Explaining corporate art collecting attitudes of employees : the role of employees’ cultural interests and experiences of the organization

(1)

Explaining corporate art collecting attitudes of employees

The role of employees’ cultural interests and experiences of the organization.

Lucia Riscado Cordas

6064892

22nd_{of June, 2018 | Final version}

Master thesis

Msc. in Business Administration

Track: Entrepreneurship and Management in the Creative Industries Thesis supervisor: dhr. J.F.E. de Groot MSc

(2)

2

Statement of Originality

This document is written by Student Lucia Riscado Cordas, who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document are original and that no sources other than those mentioned in the text and its references have been used in

creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

3

Abstract

The purpose of the present study is to examine the effects of individuals’ cultural interests outside and employees’ experiences inside of the organization on employee attitudes of the corporate art collection. A survey was conducted among a group of nine organizations in the Netherlands owning a corporate art collection and yielded valid data from 300 employees. By using a quantitative research method, this study provides empirical evidence that individuals’ participation in a diverse set of cultural activities and employees’ organizational identification positively relates to their appreciation of the corporate art collection. Additionally, employee perceptions of intrinsic art collection motives was found to fully mediate the relationship between organizational identification and employee attitudes of the corporate art collection. By observing a corporate art collection as both a cultural and corporate activity, this study was able to present a wide outlook on the determinants of employee attitudes and present new perspectives with respect to corporate art collections. Organizations can benefit from this study by enhanced understanding on the determinants of employee attitudes of the corporate art collection, which is important for preventing negative employee attitudes.

(4)

4

1. Introduction

In the past decades, organizations from around the world have increasingly been investing in a wide range of non-core activities such as environmental care-taking programs, collaborations with diverse non-profits, projects to enhance local communities or employee well-being programs (Baird & Meshoulam, 1988; McGuire et al., 1988; De Roeck et al., 2014). Academic research has approached these organizational non-core activities with heightened interest, where academics originally focused primarily on external stakeholders. The effect on, and determinants of, employee attitudes towards these corporate non-core activities are of more recent academic interest (De Roeck et al., 2014).

A novel phenomenon is corporate art collecting, which has become a popular trend among large organizations (Kottasz et al., 2008). Around the end of the 20th century, half of the Fortune 500 companies were investing their resources in a sophisticated art collection (Kottasz et al., 2007). The corporate art collection (CAC) has received marginal academic attention, and literature has mainly focused on the motivations for organizations to spend their money on distinctive artworks. These motivations may include improving the work environment (Yoon & Shin, 2014), investing in valuable assets, enjoying tax benefits (Wu, 2003) or creating a favorable reputation among important stakeholders (Kottasz et al., 2007).

Organizational non-core activities, to which owning a CAC can be ascribed, have its roots in stakeholder theory (Freeman, 1984). Stakeholder theory suggests that corporate survival depends on the attitudes of any group of individuals that have a stake in the organization’s activities (Freeman 1984). Therefore, it is important for organizations to know about the determinants of employee attitudes of the CAC, in so that they know how to respond to employees’ needs and ensure long-term legitimacy

(6)

6

and survival of their organization. Although prior work has studied corporate art collecting motives, very few have explored determinants of employee attitudes of the CAC. The purpose of the present study is to address this important gap, which has led to the formulation of the following research question:

“To what extent do individuals’ cultural interests and experiences of the organization explain employee attitudes of the corporate art collection?”

On the one hand, corporate art collecting is a cultural activity in the sense that CACs are nowadays being managed and displayed in such a professional manner that it has become highly similar to the way artworks are being managed and presented in museums (Wu, 2003). Likewise, the content of CACs usually consists of highly valuable art pieces that have similar features to the professional works in museums (Kottasz et al., 2007). On the other hand, corporate art collecting is a corporate activity in the sense that a CAC is part of an organization’s goals and output. An organization can be viewed as an important socializing agent that through its work environment gives rise to various employee experiences (Patterson et al., 2004;

Mowday, 2013). Since corporate art collecting is thus both a cultural and a corporate activity, this thesis is able to contribute to different streams of literature.

First of all, this thesis will contribute to cultural consumption literature by examining to what extent individuals’ cultural interests outside of the organization are of influence when explaining employee attitudes of the CAC. Relevant literature on cultural consumption patterns suggests that people who only consume highbrow forms of art no longer exist (Peterson & Kern, 1996; Stichele & Laermans, 2005; Berghman & Van Eijck, 2009). More specifically, formerly perfect snobs that are

(7)

7

solely oriented towards highbrow genres no longer restrict their cultural participation to highbrow consumption, but nowadays participate in lowbrow cultural activities as well (Peterson and Simkus, 1992). In accordance, this thesis will investigate if individuals’ participation in a diverse composition of cultural activities (highbrow and lowbrow) is positively related to their attitudes of the CAC. In doing so, this thesis attempts to confirm the omnivore hypothesis originally developed by Peterson (1992) and broadens the scope of literature by integrating the organization as legitimate art supplier.

Relevant literature on organizational identification (OI) argues that when employees identify more strongly with the organization, they are more likely to believe that the organization is producing valuable outputs (Dutton et al., 1994) and obtain supportive attitudes towards the organizations’ activities (Lee, 1971). Over the years, corporate art collecting has transformed into a professionalized non-core corporate activity with the usual installment of an in-house curator and a specific budget dedicated to art purchases (Wu, 2003). Within this context, this thesis will contribute to literature on OI and corporate art collecting as a legitimate non-core activity by studying the expected positive relationship between organizational identification and employee attitudes of the CAC.

Additionally, this thesis expects perceived intrinsic art collection motives to mediate the positive relationship between OI and employee attitudes of the CAC. Perceived art collection motives have scarcely been empirically researched. Academic research within the field of employee perceptions of organizational motives has mainly focused on other corporate non-core activities such as social responsibility activities (CSR) or human resource management (HRM) practices (De Roeck et al., 2014; Story & Neves, 2015). Furthermore, organizational identification as antecedent

(8)

8

of employee perceptions and attitudes is not widely empirically researched, as academics have mainly focused on organizational identification as an outcome.

Next to examining the influence of subjective employee experiences of the organization, this thesis strives to develop a coherent research framework and therefore will also take the objective measure of status consistency into account. As objective reference, status consistency occurs when individual's ranks on all status hierarchies are consistent with one another (Lenski, 1954). Status consistency is not a widely researched construct, and no studies have been found that elaborate on the relationship between status consistency and the degree that an employee values the organization’s investments in non-core activities, such as corporate art collecting. This thesis expects the positive relationship between OI and employee attitudes of the CAC to be stronger when individuals are conditioned with status consistency.

To reveal relationships between the variables, this thesis uses a quantitative research method. In accordance, a survey was conducted among a group of nine firms and yielded valid data from 300 employees that work for organizations owning an art collection. The population consists of employees in the Netherlands.

The remainder of this thesis is organized as follows. Chapter two provides a literature review that is used to analyze the thesis’s key theoretical constructs and develop the proposed hypotheses. The conceptual model, research methodology and measures will be discussed in chapter three. Chapter four will present the results of the quantitative analysis. There after a discussion in chapter five will follow that elaborates on the study results, managerial implications, limitations of the study, directions for future research and contributions to existing literature. Finally, the thesis is completed with a conclusion.

(9)

9

2. Literature Review

2.1. Corporate art collections and employee attitudes

Large organizations have expanded their investments in artworks over the past decades, and since the end of the 20th_{century organizations have even been viewed as}

the most active buyers of art (Yoon & Shin, 2014; Sussman, 2017). With organizations showing increased commitment to corporate art collecting, thereby heightening the professionalism of this activity, researchers’ perception about the role and value of the CAC has shifted. Through hiring their own curators, installing complete art departments and sometimes even arranging their art collections to tour around the globe (Wu, 2003), organizations are successfully transforming the CAC into a highly professional museum-like activity.

Employees are repeatedly mentioned as an important motive for organizations to invest in a CAC through improving employees’ work environment or stimulate employees’ cultural and creative development. Researchers have however mainly observed this phenomenon as a vehicle for organizations to enhance their reputations, thereby addressing CAC’s primarily from the view of organizations’ external stakeholders (Martorella, 1990, Wu, 2003, Yoon & Shin, 2014). A more in-depth understanding of employee attitudes towards CACs is thus lacking. Research on employee attitudes with respect to other well-known non-core activities such as CSR activities or HRM practices have received much more academic interest.

Understanding what determines positive employee attitudes towards the CAC makes good business sense, because positive attitudes towards the organization’s activities are found to be of great importantce for ensuring long term organizational survival (Bagger & Li, 2014). From the perspective of the organization, employee attitudes can significantly affect workplace sentiment, extra-role behavior, job

(10)

10

performance and turnover rates (Eisenberger et al., 1986). From the perspective of the employee, positive attitudes can significantly affect job satisfaction and emotional empowerment (Lamm et al. 2015). As a consequence, both researchers and practitioners can greatly benefit from gathering better knowledge on the factors that are determinant in influencing positive employee attitudes of the CAC.

(11)

11

2.2 Individuals’ cultural consumption: From snob to omnivore

This chapter discusses the possible effect of individuals’ cultural interests outside of the organization on their attitude of the CAC. That is, how to best predict employees’ appreciation of the CAC based on their cultural consumption patterns outside of the organization.

Traditionally, within the realm sociological research, cultural consumption patterns have mainly been studied as markers of social status (Bourdieu, 1984). However, changes in the process of status attainment have affected peoples’ cultural tastes and as a result, peoples’ consumption patterns (Roose et al., 2012). First of all, the expansion of secondary and higher education, that is now much more accessible to the great mass of the population, has led to increased social mobility (Prior, 2013). This has caused a loosening effect of the educational system on inequalities by providing opportunities for individuals from lower class backgrounds (Goldthorpe, 2007). Secondly, technological developments have had a pronounced effect on cultural consumption patterns and tastes. For example, the digitalization of music has melted genre categories by giving consumers from all classes access to an enormous amount of content (Beer & Sandywell, 2005).

In the above-described context, academics argue that formerly perfect snobs that are solely oriented towards highbrow genres no longer exist (Goldthorpe, 2007; Roose et al., 2012; Prior, 2013). With his concept of the cultural omnivore, Richard Peterson (1992) was a pioneer in identifying this shift of cultural consumption patterns in modern society. His omnivore hypothesis contends that the former perfect snobs nowadays participate in a diverse set of cultural genres, which they perceive to be beneficial for their social status (Peterson and Simkus 1992). Accordingly, in his research on the relationship between status hierarchy and musical consumption

(12)

12

patterns in the U.S., Peterson found that formerly perfect snobs were not any longer solely oriented towards highbrow music genres but also consumed music from more popular lowbrow genres (Peterson, 1992).

Originally, omnivorousness is argued to be about crossing the boundaries of the social status hierarchy, and thus is about the composition of products from different categories (Warde at al, 2007). Within this context, researchers identified the importance of determining relevant taste boundaries (Warde at al, 2007). Bourdieu (1984) distinguished two distinct cultural categories that have been utilized by various researchers in their research on cultural consumption patterns (Claessens & Dhoest, 2010). According to Bourdieu (1984), lowbrow cultural products are being consumed by the masses and can be characterized by simple storylines with a happy ending and uncomplicated characters. On the other hand, ambiguous characters, complicated storylines and complex situations characterize highbrow cultural products (Bourdieu, 1984).

After the publishing of Peterson’s omnivore hypothesis, there has been much debate and empirical research aiming to confirm this argument and to elaborate on the significance of these changed cultural consumption patterns (Van Rees et al., 1999; Stichele & Laermans, 2005; Berghman & Van Eijck, 2009). Relish (1997) examined individuals’ attitudes towards eighteen music genres using data from the 1993 General Social Survey. His findings support the omnivore hypothesis, as they indicate that a significant number of respondents have a cultural taste for lowbrow products that would traditionally not be associated with their social status (Relish, 1997).

Rees et al. (1999) investigated the existence of the omnivore within Dutch society, focusing on the cultural activity of reading in individuals’ free time. For this they used time budget data from the 1990 Dutch Time Budget Survey from over 3000

(13)

13

respondents. Their study was able to identify one group with omnivore characteristics, consuming lowbrow and highbrow books. This finding indicates proof for the omnivore argument, however the authors are hesitant to make this statement, because the focus of their research is on one singly cultural activity (Rees et al., 1999).

Vander Stichele & Laermans (2005) explored cultural consumption patterns of individuals living in Flanders, and based their quantitative analysis on survey data of the Flemish government. In order to recognize different cultural consumption patterns, Vander Stichele & Laermans (2005) used Latent Class Analysis that generated six different clusters of individuals with the same characteristics. The results of their empirical study clearly indicate the existence of some groups within the Flemish cultural public that can be characterized as omnivores (Vander Stichele & Laermans, 2005).

Redirecting our focus on the CAC, it is evident that corporate art is mainly being exhibited at work spaces that are daily visited by employees (Martorella, 1990), which make employees the key consumers of the visual artworks. Furthermore, corporate collections usually consist of highly valuable art pieces that have similar features to the professional works in museums (Kottasz et al., 2007). In addition, most companies that own a CAC are imitating art museums and galleries by having their own curators and art departments in place that are exclusively dedicated to carefully organizing and displaying the art works (Wu, 2003). Within this context, the content of corporate art collections as well as the way individuals are exposed to them is very similar to the manner in which consuming art in leisure time takes place. One evident case where the professionalization of the CAC has taken on museum-like properties is the Deutsche Bank. The bank, which owns world’s largest CAC, organizes regular staff tours to educate employees on the displayed artworks, has an online art magazine

(14)

14

and frequently invites experts from renowned museums to give lectures on their collection (Kottasz et al., 2007).

From a traditional standpoint, it would be likely to expect that employees who enjoy highbrow forms of art in their leisure time would also enjoy this in their work environment, represented by the CAC. However, this direct link between consumers of highbrow forms of art and positive employee attitudes of the CAC is less probable as it may seem. Whereas lowbrow products are usually consumed by the mass of the population (Bourdieu, 1984), highbrow culture is not a very popular genre by itself and is consumed by a small portion of the population in general (DiMaggio, 1996). In addition, according to relevant academic research, the connection between highbrow art consumption and appreciation of the CAC no longer exists, as perfect snobs that only consume highbrow forms of art are now rare (Peterson and Kern, 1996; Levine 1988; Prior 2013). More specifically, the formerly perfect snobs no longer restrict their cultural participation to highbrow products, but nowadays are open lowbrow cultural activities as well (Peterson and Simkus, 1992). This means that even when people are not perfect snobs, they can still have a positive attitude towards the CAC.

In accordance with the above, this thesis expects employees that appreciate highbrow forms of art similar to the content of the CAC can be defined as individuals that participate in a diverse set of cultural activities (highbrow and lowbrow). This type of cultural consumption can be described as vertical boundary spanning of highbrow and lowbrow cultural categories (Berghmen & Van Eijck, 2009). Therefore this thesis hypothesizes the following;

Hypothesis 1: Individuals’ vertical boundary spanning of cultural activities is positively related to employee attitudes of the corporate art collection.

(15)

15

2.3. Employees’ experiences of the organization

This chapter elaborates on possible effects of employees’ experiences of the organization on their attitude of the CAC. That is, how can we best predict employees’ appreciation of the CAC based on employees’ experiences that arise from within the work environment. First, it is proposed that organizational identification is positively related to employee attitudes of the CAC. Secondly, this thesis argues that intrinsic art collection motives positively mediate the previous described relationship. Finally, status consistency is argued to positively moderate the relationship between OI and employee attitudes of the CAC.

2.3.1. Identifying with the organization

As part of the psychological relationship between the employee and organization, organizational identification has been argued to have a unique value in explaining employee attitudes in organizations (Cole & Bruch, 2006; Lee et al., 2015). The central cognitive process within organizational identification is depersonalization, which means that an individual sees his/herself as an expression of the organization’s norms and values (Hogg et al., 1995) rather than as a unique individual.

It has been argued by researchers that when employees identify more strongly with the organization, their beliefs about the organization and its activities are likely to become more positive (Ashforth and Mael, 1989). This is because employees’ identification with the organization becomes self-defining, which means their success or failure becomes interconnected to the performance of the organization’s activities (Ashforth and Humphrey, 1993). As a consequence, this cognitive process of depersonalization increases employees’ motivation to support the organization’s goals

(16)

16

and investments (Morsing, 2006). For example, employees with high levels of organizational identification are more inclined to believe that their organization is producing valuable outputs (Dutton et al., 1994). In addition, Lee (1971) found that employees with high levels of organizational identification obtain supportive attitudes towards the organizations’ activities in comparison to employees with a low level of organizational identification. Carmeli et al. (2007) argue that when employees feel more identified with their organization, they identify more easily with the values the organization proposes to communicate to its employees.

Furthermore, Jackson (2002) showed that high levels of identification lead to exaggerated positive evaluations of an individuals’ social system in comparison to the social systems of other individuals, also known as the in-group bias. The term “bias” refers to an unjustifiable response of the individual on its system’s activities, meaning that this response is more positive than the response an objective person with no link to the social system would have (Turner & Reynolds 2001). This implies that the more employees feel identified with a social system, the more favorable their evaluations of their social system’s activities become, even though objective judgement would evaluate less positively. Jackson’s (2002) findings confirm social identity theory –in which organizational identification is based– that argues once a group identity is activated, people start to behave in a way to enhance the evaluation of their organization relative to organizations they don’t belong to, thereby enhancing the evaluation of themselves as members of the organization (Turner et al. 1987). Thus, organizational identification triggers a more positive attitude towards the own organization and its activities, which can be described as a process of maintaining and enhancing individuals own of self-esteem in comparison to others (Stests & Burke, 2000).

(17)

17

With respect to an organization’s activities, corporate art collecting has increasingly become a prominent part of organizational goals through its transformation into a professionalized non-core activity. Up until the 1980s, it was usually the job of the management team to manage and decide on purchases for the CAC (Behnke, 2007). However, after the 1980s, most organizations started hiring their own curators and dedicated a specific budget to art purchases (Wu, 2003; Behnke, 2007). With their legitimate position and and network in the world of arts, curators now choose the artworks and organize their placement throughout the organization (Behnke, 2007). It can thus be argued that the CAC has been transformed into a legitimate non-core activity, similar to other well-known corporate non-core activities such as CSR activities or HRM practices that also are being managed by experienced professionals and provided with a specific budget.

The academic debate surrounding organizational identification argues that when employees identify more strongly with the organization, they are more likely to believe that the organization is producing valuable outputs (Dutton et al., 1994) and obtain supportive attitudes towards the organizations’ activities in comparison to employees with a low level of organizational identification (Lee, 1971). In addition, the more employees feel identified with an organization, the more favorable their evaluations of their organization’s activities become, even though objective judgment would evaluate less positively (Jackson, 2002; Stests & Burke, 2000). As part of an organization’s goals and output, corporate art collecting can be perceived as a professionalized non-core activity. Within this context, this thesis expects that organizational identification is positively related to employee attitudes of the CAC. Accordingly, the following is hypothesized;

(18)

18

Hypothesis 2: Employees' organizational identification is positively related to employee attitudes of the corporate art collection.

2.3.2. The increased importance of doing good

Nowadays, employees give much importance to organizations’ social responsibilities and for them to genuinely care about the wellbeing of the society they are active in (Quintelier et al., 2017). Whereas value maximization theory argues that the only responsibility of an organization’s managers is making profits, non-core activities are expected to promote social causes and reflect the organization’s ethical stance (De Roeck et al., 2014). Furthermore, it is argued that fully understanding employee attitudes of non-core activities requires identifying employees’ underlying attributions of these activities (Donia & Sirsly, 2016). With respect to a widely researched non-core activity such as CSR, Donia & Sirsly (2016) propose this means distinguishing between employees’ extrinsic and intrinsic CSR perceptions, also known as genuine or greenwashing.

Perceived organizational motives of non-core activities such as CSR or HRM practices have been studied with increased academic interest in the past decades (De Roeck et al., 2013; Story & Neves, 2015). On the other hand, perceived art collection motives (ACM) have scarcely been empirically researched. Therefore, the theoretical embedding towards the following hypothesis will take these diverse discourses with respect to other corporate non-core activities into account.

Employees are argued to be very much aware of the character of an organization’s non-core activities. Out of all the stakeholders, employees are expected to have access to the biggest share of information about their organization and therefore are most likely to carefully evaluate their organization’s motives for conducting non-core

(19)

19

activities (Lee & Seo, 2017). This is because employees live the organization’s applications and processes (Story & Neves, 2015), are embedded in the organization’s network (Rowley, 1997) and are able to evaluate non-core activities in the context of the organization’s history, culture, values and beliefs (McShane & Cunningham, 2012). In light of the above, it is argued that employees form more extensive perceptions of the organization’s motives to invest in non-core activities compared to consumers or external stakeholders, who may simply decide not to buy the product or not to invest in the organization when detecting dishonest or unpleasant motives (Story & Neves, 2015)

Two primary types of organizational motives have been identified in the relevant literature on non-core activities: an organization’s self-serving strategic extrinsic motives or an organization’s genuine intrinsic motives (Ellen et al., 2006; De Roeck & Delobbe, 2012). When employees perceive a non-core activitiy to be extrinsically motivated, they expect the organization is being strategic in its investment and treats the investment as a ‘business case’ (Vlachos et al. 2013) or invests in the non-core activity for egoistic self-serving reasons (Ellen et al., 2006). For example, by investing in a CAC an organization may be motived to enhance its reputation, enjoy tax benefits (Wu, 2003) or fulfill managements’ personal passion for contemporary art. One can speak of intrinsic motive perceptions when employees believe the organization engages in non-core activities because they genuinly care for a social cause (Vlachos et al. 2013). This perception is related to the moral aspect of the organization, in which employees trust the charitable character of the organization. For example, the organization is perceived to be motived to support the arts community (Lindenberg & Oosterlinck, 2011) or stimulate the employees’ creative and cultural development (Wu, 2003).

(20)

20

As part of the psychological relationship between the employee and organization, organizational identification has been argued to have a unique value in explaining employee attitudes in organizations (Cole & Bruch, 2006; Lee et al., 2015). The academic debate surrounding organizational identification argues that when employees identify more strongly with the organization, they are more likely to believe that the organization is producing valuable outputs (Dutton et al., 1994) and obtain supportive attitudes towards the organizations’ activities in comparison to employees with a low level of organizational identification (Lee, 1971; Jackson, 2002). Extending this line of reasoning, this thesis expects that when employees identify more strongly with the organization, it is more likely that they perceive the organization’s investments in the CAC to be intrinsically motivated.

In their paper on employee perceptions of HRM practices, Nishii et al. (2008) propose that employees’ level of organizational identification may provide a lens through which they attach meaning to corporate non-core activity motives such as HRM practices (Nishii et al., 2008). They argue that when employees lack organizational identification, they are more likely to assume that the organization must not care about them and therefore perceive the organization’s HRM practices to be extrinsically motivated. In addition, employees that highly identify with their organization are oftentimes regarded as ambassadors of their organization, supporting their organization’s activities, and are therefore more likely to perceive the organization’s non-core activities to be driven by genuine social aspirations (Morsing, 2006).

Furthermore, this thesis expects that intrinsic motive perceptions will lead to more positive employee attitudes of the CAC. As Donia & Sirsly (2016) emphasize, while some organizations may assume that investing in any social cause is always

(21)

21

better for the employee-organization relationship than doing nothing at all, organizations oftentimes overlook the outcome of negative employee attitudes when employees believe the organization to be extrinsically motivated. When employees perceive that their organizations’ CSR activities are extrinsically motivated, employees are more likely to have a less positive attitude towards the organizations’ CSR program (Rupp et al., 2006). Within the same line of reasoning, if employees perceive that an organization is motivated to fulfill its social and ethical responsibilities through investing in non-core activities, employees are expected to respond more positively towards these activities (Lee & Seo, 2017).

Results of the study conducted by Nishii et al. (2008) provide further support for the above-described notions. Employee perceptions of HRM practices to be driven by the organization’s concern for improving service quality and employee well-being was positively related to employee attitudes of the HRM department. On the other hand, employees’ perceptions of managements’ HRM practices to be driven by reducing costs or exploiting employees was negatively associated with employee attitudes. It is interesting to mention that Nishii et al. (2008) conclude employee perceptions to be of great importance for achieving desired employee attitudes, as they find that the same set of HRM practices may exhibit different effects within a single organization.

On the one hand, relevant literature on OI suggests employees’ level of organizational identification may provide a lens through which they attach meaning to corporate non-core activity motives (Nishii et al., 2008). Employees that highly identify with their organization are more likely to perceive the organization’s non-core activities to be intrinsically motivated (Morsing, 2006). This means for example that employees believe that the organization is motivated to support the arts community (Lindenberg

(22)

22

& Oosterlinck, 2011) or to enhance the working environment of the employees (Wu, 2003). Furthermore, it stands to reason to argue that perceived intrinsic art collection motives lead to more positive employee attitudes of the CAC. If employees perceive that an organization is motivated to fulfill its social and ethical responsibilities through investing in non-core activities, employees are expected to respond more positively towards these activities (Nishii et al., 2008; Lee & Seo, 2017). As a result, this thesis posits that employees’ interpretations of the motives (intrinsic or extrinsic), which they attribute to their organization’s CAC initiatives, play an important mediating role in the relationship between OI and employee attitudes of the CAC. Within this context, the following is hypothesized;

Hypothesis 3: Perceived intrinsic art collection motives positively mediate the positive relationship between employees’ organizational identification and employee attitudes of the corporate art collection.

2.3.3. Status consistency as objective experience of the organization

Thus far, this thesis has considered subjective employee experiences of the organization that are likely to influence attitudes of the CAC. For a coherent research framework, this thesis will also take the measure of status consistency into account. As objective reference, status consistency occurs when an individual's ranks on different status hierarchies are consistent with one another (Lenski, 1954). Status hierarchies are most commonly referred to as socioeconomic status (SES) indicators that are extensively used in social science research (Duncan & Magnuson, 2003). Occupational prestige, educational attainment, and income can be considered as the “triumvirate” indicators of SES (Duncan & Magnuson, 2003).

(23)

23

Typical examples of status consistency and inconsistency are respectively a CEO with an MBA or a well-paid commercial manager with only a high school diploma. Status inconsistency is argued to lead to uncomfortable conditions for employees (Bacharach et al., 1993). Social science has examined the effect of status consistency and inconsistency on various outcomes for decades, such as social behaviors and psychological symptoms like role stress (Honjo et al., 2014). Employees may find it difficult to attain and maintain their position within the organization when facing lower educational background in comparison to colleagues (House and Harkins, 1975) or employees are confronted with a need for social mobility (Prior, 2013). Little research has however been conducted to examine the effect of status consistency on the relationship between OI and employee attitudes of an organization’s non-core activities.

As previously discussed, as part of the psychological relationship between the employee and organization, organizational identification has been argued to have a unique value in explaining employee attitudes in organizations (Cole & Bruch, 2006; Lee et al., 2015). Within the psychological relationship between the employee and organization, status consistency or inconsistency is perceived as a condition for the employee to deal with (Bacharach et al. 1993). The objective measure of status consistency or inconsistency can however not be a direct outcome of the subjective reference of organizational identification, but is a contextual condition that is likely to affect the strength of the relationship between OI and employee attitudes of the CAC (Kamdar and Van Dyne, 2007). This is because the status consistent employee is expected to feel no need to change his/her situation in the organization because all individual's ranks on different status hierarchies are consistent with one another (Lenski, 1954). Status consistency is thus expected to positively moderate the

(24)

24

relationship between OI and employee attitudes of the CAC. In accordance, the following hypothesis is formulated;

Hypothesis 4: Individuals' status consistency positively moderates the positive relationship between employees’ organizational identification and employee attitudes of the corporate art collection; with the effect becoming stronger when status consistency increases.

(25)

25

3. Methodology

By reviewing the relevant literature on individuals’ cultural interests outside and employees’ experiences inside of the organization, this thesis has been able to hypothesize four effects. The conceptual model below visualizes these hypothesized relationships.

Figure 1: Conceptual model.

This chapter will elaborate on the empirical setting and all relevant measures and variables. First the research setting will be presented, where after sampling procedures will be discussed. Consequently, all variables (dependent, independent, mediating moderating, control) and their measurement levels will be explained.

Organizational identification

Perceived intrinsic art collection motives

Employee attitudes of the CAC

Status consistency Vertical boundary spanning of

cultural activities

Employee attitudes of the CAC

H1+

H2+

H4+

(26)

26

3.1. Empirical setting

Organizations owning an art collection based in the Netherlands constitute to the research setting of this thesis. On a global level, companies have originally started collecting art around the midst of the twentieth century (Martorella, 1990; Wu, 2003). This trend, which has become increasingly popular in the last decennia, has developed at relatively the same pace in the Netherlands. However, the number of art works in Dutch corporate art collections varies greatly, and ranges from a couple of 100 of KPMG to around 15.000 owned by ING (Sterk & Elferink, 2015). It is estimated that the totality of the Dutch art corporate collection amounts to 100.000 to 200.000 pieces of art (RTL Z, 2015). The oldest Dutch corporate art collection has been founded in 1960 by Turmac Tobacco Company (Beeren, 1992). Because of the maturity and broad range of CACs in the Netherlands, this research setting allowed for the construction of a relatively large population size.

3.2. Sampling and data collection

The data collection for this study was conducted among a group of nine organizations and yielded data from almost 3.500 employees that work at organizations owning a CAC. Three organizations from the sample were multinationals, whereas the other organization solely operated in the Netherlands. The population consists of employees in the Netherlands. In line with earlier relevant research of Martorella (1990), this research only included organizations that showed serious art collecting investments instead of casual art buying. The organizations of the sample operate in the fields of finance, real estate, healthcare, consultancy or the public sector. Especially finance and real estate have been identified as leading industries that drive the art market with

(27)

27

their art collecting activities (Sussman, 2017). With respect to privacy matters, the organizations’ names and details of employees will not be displayed in this thesis.

3.3 Data collection

Convenience sampling was used as a technique to collect the data. Accordingly, the survey was distributed via the employees’ professional e-mail address, therefore making use the existing connection structure of each organization. The employees were either contacted on their direct professional email address or through a message via Intranet. Intranet is a private emailing network that is accommodated within organizations in order to professionalize and fasten the internal network capacity. Employees of organizations that use Intranet frequently log onto this platform in order to access their email and/or to read corporate announcements. Through a link in the messages, the employees were redirected to the online survey where they were anonymously asked to answer each question. The survey invitation was sent to employees based in offices in the Netherlands only, and therefore was only made available in the Dutch language. Beforehand, it was not exactly clear what response rate to expect, since previous studies on CACs with a similar research design and contextual framework are scarce. The same methods were used for all participating organizations. Responses and response rates per organization are presented in table 1.

(28)

28

Table 1 Responses and response rate per organization

Organization Responses (N) Response rate (%)

1 708 7.87% 2 248 9.90% 3 1381 16.23* 4 188 5.60% 5 107 35.67% 6 165 2.75% 7 188 4.18% 8 116 1.66% 9 375 6.94% Total 3476

As depicted in the table above, the response rate differed quite significantly per organization. The survey generated a relatively low response rate. With a total participation number of 3476, the response rate of the sample amounted to a percentage of 7,44%. Furthermore, there were a lot of missing values, varying per variable from 390 to 3232. The survey is presented in Appendix 8.1, containing all measures that have been used in this thesis.

3.4. Variables and Measurements

All of the measures used in this study were drawn from relevant existing literature and adapted to the context of the current study. Three measurements used a 5-point Likert-type scale ranging from 1 (Strongly Disagree) to 5 (Strongly Agree). Two measures made use of ordinal scales in order to classify according to rank. One measure used a ‘yes-no’ response scale, and another measure allowed respondents to manually fill in their job title. Only a part of the collected data was used in the data analyses for this thesis. As such, only the sections relevant for this thesis are discussed below and are presented in the appendix.

(29)

29

Dependent variable: Employee attitudes of the CAC

This measure was established in order to examine the degree that an employee values the organization’s investments in corporate art collecting. Five questions were set up to measure employee attitudes of the CAC, with a 5-point Likert scale (1 = Strongly agree; 5 = Strongly disagree). These questions were developed based on prominent themes identified in the literature on CACs (Kottasz et al., 2007). For example, the participants were asked if they believe the art collection to be valuable for the organization or if they believe the organization should invest more money in the CAC. A scale mean for this variable was computed to facilitate the hypotheses testing.

Independent variable: Vertical boundary spanning of cultural acitivites

Employees’ level of cultural participation was measured by using the scale originally developed by van Vander Stichele & Laermans (2006), with 9 variables (ranging from visiting a visual art museum to a popular cinema visit in the last 12 months). The original 5-point Likert response scale ranging from ‘never’ to ‘several times a month’ was adjusted into a ‘yes-no’ response scale. For each of the nine variables, the respondents were asked if they had participated in the cultural activity during the previous twelve months.

The nine items relate to the attendance of visual art museums, history museums, ballet/modern dance performances, opera/classical music performances, arthouse cinema, cabaret/theatre plays, musical performances, popular music performances and popular cinema performances visits. This thesis assumes, not without some hesitation, that the five cultural activities that were mentioned first can be categorized as highbrow cultural activities, whereas the other four cultural

(30)

30

activities are categorized as lowbrow (Berghman & Van Eijck, 2009; Toepoel, 2011). Especially the category that combines theatre and cabaret plays is difficult to categorize, since cabaret is usually categorized as lowbrow and theatre as highbrow (Toepoel, 2011).

Furthermore, Rubin (1992) and Peterson and Kern (1996) have suggested that it is possible to distinguish middlebrow as an additional cultural category. This is because academics recognize an intermediate level of cultural taste between lowbrow and highbrow, and that omnivores are aware of this intermediate category (Rubin, 1992; Peterson and Kern, 1996). Unfortunately this thesis was not able to make this detailed differentiation because the diversity between the cultural activities does not seem high enough to make three categories.

This thesis expects employees that appreciate highbrow forms of art similar to the content of the CAC can be defined as consumers that participate in a diverse set of cultural activities (highbrow and lowbrow). Therefore, and in line with relevant previous research (Berghman & Van Eijck, 2009), this thesis calculated a measure for vertical boundary crossing by counting the number of pairs of highbrow and lowbrow cultural activities that each individual consumed in the last 12 months. By doing so, this operationalization of omnivorousness takes into account both amount of activities and diversity (Berghman & Van Eijck, 2009). For example, a person who consumed two highbrow activities and three lowbrow activities is assigned with a score of two.

Independent variable: Organizational identification

To measure organizational identification a 6-item version with a 5-point Likert scale (1 = Strongly agree; 5 = Strongly disagree) of Mael & Ashforth’s (1992) measure was used. Example of items included: I am very interested in what others think about my

(31)

31

organization”, “When somebody criticizes my organization, it feels like a personal insult,” and “When I talk about my organization, I usually say “we” rather than “they”. A scale means for this variable was computed to facilitate the hypotheses testing.

Mediating variable: perception of art collection motives

Perception of art collection motives was measured based on ten prominent motives identified in the literature, as summarized in table 1 by Kottasz et al. (2007), with a 5-point Likert scale (1 = Strongly agree; 5 = Strongly disagree). With a 7-item measure, participants were for example asked to indicate how much they perceive that the organization’s CAC practices are set up to generate financial benefits or function as a tool to enhance the employees’ work environment.

Hypothesis three of this thesis described the proposed positive mediating effect of perceived intrinsic art collection motives on the positive relationship between OI and employee attitudes of the CAC. According to the relevant literature on non-core organizational motives, the seven questions were expected to divide over an intrinsic and extrinsic factor. With respect to extrinsic motives, the focus lied on questions that revealed egoistic or strategic motivates (Ellen et. al, 2006; Du et al., 2010; Vlachos et al. 2013). This meant that organizations were driven by egoistic self-serving motives or saw the investments in the CAC purely as a ‘business case’. With respect to intrinsic drivers, this thesis followed earlier studies (Weaver et al., 1999; Kottasz et al., 2007; Vlachos et al. 2013) and adopted their focus on acting out of a genuine concern for a social cause. This perception is related to the moral aspect of the organization, in which employees trust the charitable character of the organization.

(32)

32

In accordance, item one (motive for financial investment), item three (motive to fulfill managements’ passion for art) and item seven (motive for reputation/image benefits) were expected to fall in one factor characterized by extrinsic motives. Item two (motive to support the arts community), item four (motive to inspire employees and stimulate their creative development) and item five (motive to enhance employees’ work environment) were expected to fall in one factor characterized by intrinsic motives. Due to ambiguity of item six (creating a company culture/identity) with respect to intrinsic or extrinsic motives, this item was not taken into account in the analysis that follows. This is because a corporate identity is argued to be the fusion of all organizational activities such as its strategy and its internal culture, and is therefore of influence on external stakeholders as well as internally in relation to employees (Kottas et al., 2007).

To test if extrinsic and intrinsic motives could be determined by two different constructs, the discriminant validity between the different motives first needed to be investigated. To examine this, and in accordance with relevant literature (Story & Neves, 2015), an exploratory factor analysis (EFA) using an oblimin rotation method on the included six items of art collection motives was used. Researchers commonly use this method in order to examine the underlying construct(s) of measures (Fabrigar & Wegener, 2011). Appendix 8.2 visualizes all relevant tables resulting from the following EFA using an oblimin rotation method.

With a Kaiser-Meyer-Olkin score of .697 it was evident that the sample size was adequate, because KMO should at least be greater than .5. Consequently, the Bartlett’s Test of Sphericity is significant which indicates that there is least one significant correlation between two of the items. There has to be at least one correlation between the items to be able to run an EFA. In addition, the

(33)

33

Communalities table showed no extraction values that were lower than .3, which indicated a sufficient strength of extraction for each item under the corresponding factor. From the Total Variance Explained table it became evident that, as expected, there should be two new factors. This means that all items (except item number six) could be divided into two new factors that have eigenvalues greater than one, explaining over 57% of the variance.

With the knowledge of having two new factors, an oblimin rotation analysis was performed because it was assumed that there was going to be a mild correlation between the new factors. After looking at the correlation between the two new factors in this matrix, it became evident that the two new factors were not strongly correlated (.219), which means they are orthogonal (Vogt, 1993). Therefore orthogonal rotation was used to further analyze the two factors. By computing a rotated component matrix, we were able to identify which items (questions) of the measure fall under which factor. From this it became evident that items one (extrinsic), two (intrinsic), three (extrinsic), four (intrinsic) and five (intrinsic) were factorized as expected. In contrast to what was hypothesized, the item that covered the motive for retrieving reputational/image benefits was factorized with the intrinsic items. This thesis decided to use the factor with the most intrinsic items in the following regression analysis, because apart from item seven, this factor is 75% similar to what was hypothesized. A reliability test this factor resulted in a Cronbach’s alpha of .695. A scale means for this factor was computed to facilitate the hypotheses testing.

Moderating variable: status consistency

This thesis used individual measures of social status to define status consistency. Three commonly used proxies for socioeconomic status were operationalized: years

(34)

34

of schooling, level of household income and employment status (Duncan & Magnuson, 2003; Mayer & Osterle, 2014). Highest level of educational attainment was measured on a 6-item scale: (1) primary education (basisonderwijs); (2) high school (mavo/vmbo, havo, vwo); (3) lower secondary education (mbo, meao, mts); (3) higher secondary education (hbo, heao, hts); (5) University; (6) Ph.D. Each item was recoded into years of schooling. This method transformed education as a categorical variable into a continuous variable, ultimately leading to a more convenient interpretation of the final regression results. Similar to the recoding scheme in the research of Van Eijck & de Graaf (2004), primary education was attained in six years of schooling, high school in ten years of schooling, lower secondary education in 12 years of schooling, higher secondary education in 15 years of schooling, University in 17 years of schooling and Ph.D. in 20 years of schooling.

Secondly, as a link to financial barriers, level of household income was included approximated by the monthly net equivalent household income in Euro and summarized into the following 6-item scale: 1499 or less; 1.500-2.999, 3.000-4.499, 4.500-5.999, 6.000-7.499; 7.500 or more.

Finally this thesis used employment status as a status indicator. First of all, to ensure maximum accuracy, occupations were coded according to the International Standard Classification of Occupations (ISCO-08). The denoted job titles of all respondents were converted into ISCO codes, see table 7 in Appendix 8.3. Thereafter, the recoding of these ISCO codes to a 6-item employment prestige score was done using the ISEC classification, see table 8 in Appendix 8.3. ISEC-08 stands for the International Socio-Economic Classes 2008 and is a categorization of occupational information coded in the ISCO-08 and provides information on employment statuses divided into a 13 categories (Ganzeboom, n.d.). With little respondents in ISCO-08

(35)

35

major groups seven, eight and nine, these were merged into the classification of “semi- and unskilled manual workers”. The final classification is presented in table 9 (Appendix 8.3).

Having obtained comparable scores for the three status hierarchies, the last step was to establish the quantitative measure of status consistency. Similar to the research of Biron et. al (2016), this thesis calculated difference in scores between the three status indicators by regressing the income scores on the education and employment status scores and subsequently saving the standardized residual scores (Williams et al., 1987). Positive and negative values indicated status inconsistency. The value of zero indicated status consistency.

3.5. Control variables

Various control variables were included in the regression analysis of this thesis for ruling out other possible explanations for the significant relationships. Riketta (2005) has shown that variables like age, gender and tenure can influence an individual’s level of organizational identification. In addition, a number of variables have been shown to influence cultural participation and have the ability to influence the degree of omnivorousness (Peterson & Kern, 1996). These variables include education, gender and household income (DiMaggio and Useem 1978). Finally, controlling for gender, age, and tenure also makes sense for status consistency, as these variables have been identified as potentially related to absenteeism, which correlates with status consistency (Price, 1995). Finally, for ruling out all other possible explanations for the significant relationships, also contract type & hours, employment status, dummy variable for relationship status (“single” as reference category, with the other two categories being “in a relationship” and “married”), children and being in a

(36)

36

management position have been included in the regression analysis as control variables.

(37)

37

4. Results

This chapter discusses the results of the study. First descriptive statistics and the correlations are elaborated on, after which the results of the regression analysis will be presented and discussed.

4.1. Descriptive statistics

In preparation of the dataset, the integrity of the data was checked in order to inspect for errors in the data. Consequently, a frequency test was run for all variables, and showed that were quite some missing values in the dataset. These missing values were caused by respondents not completing the survey or respondents not willing to share information on certain topics. Listwise deletion was used to deal with the missing data, which entailed the process of eliminating any respondent that is missing a score on the variables that are the focus of this thesis. According to the content analysis of Harel et al. (2008) this is a common manner for a lot of researchers in dealing with missing values. All subsequent analyses are based on 300 respondents who had valid scores on all variables.

Next, normality checks were performed to examine if all the variables were normally distributed. These checks indicated that all the variables except for contract type and employment status were normally distributed, meaning that skewness and kurtosis are between -2 and +2 (George and Mallery, 2016). Contract type had a skewness of -2.351 and a kurtosis value of 3.548, which indicates the distribution to be negatively skewed and pointy with more scores in the tails. Employment status had a skewness between -2 and +2, but a kurtosis value of 2.830, indicating the distribution to be pointy with more scores in the tails.

(38)

38

In addition, an internal consistency reliability analysis through Cronbach’s alpha tested the reliability of the measurements. The most frequently cited criterion to determine a good level of reliability is an Cronbach’s alpha larger than .7 (Brown, 2002). All variables in this thesis, except intrinsic ACM, that were based on multiple-item Likert questions have a Cronbach’s alpha larger than .7. It being slightly less than .7, with a Cronbach’s alpha of .695, an acceptable level of reliability is achieved for intrinsic ACM. Table 2 on the next page presents the descriptive statistics and correlations of the variables that were used in this thesis.

(39)

(40)

The degree of organizational identification and the extent to which the respondents value the organization’s investments in the CAC showed to be quite positive with respective means of 3.576 and 3.6. The respondents of the sample showed not to be very active boundary spanners of cultural activities with a mean of 1.6 reflecting the average number of pairs of highbrow – lowbrow cultural activities. Furthermore, with a mean of 3.82, the respondents showed to believe the art collection motives to be rather intrinsic. Status consistency showed to be an average condition with a mean of .000.

The total sample consisted of 138 females and 162 men, with a mean age of 46.12 and average tenure of 15.80. Taking into account the industries that the organizations are active in, and the general aging of Western societies (North & Fiske, 2015), the relatively high mean age and tenure level is not surprising. Furthermore, highly educated individuals dominated the sample. Namely, 129 respondents had a University degree, and 96 had followed a higher secondary education. The respondents also scored high on employment status; with 169 enjoying the second highest level of employment status and 46 the highest level as being higher-level professionals. Household income was normally distributed, 87% of the respondents was in a relationship (31.3%) or married (55.7%) and around half of the respondents were living with their children aged under 18.

Table 2 also clarifies the correlations between all the variables used in this thesis. It can be observed that the independent variables vertical boundary spanning of cultural activities and organizational identification have a significant correlation with the dependent variable attitudes of the CAC (r = .183, p < .01; r = .223, p < .01). The same goes for mediating variable intrinsic ACM (r = .610, p < .01). The moderator variable status consistency is not significantly correlated with the dependent variable.

(41)

41

Furthermore, the correlations between the dependent, moderator and mediating variable are also presented in table 2. Intrinsic ACM is significantly correlated to status consistency (r = .118, p < .05). This could suggest that the more an individual is conditioned with the situation of status consistency, the higher the chance that this individual perceives the art collection motives of the organization to be intrinsic. The dependent variables OI and vertical boundary spanning of cultural activities in turn are significantly correlated to mediating variable intrinsic ACM (r = .236, p < .01; r = .140, p < .05). This firstly indicates that OI is positively related to intrinsic ACM, which is part of hypothesis 3. In additon, the latter correlation may suggest the more individuals are active boundary spanners of cultural activities, the more they perceive the motives of the organization to invest in a CAC to be intrinsic.

(42)

42

4.2. Regression analysis

Table 3 presents the regression results that are computed by using ordinary least squares (OLS) regression and Hayes Process Model 4 for mediation and Hayes Process Model 1 for moderation.

(43)

43

Model 1 in table 3 reflects a control model, that investigated the possibility of a direct relationship between the control variables and the dependent variable (Model 1: R2₌

.092, p < .05). It became evident that age is the only control variable that has a positive significant effect on employee attitudes of the CAC (β = .021, p < .05). This thus suggests that the older people get, the more they appreciate the CAC. This is not surprising, as youngsters and younger adults tend to see highbrow cultural activities such as visual art museums as boring or unapproachable (Mokhtar & Kasim, 2011).

The independent variable vertical boundary spanning of cultural activities was added in Model 2, therefore allowing the testing of hypothesis 1 (Model 2: R2 = .062, p < .01). Hypothesis 1 proposed individuals’ vertical boundary spanning of cultural activities to be positively related to employee attitudes of the CAC. As the model indicates, vertical boundary spanning of cultural activities has a positive significant on employee attitudes of the CAC, while controlling for twelve employee characteristics (β = .115, p < .05).

Model 3 reflected the testing of hypothesis 2 that proposed a positive relationship between OI and employee attitudes of the CAC (Model 3: R2 = .130, p < .001). As hypothesized, the model shows a significant positive relationship between OI and employee attitudes of the CAC (β = .263, p < .01).

Moving forward, hypothesis three was tested in model 4 (Model 4: R2 = .093, p< .01). This hypothesis proposed perceived intrinsic art collection motives to positively mediate the relationship between OI and employee attitudes of the CAC. The indirect effect is positive and statistically different from zero (β = .170, BCa95= [.068, .279]), thus providing support for a mediating effect of perceived intrinsic ACM. In addition, as the direct relationship between OI and employee attitudes of the CAC is not significant in model 4, full mediation has occurred.

(44)

44

Finally, Model 5 was used to test hypothesis 4; individuals' status consistency positively moderates the positive relationship between OI and employee attitudes of the CAC, with the effect becoming stronger when status consistency increases (Model 5: R2 = .128, p< .001). The latter three control variables were not included in this model as the moderator variable itself is based on these status indicators. The interaction term, which entails information on how much the effect of OI on employee attitudes of the CAC is different between status consistent and inconsistent persons, is insignificant with a p value of 0.276. This means the effect of OI on employee attitudes of the CAC does not depend on the level of status consistency of individuals. The table below summarizes the results discussed in this chapter. There after a visualization of the conceptual model is presented which depicts the path coefficients for all tested relationships.

Table 4 Summary of hypotheses results

Hypothesis Analysis Relationship Result

1 OLS Regression Positive relationship between vertical boundary spanning

of cultural activities and the CAC attitude

Accepted

2 OLS Regression Positive relationship between OI and the CAC attitude Accepted

3 Hayes Process Model 4 The positive mediation effect of intrinsic art collection

motives on the positive relationship between OI and the CAC attitude

Accepted

4 Hayes Process Model 1 The positive moderation effect of status consistency on the

positive relationship between OI and the CAC attitude

Explaining corporate art collecting attitudes of employees : the role of employees’ cultural interests and experiences of the organization