate that several other factors determine turnover intention: the prospect of higher salary, conflicts with colleagues, personal reasons and career possibilities Martin (2010) and authentic leadership (Azanza et al., 2015).
Work-life balance
Work-life balance is important for both the individual and the organisation (The Work Foundation, 2019). For organisations, work-life balance is important for job satisfaction, productivity and the recruitment and retention of staff (Baltes et al., 1999). When it comes to the individual, work-life balance affects life satisfaction, family, health and most importantly the employees’ intent to leave an organisation (Allen et al., 2000; Dembe, 2009; Vries & Herrijgers, 2018). Since it is hard to achieve a work-life balance in multinational accounting firms, Ladva and Andrew (2014) report that still a lot of auditors have to immolate their private time to meet targets and perform at work (Ladva & Andrew, 2014). For young accountants, the challenge to combine work and life is even harder. That is, they have to study besides their ‘normal’ work to achieve their title as certified public accountant (CPA) (NBA, 2019). In addition, due to the nature of the profession, each year is characterized by a period called busy season (Sweeney & Summers, 2002). In this period, auditors work long hours, spending their weekends and evenings on work. This overtime is concerning, given that work-life balance is the most important reason to leave an audit firm (Vries & Herrijgers, 2018; Crompton & Lyonette, 2011; Haworth & Lewis, 2005). Ladva and Andrew (2014) suggest that employees with work-life imbalance all want to leave at some point, they only need the right impulse to make the decision. Such an impulse is often translated into a ‘really good’ opportunity. This opportunity should give them a work-life balance and a reward for their contemporary sacrifice (Ladva & Andrew, 2014).
As work-life balance seems to be so important for employees, it is helpful to discover why it is that important. According to Byrne (2005), employees generally ask for work-life balance as a result of a number of things. Changes in personal priorities, age, having children and nearing retirement are among those causes (Byrne, 2005). Changes like these can have a big impact on the time an employee has to devote to their family. For example, an employee who recently became a father may experience that
7
the balance he used to have is no longer sufficient to obtain work-life balance. This example indicates that work-life balance is subject to personal judgement. Hence, it is hard to identify a ‘good’ work-life balance. As work-life balance is a much studied subject, many definitions exist: Sirgy and Lee (2015) describe that work-life balance is achieved when people are fully committed in their various roles in a balanced way. Others describe work-life balance as: “the distribution of time over work and life outside of work” (Guest, 2002, p. 264), “satisfaction and good functioning at work and at home, with a minimum of role conflict” (Clark, 2000, p. 751) and “people having a measure of control over when, where and how they work” (The Work Foundation, 2019).
Regardless of the definition used, work-life balance thus matters to employees, for various reasons. Employers must acknowledge the importance of work-life balance, since it gives employees the feeling that they are valued and brings them more time to focus on life outside of work (Byrne, 2005). If firms do not see the importance of work-life balance, multiple conflicts can arise between the employer and employee as well as between the employee and his family (Greenhaus & Beutell, 1985). Both can cause stress and other issues which may result in resignation or sick leave days (Gatrell, Burnett, Cooper, & Sparrow, 2013). As work-life balance is thus very important for the wellbeing of an employee, and influences decisions to stay or leave the company, I expect a negative relation between work-life balance and turnover intention:
Hypothesis 1: work-life balance is negatively associated with employee turnover intention
Given the importance of a good work-life balance, it is very valuable to investigate the causes of a distorted balance between time for work and time for non-work activities. Until the final decade of the twentieth century, work-life balance was rather unproblematic since usually the man had a full-time job and the women were assigned to caring and domestic work (Crompton R. , 1999). However, rising levels of unemployment, de-industrialization and the shift to service employment have made considerable changes since then (Crompton & Lyonette, 2006). Many low-level service jobs were not well paid, and dual earning became an economic necessity for an increasing number of families. As a result, individuals and families had to struggle with many pressures and tensions in order to combine employment and family
8
responsibilities. These struggles indicate work-life imbalance, people cannot combine their commitment in their various social roles (Sirgy & Lee, 2015).
The same issues are faced by auditors today: it is hard for them to find a good balance between work and life. One of the most important causes lies in the characteristics of the audit: having multiple bosses. As financial statements are generally audited in the first quarter of the year, auditors have to work simultaneously in multiple engagements, each having a different boss. This would not be such an issue if there were flawless time schedules, but according to Vries & Herrijgers (2018), the audit plan is often too optimistic. For the reason that there is usually no gap between engagements for unforeseen work, this is very problematic. If for example an engagement is delayed, the members of the engagement team are likely to be asked by their boss to perform more work. Due to the tight planning, another planned engagement will start during the delay which makes it hard to complete both simultaneously. In such a situation the auditor has to be determined. He must step up for his- or herself and refuse the additional work. Otherwise, the auditor will place the needs of others before his/her own. However, this requires the capability to do so. Saying no in not an easy thing to do, especially to a superior (Gosselin, 1993). Auditors have to choose for themselves how to act, standing up for their own basic human rights in an appropriate manner (Ryan, 1981). Saying no is part of being assertive, sticking up for yourself and knowing what you want (Hofstee, Raad, & Goldberg, 1992). Being assertive thus enables auditors to reject additional work, and I therefore expect assertive auditors to have a better work-life balance.
Hypothesis 2: assertiveness is positively associated with work-life balance
As discussed before; both turnover intentions and work-life balance are known issues from the audit sector. However, if auditor assertiveness positively impacts auditor work-life balance, it may also by indirectly impact auditor turnover intention (through increased work-life balance). In this indirect relationship, work-life balance may therefore mediate the relationship between auditor assertiveness and turnover.
Hypothesis 3: Work-life balance mediates the negative effects of assertiveness on turnover intention.
9
Moderating role of organizational identification
It is argued that social identification is a perception of oneness with a group of persons (Ashforth & Mael, 1989). According to social identity theory, people tend to classify themselves and others into various social categories, such as organizational membership (Tajfel & Turner, 1985). Social categorization enables the individual to locate or define him- or herself in the social environment (Ashforth & Mael, 1989), making social identification an important part of who someone really is (Stryker & Serpe, 1982). Social identity theory proposes that individuals’ identities that stem from group membership are essential for their self-concept (Tajfel & Turner, 1979), which in turn impacts their affect, cognitions and behaviour.
According to Ashforth and Mael (1989), the individuals’ organization may provide one explanation for one’s social identity. They argue that organizational identification is a specific form of social identification, an individual may derive his or her social identity from the organization (Ashforth & Mael, 1989). Karinka-Murray, Duncan, Pontes & Griffiths (2015) define organizational identification as ‘a psychological linkage between the individual and the organization whereby the individual feels a deep, self-defining affective and cognitive bond with the organization as a social entity’ (Edwards & Peccei, 2007). Hence, for some employees, the organization plays a very important role in defining their personal identity.
As some auditors will derive much of their personal identity from their audit firms, they will get used to some characteristics of the organization. The more an employee identifies with his or her organization, the more this employee’s self-image incorporates the organization characteristics (Dick, et al., 2004). In other words, the more I identify myself with my organization, the more I define myself in terms of this particular membership (Dick, et al., 2004). As such, if it is normal to work long hours and weekends, auditors with high organizational identification will incorporate this in their self-image. For that reason, auditors with high organizational identification respond differently to work-life imbalance than those with low organizational identification. Auditors with high organizational identification are less likely to experience the negative consequences of work-life imbalance since working for the audit firm defines who they are (Dutton et al., 1994). They will have enhanced support for, and commitment and loyalty to the firm (Ashforth & Mael, 1989). In addition, Dutton et al. (1994) describe that individuals with high organizational identification are more likely to be actively involved in the organizations’ goals, and in turn be more
10
motivated to work harder to achieve these goals. Therefore, auditors with high organizational identification will experience the work-life balance as normal, and will be shielded from the negative effects of work-life balance. On the contrary, auditors with low organizational identification will derive less of their personal identity from their organizations. They less incorporate the characteristics of the organizations, have less support for, and commitment to the organization and hence, will be less shielded from the negative effects of work-life imbalance and thus suffer more (Ashforth & Mael, 1989; Dutton et al., 1994). However, as auditors with high organizational identification are shielded from the negative effects of work-life imbalance, they are also less likely to benefit from work-life balance. And although auditors with low organizational identification are not shielded from the negative effects of work-life imbalance, they are more likely to benefit from work-life balance.
A situation in which auditors with low organizational identification have work-life balance, will therefore strongly reduce their turnover intention and vice versa. Auditors with high organizational identification will be shielded from the negative effects from work-life imbalance and thus will strongly reduce their turnover intention in case of work-life imbalance. Thus, organizational identification moderates the relation between work-life balance and turnover intention, such that auditors with low organizational identification will strongly reduce their turnover intention:
Hypothesis 4: Organizational identification will moderate the negative effect of work-life balance on turnover intentions, such that the effects will be strengthened
for auditors who have low organizational identification.
By extension, if this moderating effect is supported, and given that I also postulated an indirect effect of auditor assertiveness on auditor turnover intention (operating through work-life balance), I also expect that organizational identification may similarly moderate the negative indirect effect of assertiveness through work-life balance on turnover intention. There is more likely to be a negative effect of auditor assertiveness on auditor turnover intention – through work-life balance – when auditor organizational identification is low.
Hypothesis 5: Organizational identification will moderate the negative indirect effect of assertiveness on turnover intentions, such that this effect is stronger for low
11
Figure 1: Hypothesized model
Research method
This section elaborates on the methodology used in order to conduct the research. The sample, research model and data analysis will be discussed.
Sample and procedures
A total of 135 auditors consented to participate in this study and provided their contact information to a student that contacted them. All auditors had at least 2 years of audit-experience and were currently employed by a big audit firm in the Netherlands. I used a self-administrated web-based survey to collect data. The survey with an invitation link was sent to 136 auditors. Auditors received a first reminder one week after the survey was sent out and a second reminder after two weeks. Respondents had to answer all survey questions to complete the survey. However, if auditors quit the survey early, the answered questions would be saved. From the initial group, 114 auditors completed the survey, which is equal to a response rate of 85%. The auditors that completed the survey received a personal benchmark report in which their scores were compared against the scores of other auditors who filled out the survey. Upon receiving the benchmark report, auditors had the opportunity to send a survey link to their colleagues to invite them to also participate. All in all, a total of 131 usable responses were received.
Given the sensitivity of the subjects studied, people generally tend to avoid answering, unless they believe their confidentiality and anonymity can be guaranteed (Muliawan et al., 2009). Therefore, anonymity has been assured to the participants multiple times (e.g. anonymization of given answers).
12
Measures
Participants responded to all scales using a 7-point Likert scale ranging from ‘strongly disagree’ =1, to ‘strongly agree’ =7. All of the measures used in this study were adapted from scales validated in prior research. For each measure, the Cronbach’s alpha (α) was calculated. Generally, the Cronbach’s Alpha should be higher than 0.7 in order to assure internal consistency (Hair, Black, Babin, & Anderson, 2010).
Assertiveness. ASSERT is a measure used for the assertiveness of the respondents. A 5-item unmitigated communion measure was used to capture a lack of assertiveness (Fritz & Helgeson, 1998). High unmitigated communion reflects placing others needs above one’s own, to the detriment of oneself and captures the inability to ‘say no’ to others. Example items are: ‘I can’t say no when someone asks me for help’, ‘I always place the needs of others above my own’, (α = 0,069)2. I reversed this measure such that higher values reflect higher assertiveness (i.e. lower unmitigated communion).
Work-life balance. WLB is a measure of work-life balance, indicating the extent of work-family conflict using five items. All of the statements have been used in previous research (Netemeyer, McMurrian, & Boles, 1996; Dahm, Glomb, Flaherty Manchester, & Leroy, 2015). Example items are: ‘the demands of my work interfere with my home and family life’, ‘the amount of time my job takes up makes it difficult to fulfil family responsibilities’ (α = 0,908). I reversed this measure such that higher values reflect higher work-life balance (i.e. lower work-family conflict).
Turnover intention. TURNINT is a measure of turnover intention which indicates to what extent employees want to leave the firm that they are currently working for. All items were used in previous research (Dalton, Viator, & Davis, 2015; Viator & Pasewark, 2005). Example items are: ‘I often think about leaving the firm I work for’, ‘I will probably look for a job outside of the firm I work for within the next 3 years’ (α = 0,867).
Organizational identification. ORGID is a measure of organizational identification, the extent to which respondents identify themselves with their audit firm. It is captured using 6 items that have been deployed by previous studies already (Mael & Ashforth, 1992). Example items are: ’When someone criticizes this audit firm, it feels like a
2 All questions measure the same underlying factor. The Kaiser-Meyer-Olkin test statistic has statistic 0,698, which is higher
13
personal insult’, ‘I am very interested about what others think about this audit firm’ (α = 0,874).
Control variables. In prior research, already tenure, seniority and age have been used as control variables (Azanza et al., 2015; Stoner & Kim, 2008). Azanza et al. (2015) find that employees with more organizational tenure reported lower levels of turnover intention. One explanation for this effect may be that with increasing seniority within an organization, workers are more likely to gain status. Leaving would mean that they have to build up their value in a new organization, which is in most cases less attractive then remaining their current function (Hellman, 1997). Furthermore, I deploy gender as a control variable. Multiple studies suggest a difference in gender in the area of work-life balance: ‘Particularly for women, it is difficult to balance the work burdens due to family responsibilities’ (Faiz, 2015) and ‘female employees may face more problems both at work and family’ (Lambert, Hogan, Keena, Williamson, & Kim, 2017). The final control variable is whether is based on the combination of marital status and having children. According to Byrne (2005), having children is a significant change in life, and it is assumable that it will also change work-life balance (Byrne, 2005).
Data analysis
As a start, the Cronbach’s Alpha and a confirmatory factor analysis are performed. The Cronbach’s Alpha is calculated to test internal consistency of the items. I also performed additional factor analysis to assess the validity of the measures.
In order to statistically test the hypothesis, IBM SPSS Statistics 25 was used. Hypothesis 1 and 2 were tested using an ordinary regression analysis in SPSS. To be able to test the fourth hypothesis, the independent variables (WLB & ORGID) were standardized (Aiken & West, 1991). Thereafter, a two-way interaction term is calculated using the two standardized independent variables (Dawson, 2014). Using the two-way interaction term, I tested hypothesis 4. Finally, to test the mediation (hypothesis 3) and the moderated mediation (hypothesis 5), I used a macro developed by Hayes (2017). Hayes wrote the process macro as a tool for implementing the methods discussed in the book (e.g. moderated mediation). The macro is accessible in SPSS.
14
Results
This section elaborates on the results from the conducted research. Firstly, descriptive statistics will be discussed after which I test the hypothesis.
Descriptive statistics
Table 1 shows descriptive statistics and Pearson’s correlations among all study variables. On average, the respondents rated their work-life balance 3,74 out of 7, 27% of the respondents was a female and 26% was married and had children. As shown in Table 1, turnover intention and work-life balance are significantly negatively related, indicating that auditors with a better work-life balance have less intent to leave their current employer. Assertiveness and work-life balance are significantly positively related, signifying that auditors that are more assertive have a better work-life balance. Seniority and parenthood are positively related, which is logically given the fact that the average age to have a baby for women is 29,9 years3 (CBS, 2019). Furthermore, although not hypothesized, organizational identification is significantly negatively related to turnover intention.
Table 1: Means, Standard Deviations and Correlations Among Variables
Variable Mean S.D. 1 2 3 4 5 6 1 ASSERTV 3,88 0,96 - 2 WLB 3,74 1,23 ,22* - 3 TURNINT 3,71 1,49 -,14 -,41*** - 4 ORGID 4,82 1,09 -,19* ,13 -,36*** - 5 Seniority 3.69 1.24 -,17 ,17 -,06 ,14 - 6 Gender 0,27 0,45 ,00 ,03 ,12 ,04 -,11 - 7 Parenthood 0,26 0,44 ,11 ,12 -,06 ,19* ,57*** -,052 n=131, *p <.05, **p <.01 , ***p <.00
15
Table 2: Hierarchical Analysis Examining Moderating Effects of Organizational Identification and Work-Life Balance on Turnover Intention
Turnover Intention Model 1 Model 2 Model 3
Step and variables B SE B SE B SE
Intercept 3,78*** (0,47) 3,40*** (0,41) 3,41*** (0,41) Control Seniority -0,04 (0,13) 0,05 (0,11) 0,04 (0,11) Gender 0,37 (0,30) 0,41 (0,26) 0,40 (0,26) Parenthood -0,12 (0,36) 0,11 (0,32) 0,12 (0,32) Main effects Work-life balance -0,56*** (0,12) -0,55*** (0,12) Organizational identification -0,49*** (0,12) -0,48*** (0,12) Two-way interaction Work-Life Balance x Organizational Identification 0,09 (0,11) R Square ,02 ,28 ,29 ∆ R Square ,26*** ,00
n=131. One-tailed test for hypothesized effects and two-tailed for controls. *p <.05, **p <.01 , ***p <.001
In table 2, the moderating effects of organizational identification and work-life balance on turnover intention are shown. According to hypothesis 1, there will be a negative relation between work-life balance and turnover intention. As presented in table 2 (model 2), the coefficient of work-life balance is negative and highly significant (b = -0,56, p <0,001), a result that is consistent with the hypothesis.
Table 3 presents the results of the regression analysis for work-life balance. According to hypothesis 2, assertiveness is positively associated with work-life balance. As shown in table 3 (model 2), the coefficient of Assertiveness is positive and significant (b = 0,25, p <0.05), a result that confirms the hypothesis.
According to hypothesis 4, organizational identification will moderate the negative effect of work-life balance on turnover intentions, such that the effects will be strengthened for auditors who have lower organizational identification. In order to test this hypothesis, I added an interaction term between work-life balance and
16
organizational identification to model 3 in table 2. To simplify interpretation, I plotted the relationship between organizational identification and turnover intention. Figure 2 presents the resulting graph.
Figure 2: Work-life Balance and Turnover Intention at Different Values of Organizational Identification
The results signify no significant relation between the interaction variable and turnover intention (b = 0,09, p = 0,39). Hence, I cannot accept the fourth hypothesis. Model 3 (table 2) also demonstrates that the relation between organizational identification is highly significant. Hence, although not hypothesized, also a highly significant relation was found between the moderator, organizational identification, and turnover intention.
Table 3: Results of Regression Analysis on Work-life Balance
Work-life Balance Model 1 Model 2
Step and variables B SE B SE
Intercept 3,20*** (0,38) 2,37*** (0,53) Control Seniority 0,14 (0,11) 0,11 (0,11) Gender -0,04 (0,24) -0,05 (0,24) Parenthood 0,10 (0,30) 0,09 (0,29) Main effect Assertiveness 0,25* (0,11) R Square ,03 ,06 ∆ R Square ,04*
n=131. One-tailed test for hypothesized effects and two-tailed for controls. *p <.05, **p <.01 , ***p <.001 1 2 3 4 5 Low WLB High WLB T u rn ove r In te n tion Low ORGID
17
I present the indirect effects estimated through bootstrapping, in order to test whether assertiveness affects turnover intention through work-life balance. Bootstrapping methods are more powerful than stepwise procedures and generate the recommended bias corrected confidence intervals (Veltrop, Hermes, Postma, & Haan, 2015). By applying bootstrap procedures, it is possible to gauge the significance of the indirect effect while avoiding power problems stemming from non-normal sampling distributions of the indirect effect (Edwards & Lambert, 2007). The results are presented in table 4 and show that the 95 percent confidence interval does not include zero for work-life balance (-0,255, -0,10) and hence, hypothesis 3 receives support.
Table 4: Results for Indirect Effects on Turnover Intention through Work-Life Balance.
Boot indirect effect Bootstrap 95% confidence interval Lower bound Upper bound
WLB -0,12 -0,255 -0,010
N=131. Unstandardized coefficients are reported. Bootstrap sample size is 5,000. Bootstrap 95% confidence interval.
Table 5: Results for Conditional Indirect Effect on Turnover Intention Conditional indirect effect exerted on turnover intention through work-life balance
Turnover intention Boot indirect effect Bootstrap 95% confidence interval Lower bound Upper bound Conditional indirect effect at turnover intention = M +/-SD
0,16 * M (3,67) -0,12 -0,27 -0,01
M (5) -0,10 -0,21 -0,01
0,84 *M (6) -0,08 -0,19 0,00
n = 131. Bootstrap sample size is 5,000. Bootstrap 95% confidence interval.
In order to acquire additional information about the nature of the interaction effect, I divided the relationship between work-life balance and turnover intention at low (16th
percentile), high (84th percentile) and average values of work-life balance. Table 5
presents these conditional indirect effects for turnover intention. The 95 percent bootstrapped confidence interval excludes zero for both low (-0,27, -0,01) and mean
18
work-life balance (-0,21, -0,01). However, when work-life balance is high, the bootstrap does not exclude zero (-0,19, 0,00). These results indicate that when work-life balance is low or at the mean, the indirect effects are significant for turnover intention. Thus, for turnover intention, I observe a negative indirect effect of assertiveness through work-life balance when work-life balance is low or at the mean, but not when work-life balance is high. Hence, these results support Hypothesis 5. In table 6, an overview of the results of the hypothesized model is presented.
Table 6: Results Hypothesized Model Hypothesis Independent variable Dependent variable Coefficient Conclusion H1 WLB TURNINT - 0,56 Supported at p < 0.001 H2 ASSERT WLB 0,25 Supported at p < 0.05 H3 ASSERT/WLB TURNINT - 0,12 Supported at p < 0.05 H4 ORGID*WLB TURNINT 0,09 Not Supported
H5 ORGID/ASSERT TURNINT Conditionally Supported4
Discussion and Conclusion
In this study I investigated the indirect effects of assertiveness on auditor turnover intention using work-life balance as mediating variable and the moderating effect of organizational identification on this mediation. Existing literature is primarily focussed on the determinants and effects of work-life (im)balance (Byrne, 2005; Allen et al., 2000). I add to this literature by demonstrating the mediating effects of work-life balance, instead of solely focusing on work-work-life balance. This study displays that work-life balance is negatively related to turnover intention and positively related to assertiveness. Assertiveness can be viewed as a cause of work-life balance, whilst turnover intention can be seen as outcome of work-life imbalance. Thus, this papers adds to existing literature by establishing a mediated relation between assertiveness and turnover intention.
19
Implications and future research
Researchers have performed a lot of studies to gain insight in work-life balance and its effects (e.g. (The Work Foundation, 2019; Dahm et al., 2015). Although the distribution of time between work and life seems to be a present-day discussion only, it has been subject to empirical research for many years (e.g. Zedeck & Mosier 1990). Since then, a wide variety of explanatory variables have been found in research. For example, Rasheed et al. (2018) found that gender and received social support are associated with work-life balance. Given the vast amount of research on work-life balance, it is rather surprisingly that only very few studies have focussed on the effects of individual character traits. This study extends research on work-life balance by illustrating that assertiveness is positively related to work-life balance.
This study makes several contributions to research on turnover intentions. First, as the financial costs of actual turnover are relatively high (Altman, 2017), much studies have been performed to identify determinants of turnover intention (e.g. (Muliawan et al., 2009). Presumably the most important factor identified yet is job satisfaction (Keng et al., 2018; Alam, 2012). My study contributed to this research stream with additional evidence for the negative effects of work-life balance on turnover intentions. Work-life balance was highly significant negatively related to turnover intentions. As auditor shortage is a major problem facing audit firms at this moment, my study is relevant for audit firms as it provides a better understanding of the issue. Work-life balance is an explanatory variable that explains much of the variance in turnover intention, so audit firms must try to improve employees’ work-life balance in order to retain employees. Similarly, by emphasizing the importance of work-life balance to audit firms, research on auditing can be expanded using this study. As this is one of the first studies to relate assertiveness and work-life balance to auditor turnover intention, future auditing research may examine other determinants of turnover in the audit sector.
Second, considering that assertiveness is significantly positive related to work-life balance, and work-life balance and turnover intention are negatively significant associated, this study extends research to turnover intention by proving an indirect negative relation between assertiveness and turnover intention, mediated by work-life balance. Although researchers have identified a lot of factors that influence turnover intention (Hellman, 1997; Stoner & Kim, 2008), no study has yet related it to assertiveness. As there is much to explore, many possibilities for future research are
20
open. An interesting avenue for future research would be to examine more individual character traits (e.g, leadership, poise), and explore the boundary conditions of this association in terms of position (e.g. staff or manager) and period of the year (e.g. busy season). After all, assertiveness is only one possible trait an auditor can have, and many other factors play a role in making turnover decisions. This study can have practical implications for audit firms as it provides further insight into determinants of turnover intentions. Audit firms must acknowledge the importance of work-life balance, and hence, consider changing their HR strategy. Additionally, future research may also focus on other predictors of work-life balance itself. A number of studies have already related a diversity of variables to work-life balance (e.g. parenthood and marital status) (Dahm et al., 2015), and hence, also other factors than auditor traits can predict work-life balance.
Finally, although not hypothesized, the results of this study point out that organizational identification is negatively related to turnover intention. I conceptualized organizational identification as moderator, but the results show a stronger direct effect than the hypothesized indirect effect. This relation was, however, already established by Bamber & Venkataraman (2002) who found that organizational identification was significantly negatively related to turnover intention at the 0.05 level. As such, this research provides the same relation with a higher significance; at the 0.001 level.
Limitations
The implications of this study should be considered in light of its limitations. A main limitation of this research is its cross-sectional design. Although my theoretical model implies a specific causal order, the cross-sectional design only allows for testing whether the relationships are consistent with my hypothesis.
This study was conducted during the so called ‘busy season’, which may have biased the results. First, in busy season, auditors work long-hours and weekends to finish their assignments. As such, the busiest auditors might not have responded to the survey, thereby excluding the extremes of the sample. However, given that only the busiest auditors did not participate, hypothesis testing with these auditors included would have been even more significant. Second, auditors may have responded differently if the survey would have been conducted outside of busy season, considering that they would have had a better work-life balance in that period. This limitation is
21
partly mitigated by the fact that all surveys were filled in during busy season, so there are no differences between the respondents.
Many complaints regarding work-life balance comes from young auditors that have just started both studying for their title as registered public accountant, as well as working for their audit firm. For the reason that the study oftentimes only takes two years to finish, the group that has a hard time combining obligations both at school and at home are excluded from the sample. As the issues in the audit sector do not only arise for people with two years of experience, I think it is better to include those in future research as well. However, as most of these new employees are generally new to a full time job, it may be justified as they do not know yet what distribution of time is work-life balance for them. Nevertheless, I suggest that in future research also these auditors should be included.
The sample of this survey was mainly composed of auditors employed by big audit firms, making it a rather homogeneous group. As work-life balance is mostly an issue for the biggest four audit firms in the country (Vries & Herrijgers, 2018), a more equally divided sample could have increased validity. In that case, testing for both big four and non-big four would have been enabled and the findings could have been reflected to all Dutch audit firms.
Acknowledgement
I highly appreciate the helpful comments of Dr. D. B. Veltrop during the entire process of writing this Master Thesis. I also thank Vincent Koele and Sjoerd Harmens for reviewing and guidance.
References
Aiken, L., & West, S. (1991). Multiple regression: testing and interpreting
interactions.
Alam, S. (2012). Factors affecting job satisfaction, motivation and turnover rate of medical promotion officer (mpo) in pharmaceutical industryL a study based in khulna city. Business Intelligence Journal, Vol 1, pp.126-131.
Allen, T., Herst, D., Bruck, C., & Sutton, M. (2000). Consequences Associated With Work-to-Family Conflict: A Review and Agenda for Future Research. Journal
22
Altman, J. (2017). How much does employee turnover really cost? Huffpost.
Ashforth, B., & Mael, F. (1989). Social identity theory and the organization. Academy
of Management Review, Vol. 14 No. 1, pp. 20-39.
Azanza, G., Moriano, J., Molero, F., & Mangin, J.-P. (2015). The effect of authentic leadership on turnover intention. Leadership & Organization Development
Journal, Vol. 36 No. 8, pp. 955-971.
Baltes, B., Briggs, T., Huff, J., Wright, J., & Neuman, G. (1999). Flexible and compressed workweek schedules: a meta-analysis of their effects on work-related criteria. Journal of Applied Psychology, Vol. 84 No. 4 pp. 496-513. Byrne, U. (2005). Work-life balance. Why are we talking about it at all? Business
Information Review, Vol. 22, No. 1, pp.53-59.
CBS. (2019, May 11). Leeftijd moeder bij eerste kind stijgt naar 29,9 jaar. Retrieved from CBS: https://www.cbs.nl/nl-nl/nieuws/2019/19/leeftijd-moeder-bij-eerste-kind-stijgt-naar-29-9-jaar
Clark, S. (2000). Work/family border theory: a new theory of work/family balance.
Human Relations, Vol. 53, pp. 747-770.
Crompton, R. (1999). Restructuring gender relations and employment. Oxford: Oxford University Press.
Crompton, R., & Lyonette, C. (2006). Work-life 'balance' in Europe. Acta Sociologica, Vol. 49, No.4, pp. 379-393.
Crompton, R., & Lyonette, C. (2011). Women's career succes and work-life adaptions in the accountancy and medical professions in Britain. Gender, Work &
Organization, Vol. 18 No. 2, pp. 231-254.
Dahm, P., Glomb, T., Flaherty Manchester, C., & Leroy, S. (2015). Work-family conflict and self discrepant time allocation at work. Journal of Applied Psychology, Vol. 100, No. 3, pp.767-792.
Dalton, D., Viator, R., & Davis, A. B. (2015). The joint effect of unfavorable supervisoty feedback environments and external mentoring on job attitudes and job outcome in the public accounting profession. Behavioral research in
accounting, Vol. 27, No. 2, pp.53-76.
Dawson, J. (2014). Moderation in management research: what, why and how. Journal
of Business and Psychology, Vol. 29, pp. 1-19.
Dembe, A. (2009). Ethical Issues Relating to the Health Effects of Long Working Hours. Journal of Business Ethics, Vol. 84 No. 2, pp. 149-169.
23
Dick, R. V., Christ, O., Stellmacher, J., Wagner, U., Ahlswede, O., Grubba, C., . . . Tissington, P. (2004). Should I stay or should I go? Explaining turnover intentions with organizational identification and job satisfaction. British
Journal of Management, Vol.15, pp. 351-360.
Dutton, J., Dukerich, J., & Harquail, C. (1994). Organizational images and member identification. Administrative Science Quaterly, Vol. 39, No.2, pp. 239-263. Edwards, J., & Lambert, L. (2007). Methods for integrating moderation and
mediation: a general analytical framework using moderated path analysis.
Psychological Methods, Vol 12, pp. 1-22.
Edwards, M., & Peccei, R. (2007). Organizational identification: development and testing of a conceptually grounded measure. European Journal of Work and
Organizational Psychology, Vol. 16, No. 1, pp.25-57.
Faiz, R. (2015). Work-family conflict: a case study of women in Pakistani banks. Flynn, P., & Leeth, J. (1997). The accounting profession in transition. CPA Journal,
Vol. 67, Issue, 5 pp.42-47.
Fritz, H., & Helgeson, V. (1998). Distinctions of unmitigated communion from cómmunion: self-neglect and overinvolvement with others. Journal of
Personality and Social Psychology, Vol. 75, No. 1, pp.121-140.
Gatrell, C., Burnett, S., Cooper, C., & Sparrow, P. (2013). Work-life balance and parenthood: a comparative review of definitions, equity and enrichment.
International Journal of Management Reviews, Vol.15, pp 300-316.
Gosselin, T. (1993). Negotiating with your boss. Training and development, 37-41. Greenhaus, J., & Beutell, N. (1985). Sources of conflict between work and family roles.
Academy of Management Review, Vol. 10, No. 1, pp76-88.
Greenwood, R., Li, S., Prakash, R., & Deephouse, D. (2005). Reputation, diversificationm and organizational explanations of performance in professional service firms. Organization Science, Vol. 16, No. 6, pp. 661-673. Guest, D. (2002). Perspectives on the study of work±life balance. Social Science
Information, 255-279.
Hair, J., Black, W., Babin, B., & Anderson, R. (2010). Multivariate Data Analysis. Prentice Hall.
Hall, D., Schneider, B., & Nygren, H. (1970). Personal factors in organizational identification. Administrative Science Quarterly, Vol. 15, pp. 176-190.
24
Haworth , J., & Lewis, S. (2005). Work, Leisure and Well-Being. Britisch Journal of
Guidance & Counselling, Vol. 33 No. 1, pp 67-79.
Hayes, A. (2017). Introduction to mediation, moderation, and conditional process
analysis. Guilford Press.
Hellman, C. (1997). Job satisfaction and intent to leave. Journal of Social Psychology, Vol. 137, No. 6, pp.667-689.
Hofstee, W., Raad, d. B., & Goldberg, L. (1992). Integration of the Big Five and Circumplex Approaches to Trait Structure. Journal of Personality and Social
Psychology, 146-163.
Jones, A., Strand, C., & Wier, B. (2010). Healthy lifestyle as a coping mechanism for role stress in public accounting. Behavioral Research in Accounting, 21-41. Kaiser, H. (1970). A second generation little jiffy. Psychometrika, Vol. 35, No. 4, pp.
401-415.
Karanika-Murray, M., Duncan, N., Pontes, H., & Griffiths, M. (2015). Organizational identification work engagement, and job satisfaction. Journal of Managerial
Psychology, Vol.30 No8., pp.109-1033.
Keng, T., Nor, N., & Ching, Y. (2018). Turnover intention and job satisfaction among quantaty surveyors. International Journal of Technology, Vol. 9, No. 8, pp. 1551-1560.
Ketchand, A., & Strawser, J. (2001). Multiple dimentions of organizational commitment: implications for future accounting research. Behavioral Research
in Accounting, Vol. 13, pp.221-251.
Ladva, P., & Andrew, J. (2014). Weaving a web of control: 'The Promise of Opportunity' and work-life balance in multinational accounting firms'. Accounting, Auditing
& Accountability Journal, Vol. 27 No. 4, pp. 634-654.
Lambert, E., Hogan, N., Keena, L., Williamson, L., & Kim, B. (2017). Exploring the association between different types of social support with role stress, work-family conflict, and turnover intent among private prison staff. Journal of
Aplied Security Research, Vol. 12, No.2, pp.203-223.
Lambert, E., Qureschi, H., Frank, J., Keena, L., & Hogan, N. (2016). The relationship of work family conflict with job stress among Indian police officers: a resarch note. Police Practice and Research, Vol. 18, No. 1, pp.37-48.
25
Lewis, S. (2007). Working time, client time and family tIme: accounting for time in the accountancy profession. in van der Lippe, T. and Peters, P. (Eds), Competing claims in work and family. 143-160.
Lightbody, M. (2008). No time to stand and stare: imagery of flexible work arrangements in Australien professional accounting journals. Pacific
Accounting Review, Vol. 20 No. 2, pp 138-157.
López, D., & Peters, G. (2012). The effect of workload compression on audit quality.
Auditing: A Journal of Practice & Theory, Vol. 31, No. 4., pp. 139-165.
Mael, F., & Ashforth, B. (1992). Alimni and their alma mater: A partial test of the reformulated model of organizational identification. Journal of Organizational
Behavior, Vol. 13, No. 2, pp.103-123.
Mello, J. (2011). Strategic Human Recource Management. Cincinnati, OH: Cengage Learning.
Meyer, J., & Allen, N. (1997). Commitment in the workplace: Theory, Research and
Application.
Meyer, J., Allen, N., & Gellatley, I. (1990). Affective and continuance commitment in the organization: evaluation of measures and analysis of concurrent and time-lagged relations. Journal of Applied Psychology, Vol. 75, No. 6, pp.710-720. Morris, T., & Empson, L. (1998). Organisation and expertise: an exploration of
knowledge bases and the management of accounting and consulting firms.
Accounting, Organizations and Society, Vol. 23, No. 5/6 pp. 609-624.
Muliawan, A., Green, P., & Robb, D. (2009). The turnover intentions of information system auditors. International Journal of Accounting Information Systems, 117-136.
NBA. (2019). The Dutch educational system for chartered accountants. Retrieved from NBA: https://www.nba.nl/over-de-nba/english-information/ra-qualifications/the-dutch-educational-system-for-register-accountants/
Netemeyer, R., McMurrian, R., & Boles, J. (1996). Development and validation of work-family conflict and family-work conflict scales. Journal of Applied
Psychology, Vol. 81, No. 4, pp. 400-410.
PWC. (2017, May). Understanding afinancial statement audit. Retrieved from PWC:
26
Rasheed, M., Iqbal, S., & Mustafa, F. (2018). Work-family conflict and female employees' tunover intentions. Gender in Management: An International
Journal, Vol. 33, No. 8, pp.636-653.
Rusbult, C., & Farrell, D. (1983). A Longitudal test of the investment model: the impact on job satisfaction, job commitment and turnover of varations in rewards, costs, alternatives and investments. Journal of Applied Psychology, Vol. 68 pp. 429-438.
Ryan, M. (1981). The practice of assertiveness. Occupational Health Nursing, 7-9, 47. Sirgy, M., & Lee, D.-J. (2015). Work-life balance: a quality-of-life model. Applied
Research Quality Life, 1059-1082.
Stoner, M., & Kim, H. (2008). Burnout and turnover intention among social workers: effects of role stress, job autonomy and social support. Administration in Social
Work, Vol. 32, No. 3 pp.5-25.
Stryker, S., & Serpe, R. (1982). Commitment, identity salience, and role behavior: Theory and research example. Personality, Roles, and Social Behavior, pp. 199-218.
Sweeney, J., & Summers, S. (2002). The effect of busy season workload on public accountants' job burnout. Behavioral Research In Accounting, Volume 14, pp. 223-245.
Tajfel, H., & Turner, J. (1979). An integrative theory of intergroup conflict. pp. 33-47. Tajfel, H., & Turner, J. (1985). The social identity theory of intergroup behavior. The Work Foundation. (2019). Definition of work-life balance. Retrieved from The
Work Foundation:
www.theworkfoundation.com/difference/e4wlb/definition.aspx
Veltrop, D., Hermes, N., Postma, T., & Haan, J. d. (2015). A tale of two factionsL why and when factional demographic faultlines hurt board performance. Corporate
GovernanceL An International Review, Vol. 2, No. 4 .
Viator, R., & Pasewark, W. (2005). Mentorship separation tension in the accounting profession: the consequences of delayed structual separation. Accounting,
Organizations and Society, Vol. 30, pp.371-387.
