Final Version Master Thesis
Name: Max Dobbe
Student Number: 11152249
First Supervisor: Eloisa Federici Second Supervisor: Corine Boon Date of Submission: 17-08-2018
Programme: MSc. Business Administration
Specialisation: Leadership and Management
Institution: Amsterdam Business School
University of Amsterdam
The role of perceived HR practices in predicting organizational commitment and job satisfaction in the gig economy, mediated by the perceived organizational support of gig
1 Statement of originality
This document is written by Student Max Dobbe who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is 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.
2 Abstract
The present study examines the role of perceived HR practices in predicting organizational commitment and job satisfaction with a mediating effect of perceived organizational support of gig workers. The study explores these relationships in a new phenomenon referred to as the gig economy. The gig economy consists of a growing group of workers that perform short-term tasks called gigs, and no longer have a long-short-term employment with an organization. The study builds on current literature of Human Resource practices and organizational support to investigate the underlying mechanisms of organizations in the gig economy. A sample of 181 gig workers from the Netherlands, United Kingdom, and the United States filled in a
questionnaire. The study found that perceived HR practices had a significant direct effect on organizational commitment. In addition, perceived organizational support partly mediates the relation between perceived HR practices and organizational commitment, Furthermore, perceived organizational support fully mediates the relationship between perceived HR practices and job satisfaction. Implications and future research avenues are discussed.
Key words: gig economy, perceived HR practices, perceived organizational support,
3
Table of Content
1. Introduction ... 4
2. Literature review ... 9
2.1 Gig economy ... 9
2.2 Human resource management (HRM) ... 12
2.2.1 HR practices ... 13
2.2.2 HRM outcomes ... 14
2.2.3 Composition of HR practices in the gig economy ... 15
2.2.4 Perceptions of HR ... 16
2.2.5 HR in the gig economy ... 17
2.3 HR and organizational commitment ... 18
2.3.1 Social Exchange Theory (SET) ... 19
2.4 HR and organizational commitment and the mediating role of POS ... 20
2.5 HR and job satisfaction ... 21
2.6 HR and job satisfaction and the mediating role of POS ... 23
3. Method ... 24
3.1 Sample & Procedure ... 24
3.2 Analytical strategy ... 25
3.3 Measures ... 26
4. Results ... 28
4.1 Results of the correlation matrix ... 28
4.2 Results of Process Model 4... 29
5. Discussion ... 33
5.1 Practical implications ... 35
5.2 Limitations and suggestions for future research ... 36
6. Conclusion ... 38
Bibliography ... 39
Appendix ... 51
Survey Platform Economy ... 51
4 1. Introduction
Stephanie Leffer, CEO of OneSpace believes the gig economy is a new composition of how people are making money and structuring their employment (Reus, 2015). The gig economy consists of a growing group of gig workers who no longer have the so-called long term-employment with an organization, but are paid for short tasks, assignments or sales, called gigs, that are defined by a high degree of autonomy and flexibility (De Stefano, 2015; Kuhn, 2016). Digital platforms are defining the gig economy by transforming the way of work,using digital innovation to reach enormous pools of workers and customers with lightning speed (Manyika e.a., 2016). A large study, both within in the United States and Europe, has shown that today these digital platforms are used by 15 percent of independent workers (Manyika e.a., 2016). Forms of work in the gig economy consist of “work-of-demand via apps” or “crowd work”, where the demand or supply of work is matched online or via mobile apps (De Stefano, 2015).
The rapidly accelerating growth of the gig economy represents one of the most significant and difficult challenges for Human Resources departments by redefining the way work is viewed (Horney, 2016). In addition, as platforms become more ubiquitous, there is a severe lack in research regarding the impact of this rapid growth on gig workers and the way HR processes of platforms are interacting with gig workers (Horney, 2016; Kuhn, 2016). Therefore, this study will focus on HR in the context of the gig economy.
HRM is playing an important role in matching people with jobs (Boon, Den Hartog, Boselie, & Paauwe, 2011) and improving organizational performance (Huselid, 1995). Traditional HRM studies focused on how individual HR practices affect organizational outcomes, while recent research focused more on the bundles of HR practices as a system and their combined effect on organizational outcomes (Wright & Boswell, 2002). This study focuses on HR practices as a system to better represent the platforms’ context and argues the relationship between gig workers and platforms is contingent on an HR system rather than a single HR practice. According to MacDuffie (1995), HRM systems resemble an organization’ culture and strategy. HRM systems incorporate a set of interconnected high performance practices such as selection, recruitment, training and development, job design, performance management, performance-based pay, and participation in decision making that are expected to lead to improved organizational performance (Jiang, Lepak, Hu, & Baer, 2012; Kooij & Boon, 2018). This study focuses on the bundle of HR practices involving selection, training, job design, and participation. These practices have been selected for multiple reasons. First, the competition for talent in the gig economy resulted in a focus on skill development and the
5
phasing out of career paths (Best, 2017). Secondly, the nature of work in the gig economy is fundamentally different compared to traditional jobs (e.g. more autonomy and flexibility) (Donovan, Bradley, & Shimabukuru, 2016). Thirdly, the lack of social interaction between platforms and gig workers raises concerns for the quality of communication between
platforms and gig workers (Huws, Spencer, Syrdal, & Holts, 2017). Communication plays a key role in the way managers and workers are perceiving HRM (Den Hartog, Boon, Verburg, & Croon, 2013). Therefore, this study proposes that these practices play a significant role in the gig economy and will represent the HR systems of platforms.
Research has shown that HR practices as a system and organizational performance have a significant relationship (Huselid, 1995; MacDuffie, 1995). Huselid (1995) found that HR as a system had a significant relation with turnover, productivity, and financial
performance. Kwon, Bae, & Lawler (2010) found a significant relationship between HR practices as a system and organizational commitment. In addition, Ileana Petrescu &
Simmons (2008) found a significant relationship between HR practices and job satisfaction. However research between HR and organizational performance suffers from an issue called the “black box”, which represents that studies are unable to link HRM and the inner workings of HR processes in organizations to organizational performance (Boselie, Dietz, & Boon, 2005; Boxall, Ang, & Bartram, 2011). Furthermore, another issue that HRM research faces have to do with the perceptions of HR practices. Boxall et al. (2011) argue that it is a fallacy to expect that all employees in an organization are managed the same way, which is an issue for surveys that ask HR managers to report on the practices of their organization. Therefore, researchers have started to focus on employee “perceptions” of HRM practices that may differ from manager’ intentions in implementing set practices (Kooij & Boon, 2018). Furthermore, employee perceptions are important because they are the ones expected to experience the intended HR practices (Ostroff & Bowen, 2016; Whitener, 2001). Thus, this study looks at how gig workers perceive the HR practices of platforms.
The gig economy is also subject of another discussion that revolves around the employment status of gig workers. Platforms claim that gig workers are fully self-employed, but other researchers claim that platforms have significant control over gig workers’
compensation and working environment (Friedman, 2014; Kuhn, 2016). While this debate is continuous, platform work can provide autonomy and flexibility, allowing people who may otherwise find it difficult to participate in certain types of employment, thus certain people might feel more committed to platforms than traditional organizations (Eurofound, 2018).
6
Organizational commitment is often defined as the extent to which an employee can identify himself with and is involved in the organization (Curry, Wakefield, Price, & Mueller, 1986). HR practices such as training will induce a bond between the organization and the worker, resulting in higher levels of organizational commitment (Klein & Weaver, 2000). HR practices of platforms in the gig economy have changed from a traditional full-time employee philosophy to a more talent focused view (Horney, 2016). Digital innovations create the opportunity organizations to have a flexible workforce which requires fewer resources and thus organizations are less likely to invest a lot of in HR practices, however, platforms can still benefit from committed gig workers (Friedman, 2014). As a result, the traditional
employment benefits for both sides are less relevant in the gig economy, because gig workers tend to have a short-term relationship with their platform (Manyika e.a., 2016). Untying this underlying foundation for gig workers and platform alike, this study will examine the effect of this change on gig workers. With the rising number of people that are working in the gig economy, it is now more than ever essential to look at how HR practices as system affect gig workers’ organizational commitment.
Perceived HR practices that signal investment in employees and recognize the needs of employers contribute to the development of perceived organizational support (POS) (Allen, Shore, & Griffeth, 2003). POS has been studied intensively through the years and has been associated with several variables such as HR practices, job satisfaction, and organizational commitment. (Allen & Shanock, 2013; Allen e.a., 2003; Eisenberger, Huntington, Hutchison, & Sowa, 1986; Whitener, 2001). Whitener (2001) looked at the mediating role of HR
practices on the relationship between POS and organizational commitment, while Allen et al. (2003) looked at the mediating role of POS on individual HR practices and organizational commitment. Allen et al. (2003) found that HR practices are significantly and positively related to POS. This implies that if HR practices are seen as supportive by workers this leads to an affective attachment to the organization (Allen e.a., 2003). Furthermore, HR practices that imply that the organization values and cares about employee contributions are more likely to lead to and establish a social exchange relationship with their workers (Allen e.a., 2003).
The social exchange theory (SET) and the motivational processes it entails combined with the norm of reciprocity (Blau, 1964; Gouldner, 1960) may explain the relationships among human resource practices, POS and employee commitment. Building upon the social exchange theory, this study argues that perceived feelings of organizational support mediate the relationship between perceived HRM practices and organizational commitment.
7
at HR practices as a system compared to individual HR practices. Secondly, Allen et al. (2003) included fairness, growth, and participation as individual HR practices, while this study looks at training, selection, job design and participation as HR practices in the context of the gig economy.
At this time, there is no consensus about the satisfaction of gig workers. Are gig workers satisfied with their current way of working or are they dissatisfied? The current research on gig economy is split on this debate. On one hand, some studies have shown that gig workers are more satisfied compared to traditional work arrangements (Manyika e.a., 2016), while on the other hand researchers and studies show cause for concern regarding the limited contractual protection, isolation in work and lack of social contact of gig workers (De Stefano, 2015; Friedman, 2014). While gig workers typically value the flexibility and
autonomy of crowd work, there are complaints about many aspects of the job (Huws e.a., 2017). This resulted in stress and grievance included difficulty in communicating with platform personnel and arbitrary terminations. Therefore it is not a stretch to imagine this negatively impacts job satisfaction. High degrees of autonomy, however, can also lead to higher job satisfaction (Batt, 2002). Building upon the signaling theory (Spence, 1978), this study argues that platforms can send positive signals via HR practices to gig workers by focusing on positive organizational attributes such as a high degree of autonomy and increase gig workers levels of job satisfaction.
Some researchers suggest that the relationship between perceived HR practices and job satisfaction is mediated by POS (Eisenberger, Cummings, Armeli, & Lynch, 1997). The feelings of perceived organizational support by gig workers increase the attachment to the organization, resulting in higher levels job satisfaction. Eisenberger et al. (1997) provide evidence in their research regarding job satisfaction and POS as distinctive concepts. This study is proposing that based on SET (Blau, 1964), that if gig workers perceive organizational support as genuine, this will lead to a mediating effect of POS on the relationship between perceived HR practices and organizational commitment.
Concluding, this study examines the relationship between perceived HR practices and job satisfaction and organizational commitment with a mediating effect of POS. This study has several aims. First, it explores and more importantly expands the knowledge on the new phenomena called the gig economy by investigating how gig workers perceive HR practices. Building on SET, the second aim of this study is to investigate the relationship between HR practices and gig workers’ commitment. Building on the signaling theory, the third aim of this study is to investigate the relationship between HR practices and job satisfaction.
8
Furthermore, based on SET, this study looks at the mediating role of perceived organizational support on job satisfaction and organizational commitment. In addition, this study explores possible new antecedents of perceived organizational support.
This study begins with a review of the relevant literature on the different concepts in the gig economy, their relationships, and supportive theories. Subsequently, a description of the research methods is presented and followed by the results of this study. Finally, in the discussion section the theoretical and practical implications, the limitations, future research avenues and the conclusions are presented.
The study’s research model is as follows.
Figure 1: Conceptual model
Perceived HR practices Perceived organizational support Job satisfaction Organizational commitment H2a + H3 + H4b + H2b + H1 + H4 H2
9 2. Literature review
This section consists of the most relevant findings regarding the gig economy, Human Resource practices, perceived organizational support, organizational commitment and job satisfaction. The hypotheses between the variables in this research will also be stated.
2.1 Gig economy
A full-time job with one employer has been considered the norm for decades, but in recent years, this is no longer an adequate norm to capture how a large share of the workforce makes a living (Manyika e.a., 2016). New advances in digital technologies and the rise of new employment methods will lead to associated organizational change (Eurofound, 2018). Furthermore, digitalisation and platforms are likely to extend to more sectors of the economy, resulting in more contradictions (Eurofound, 2018). Therefore, we stand at this historical conjunction, whereby more information is needed to understand how the digital revolution affects the nature of economic activity, work and employment The economy of tomorrow will not consist of “jobs”, but will move in the direction of “gigs” (Stewart & Stanford, 2017). The traditional employment or the so-called long-term employment is shifting towards a future without employers (Friedman, 2014). In this new way of working, workers will work through online platforms that have a short-term employment focus. Hence, the gig economy is born, though different terms have been used to describe the gig economy, such as ‘crowdsourcing’, ‘sharing economy’ or ‘collaborative economy’ (Huws, Spencer, Syrdal, & Holts, 2017; Stewart & Stanford, ). The gig economy is an abstract concept that is difficult to specify in one definition because gigs apply to many different functions, businesses and even industries. Therefore, distinctions and clarifications are needed to define the gig economy. The gig economy consists of an independent workforce who can be divided into categories such as the self-employed, independent contractors, freelancers, some small business owners, and many temporary workers, including workers who get short-term assignments or tasks through staffing agencies or via platforms (Manyika e.a., 2016; Stewart & Stanford, 2017). The first classification to make is that there are several forms of platforms. According to (Eurofound, 2018 p. 20): “the definition of platforms refers to the coordination of economic transactions
(involving the exchange of goods and services); it therefore excludes online spaces that are sometimes considered as platforms – most importantly, social networks”. Further, a
distinction has to be made between privately owned platforms (e.g. Uber or Airbnb) and platforms most often owned by their users (e.g. Blockchain) (Eurofound, 2018). Transactions can be paid or unpaid services provided via a platform. The category of unpaid services is closely associated with the ideology of the “sharing economy” because goods and services are
10
shared or exchanged instead of being purchased. Platforms can focus on the exchange of goods (e.g. eBay) or the exchange of services (e.g. Deliveroo) (Eurofound, 2018). Service platforms can operate online or local, while online platforms are similar to “crowd work”, local platforms provide personal services and commonly referred to as the “gig economy” (Eurofound, 2018). De Stefano (2015) makes a further distinction between “crowd work” systems and “work-on demand systems”. The former involves jobs that can be delivered and completed online and whereby the platform functions as “matchmaker”. The latter involves more traditional tasks and jobs, by which the jobs and tasks are managed through an online platform which keeps control over different aspects of the work. The heterogeneity and the complex nature of the different platforms require further clarification. Therefore, this study focuses on local service platforms or the so-called “work-on demand systems” (e.g. Uber) that exchange services, because service platforms have a closer resemblance to the traditional worker and employer relationship compared to platforms that exclude personal services (e.g. eBay).
Despite the abstract concept of the gig economy, there are several key features that embody most forms of gig work (Stewart & Stanford, 2017). Gig work usually consists of irregular work schedules, that are heavily influenced by fluctuations in demand. In addition, gig workers typically have to supply most of their own equipment used in their work. Work equipment can vary from a bicycle for food delivery to more complex equipment, such as a computer. Gig workers also need to supply their own workspace, such as their home or a vehicle. Compensation of gig work can present itself in many forms, but mostly in task, assignment or sale related payments. In contrast to traditional employment, gig workers have no employer in the traditional sense. Instead, gig jobs are organised around digital mediation in the form of digital platforms (Stewart & Stanford, 2017). Mckinsey & Company (2016)
tried to illuminate the gig economy by surveying more than 8,000 respondents across six different countries (Manyika e.a., 2016); the United States, United Kingdom, Spain, France, Germany, and Sweden. They concluded that gig work is defined by a high degree of
autonomy, payment that is based on a task or sales and a short-term relationship between worker and client (Manyika e.a., 2016). The gig economy is a rising phenomenon that exists all around the world. For instance, a growing number of Dutch workers are no longer
employed in traditional jobs with a long-term focus on employment and connection with the organization (Dhondt, Huws, Zander, & Dragonimerescu-Gaina, 2016). Instead, this group is hired for “gigs” under flexible work arrangements. A study conducted by Dhondt et al. (2016) included more than 2,000 Dutch respondents, aged between eighteen and seventy years, and
11
found that 18% of the Dutch adult workforce tried to find a job via the “platform” or “gig” economy. In addition, Dhondt et al. (2016) showed that one in eight people of the Dutch workforce made money via jobs in the gig economy. Additionally, for 6% of the respondents’ gig work was the main source of income. In the United States, the number of alternative work arrangements, which includes gig work, is rising every year compared to traditional work arrangements (Friedman, 2014). Huws et al. (2017) collaborated with the foundation for European Studies and UNI Europa to create a large-scale questionnaire exploring gig work in the European Union. Huws et al. (2017) estimated that 4.9% of the working population in the Netherlands was undertaking in crowd work, the equivalent number of 600,000 people. In addition, 1.2% of the Dutch working population was dependent for 50% of their income for gig work and the use of an app. Additionally, 4,9% of the working population in the United Kingdom was undertaking in crowd work, the equivalent number of 2,260,000 people. In contrast to the Netherlands, 1.5% of the United Kingdom working population was dependent for 50% of their income for gig work and the use of an app. A study by Katz and Krueger (2015) in the United States sent out more than 6,000 questionnaires to subjects, and a total of 3,850 subjects completed the questionnaire. The goal of their study was to fill the gap
regarding the number of alternative or not standard work arrangements including gig work (Katz & Krueger, 2016). Katz and Krueger (2016) found an increase in alternative work arrangements from 10.7% in February 2005, to 15.8% in late 2015. Workers who provided services via online intermediaries, such as Uber, accounted for 0.5% of all working
population in 2015 (Katz & Krueger, 2016), however, Mckinsey & Company estimated that 4% of the United States working population worked this way (Manyika e.a., 2016). Reasons for these differences can be distributed to shortcomings in government data and different definitions used by researchers, yielding in different ranges and outcomes (Manyika e.a., 2016). This growth in gig work is a steady trend over the years and occurs in a lot of different functions, sectors, markets, and even countries (Dhondt e.a., 2016; Friedman, 2014; Huws e.a., 2017; Katz & Krueger, 2016; Manyika e.a., 2016).
What are the potential implications of platforms and employment, given the rising growth in gig work? A direct implication is the employment conditions of gig workers because platforms do not fit conventional categories of self-employment and dependent employment (Eurofound, 2018). Concerns have been raised about the current situation of gig workers; the limited contractual protection, isolation in work, lack of contact (except with the client) and lack of autonomy (Eurofound 2018; (De Stefano, 2015; Friedman, 2014; Huws
12
e.a., 2017). In contrast, reports have shown that gig workers by choice are more satisfied with their job compared to traditional jobs (Manyika e.a., 2016). In addition, gig work can benefit organizations and the economy by stimulating employment, raising productivity, increasing flexibility and increasing the availability of services (Manyika e.a., 2016).
2.2 Human resource management (HRM)
Human resources are an organization’s greatest assets because they are the key drivers to everyday business functions, such as managing cash flow, performing business transactions, communication and dealing with customers (Haslinda, 2009). In addition, human resource management practices are an important factor for matching people with the organizations and jobs they work in. Human resources are therefore essential for organizational success (Boon, Den Hartog, Boselie, & Paauwe, 2011; Haslinda, 2009; Schneider & Bowen, 1993).
Therefore, it is no surprise that management of human resources has been of keen interest to human resource practitioners and academics. Since the early 1980s, there has been much debate about the field of HRM, resulting in different subfields of strategic and functional HRM (Wright & Boswell, 2002). The review of Mahoney & Deckop (1986) identified a number of trends, such as HR departments moving away from personnel administration to HRM and from personnel planning to a more strategic role of HRM (SHRM). SHRM as a concept has been used and debated intensively, but there is not one commonly adopted definition. Wright and McMahan (1992) defined SHRM as “the pattern of planned human
resource deployments and activities intended to enable the firm to achieve its goals” (Wright
& Snell, 1998). Research in the field of SHRM tries to explain and increase the knowledge on the linkage between the strategy and outcomes of an organization. In contrast, functional HRM research explores the impact of HR practices on individuals (Wright & Boswell, 2002). Functional HRM entails examining HR practices in an experimental context and then
exploring the effect on individuals. Moreover, individuals can be assessed on different traits (e.g. skills, abilities, opportunities, etc.) and the impact to organizational performance measures (e.g. productivity, turnover, absenteeism, etc.) (Wright & Boswell, 2002). In both areas, functional HRM research tries to account and identify for variance across individuals. While SHRM operates at a macro level, functional HRM operates more at a micro level, thus there are multiple levels of analysis within HRM research. However, this is not the only distinction, another distinction concerns the number of HR practices used in HRM research (Wright & Boswell, 2002).
13 2.2.1 HR practices
A great number of studies has looked at the effect of a single HR practice (e.g. selection), independent of any other present HR practices at that time. However, other researchers have examined the effect of multiple HR practices as a system, or the so-called high involvement-, high commitment-, high performance work systems (HPWS’s) or HR practice configurations (B. E. Becker & Huselid, 1998; Huselid, 1995; Lepak & Snell, 2002; MacDuffie, 1995; Whitener, 2001). Huselid & Becker (1998) argue that an organization’s HRM system and financial performance are strongly related. On the one hand, organizations see the strategic importance of a skilled and motivated workforce. On the other hand, organizations do not always value HR as a strategic asset. This difficult relationship embodies the need for strong HRM systems that go beyond HR practices and their effects. HRM systems can leverage the strategic advantage of organizations through internally aligning HR practices and policies with the organization’s outside competitive strategy. Such internally aligned HRM systems often consist of strongly developed recruitment, section, training, participation, and
compensation practices that have been well balanced and bundled (B. E. Becker & Huselid, 1998; Huselid, 1995; Pfeffer, 1994). The HR practices as a system approach takes into account that individual HR practices can have synergistic, complementary or conflicting effects with other practices. A possible explanation for these effects is that organizations that align their HR practices can send consistent messages to employees, regarding the
expectations from both sides (Rousseau & Greller, 1994). In addition, according to MacDuffie (1995), there are three conditions that need to be met for achieving additional value with an HRM system. First, workers need to have skills and knowledge that are not available to managers in the organization. Second, workers need to be motivated and have the autonomy to execute their skills and knowledge. Last, the bundle of HR practices needs to be integrated with the organization’s strategy. The system of HR practices is needed to ensure that the conditions are evaluated and achieved. Organizations often create HRM systems that are based on the organization’s culture and strategy (MacDuffie, 1995). The interactions between managers and employees are contingent on the combination of HR practices in a system, rather than the individual practices (Cutcher-Gershenfeld, 1991). Therefore, focusing only on the impact of a single HR practice on performance instead of the whole HR system may be a fallacy that leads to an HR practice capturing the value of the HR system (Ichniowski, Shaw, & Prennushi, 1997). Thus, to truly capture and examine the impact of HR practices on
14
Ketchen, 2006; Delery & Doty, 1996). Therefore, this study focuses on the impact of HR practices as a system.
2.2.2 HRM outcomes
Significant research attention has been devoted over the years to understand the impact of HR practices as a system on organizational outcomes (Bae & Lawler, 2000; Huselid, 1995; Wright & Boswell, 2002). Organizations can use HR practices to influence organizational outcomes, this notion is based on the perspective of Ostroff & Bowen (2000) that HR practices shape the skills, attitudes, and behaviors of workers, which in turn can positively affect organizational behavior and outcomes. Bailey (1993) argued that human resources are frequently “underutilized” because workers are unlikely to perform at their maximum potential. An organised effort to draw out discretionary effort can be worth the investment. HR practices can assist in drawing out discretionary effort by influencing workers motivation and skills (Bailey, 1993). Another insight into how HR practices can influence organizational performance builds upon Barney (1991) resource-based view of the firm. Wright & McMahan (1992) argue on the basis of the resource-based theory of firms that human resources can provide a source of sustained competitive advantage if they met four basic conditions. First, human resources need to add additional value to the organization’s production process. Second, the skills the organization is looking for must be rare. The third point is that the investments in human capital cannot be easily imitated. Last, an organization human resources must be well protected against replacement by competitors or technological
advancements. In addition, Ostroff (2000) found some support for a universalistic relationship between HR practices and performance. However, this relationship was dependent on the firm’s strategy. Since these findings can explain the relationship between HR practices and organizational performance, it is important to look at specific outcomes. For example, some studies have shown that HR practices influence organizational support (Allen, Shore, & Griffeth, 2003; Rhoades & Eisenberger, 2002; Tremblay, Cloutier, Simard, Chênevert, & Vandenberghe, 2010; Wayne, Shore, & Liden, 1997; Whitener, 2001). In addition, HR practices have been linked to several organizational outcomes, such as organizational commitment, job satisfaction, turnover intention, absenteeism and productivity (Allen e.a., 2003, 2003; Bae & Lawler, 2000; B. E. Becker & Huselid, 1998, 1998; Delery & Doty, 1996; Huselid, 1995; Ileana Petrescu & Simmons, 2008; Kinnie, Hutchinson, Purcell, Rayton, & Swart, 2005; Lepak & Snell, 2002; Murphy, Torres, Ingram, & Hutchinson, 2018; Whitener, 2001; Wright & Boswell, 2002). Although these studies demonstrated relationships between HR practices and organizational outcomes, there remains some controversy between these
15
two concepts. Research that is concerned with understanding the drivers of HRM and
organizational outcomes is subject to the “black-box” problem (Purcell, 2003). The black-box refers to the issue of not exactly knowing how HR processes improve organizational
performance (Wright, Gardner, & Moynihan, 2003).
2.2.3 Composition of HR practices in the gig economy
HR practices can affect workforce outcomes through three dimensions (Gardner, Wright, & Moynihan, 2011). The first dimension is motivation-enhancing practices, such as feedback, both individual and group feedback. The second dimension is empowerment-enhancing practices, such as information sharing or participation and the third dimension consists of skill-enhancing practices, such as recruiting and training. Lepak, Liao, Chung, and Harden (2006) proposed that HR practices can be categorized into one of these dimensions. The dimensions of motivation, empowerment, and skill-enhancing practices are similar to the Ability, Motivation and Opportunity (AMO) theory that consists of ability, motivation and opportunity (Appelbaum, 2000). Skill-enhancing practices are practices to obtain and keep skilled employees. Examples of skill-enhancing practices are selection, recruitment, and training. Motivation-enhancing practices are designed to keep employees motivated. Motivation-enhancing practices are compensation benefits, performance appraisals, job security, career development and incentive rewards. Empowerment-enhancing practices or opportunity-enhancing practices are designed to give employees the freedom to enhance their motivation and skills to obtain organizational outcomes. Empowerment-enhancing practices consist of participation, information sharing, work teams and flexible job designs.
Considering the value of human capital in long-term arrangements between employer and employee and the absence of this in the gig economy, new insights are needed to better understand the effects of HR practices in a different context. Horney (2016) described the gig economy as a potential disruptor for HR practices. To address these concerns, organizations need to change their HR practices. According to Horney (2016), organizations should adapt and align their HR practices, such as selection, job design, training, and participation, to match the needs of gig workers. The HR practice of selection is a highly researched topic because it is considered as a possible predictor of employee effectiveness (Wright & Boswell, 2002). This is likely due to the emergence of a well-known and accepted personality
framework called the “Big Five” (Goldberg, 1990). Personality has been shown to predict organizational and work outcomes such as task and contextual performance, and employment status (De Fruyt & Mervielde, 1999; McManus & Kelly, 1999). Training as an HR practice has been researched at many different levels, one of these areas is socialization and
16
orientations of new hires (Klein & Weaver, 2000). Klein & Weaver (2000) found out that workers that attended new hire training programs were more accustomed to the organization’s history, goals and people and a result were more likely to stay committed. Considering the short-term relationship between platforms and gig workers and non-traditional hiring methods of platforms, it is interesting to see how gig workers are affected by the practice of training and selection. Participation as an HR practice is associated with the changing nature of work (Wright & Boswell, 2002). Past research has shown a positive relationship between
participation in decision making, and outcomes such as employee performance and job satisfaction (Locke, Alavi, & Wagner III, 1997). Perhaps the biggest change in the nature of work in recent years is the gig economy (Horney, 2016). The influence of participation in the gig economy can give more insight into how gig workers perceive their relationship with their platform. Job design refers to the actual structure of jobs that employees perform (Oldham & Fried, 2016). Thus, tasks or activities that gig workers perform and execute for their platform on a daily basis (Oldham & Fried, 2016). Studies so far have looked at the relation of job design with traditional outcomes such as satisfaction and performance, however recent research is expanding to other areas such as the impact on individuals and their employing organization (Oldham & Fried, 2016). In the gig economy, platforms have a different relationship with their gig workers compared to traditional work arrangements (Friedman, 2014). Not much is known about this relationship, therefore job design can explore this relationship in the gig context to find new insights. A better understanding is needed on how job design impacts the relationship between individuals and non-employing platforms because platforms need to understand how their HR practices influence organizational outcomes and gig workers. This research will look at the relationship between perceived HR practices by gig workers, such as selection, job design, training and participation, and organizational
outcomes.
2.2.4 Perceptions of HR
Previous research has shown that people’s perceptions of HRM practices might play a significant role (Boon e.a., 2011). Bowen & Ostroff (2016) mention that there are
discrepancies between perceptions of HR practices by worker and leaders in the same organization. Relations between line managers’ reports of practices in place and employees’ reports of practices and their experiences are relatively weak (Ostroff & Bowen, 2016).
HR practices allow organizations to communicate with their workers through sending signals about which behavior and goal standards are expected (Rousseau, 1995). Because people have different values, expectations, and preferences, signals send by the organization
17
tend to be interpreted differently, thus HR practices might be perceived differently (Bowen & Ostroff, 2004). Bowen & Ostroff (2004) make the distinction between HR practices and HR process. HR practices are considered the “what”, and HR process is the “how” through which the HR system delivered by sending signals to employees (Ostroff & Bowen, 2016). HR practices sent signals and communication to employees from which employees draws attributions about. These attributions are used by employees to develop an understanding of what is valued, expected and rewarded, which represent the intended message (Ostroff & Bowen, 2016). In contrast, HR process focuses on several mechanisms through which a message can be delivered explicitly, which represent the medium (Ostroff & Bowen, 2016). Therefore, a strong HRM system entails process mechanisms that make the bundle of practices coherent, salient, clear, visible and understandable, which result in a shared perception of the organizational climate and builds consensus (Ostroff & Bowen, 2016). Perceptions have a significant impact on the relationship between HR practices and organizational outcomes (Guest, 1999). In addition, the impact of HR practices is partly determined by the way people perceive HR practices (Nishii & Wright, 2007). Considering the following; since there is limited research that has investigated the effects of employee perceptions and discrepancies between manager and employee perceptions of HR practices, this research will focus on employee perceptions (Guest, 1999a; Ostroff & Bowen, 2016). In addition, gig workers do not always have managers in the traditional sense but are often reviewed by clients via apps (e.g. Uber, Deliveroo etc.), thus making employee perceptions more relevant in the gig economy (Cunningham-Parmeter, 2016).
2.2.5 HR in the gig economy
The rise of digital platforms brings with it, many challenges for organizations. For example, Manyika e.a., (2016) argue that digital platforms could challenge the established notions of how organizations are structured. In addition, the nature of work may be evolving again (Manyika e.a., 2016). For organizations, hiring gig workers require careful
consideration. On the one hand, gig workers can allow organizations to become more nimble, agile and productive (Manyika e.a., 2016). On the other hand, trade-offs must be made to entrust workers with confidential or high-profile projects which have less of a connection with the organization (Manyika e.a., 2016). These challenges pose questions for HR departments, on how to handle these difficulties. Exploring HR practices in the gig economy can provide answers for these challenges.
18 2.3 HR and organizational commitment
The potential impact of HR practices on organizational commitment has received less interest than it deserves (Meyer & Smith, 2000). Employers often value workers that are dedicated and loyal to their organization (Rhoades & Eisenberger, 2002). Organizations can benefit greatly from committed workers. Previous studies suggested that emotionally
committed workers have greater performance, are less likely to quit their job and have reduced absenteeism (Rhoades & Eisenberger, 2002; Watson, 1999). There have been some studies that looked at the relationship between HR practices and organizational commitment and these suggest that HR practices can influence employees commitment (Allen e.a., 2003; Arthur, 1992; Kinnie e.a., 2005; Meyer & Smith, 2000; Whitener, 2001).
Organizational commitment (OC) as a concept has received a great deal of interest, mainly because it has been used in studies as a consequence or antecedent of work-related variables (Mathieu & Zajac, 1990). A better understanding of commitment results in more insight in employees and organizations. OC has been studied in many forms in the academic field, resulting in different types of OC. Attitude commitment is arguably the most well-known form of OC. Mowday, Steers, & Porter (1979) state that attitude commitment represents a state in which an individual identifies with a specific organization and its goals and wishes to maintain membership to obtain these goals. Calculated commitment is another form of OC. Calculated commitment is a state in which employees invested in an organization and cannot “afford” to separate themselves from it (Mathieu & Zajac, 1990). Other forms of OC are normative and organizational identification. These forms, however, are nearly identical to attitude or calculated commitment and therefore subsumed into attitude or calculated commitment. (Mathieu & Zajac, 1990). Calculated commitment is build upon the work of (Becker, 1960)OC is related to performance and employee’s absenteeism (Mathieu & Zajac, 1990). This study, however, is focused on OC at an overall level, because while
distinctions can be relevant for finding what causes turnover, this study is interested in the general levels of OC in the gig economy.
OC as a concept in the gig economy is different compared to traditional work environments. While many younger workers have been disappointed that they are unable in finding a secure employment, the gig economy shocked older workers and they struggle to find a role in this new way of working (Friedman, 2014). The gig economy thrives on short employment relations, making the concept of lifetime employment at one organization
redundant. This gives a different meaning at the concept of commitment and begs the question if platforms can keep gig workers committed? Traditionally long employment contracts
19
assured labor in times of labor shortage for organizations, while being counterproductive in times of crisis (Friedman, 2014). For organizations in the gig economy this works differently, they benefit from recessions to obtain cheap labor, however in times of prosperity they have to pay more for the same workers (Friedman, 2014). Therefore, keeping gig workers
committed can be beneficial for platforms.
2.3.1 Social Exchange Theory (SET)
Social exchange theorists argue that resources received from others are valued higher if they are given by discretionary choice instead of obligatory reasons (Rhoades &
Eisenberger, 2002). The social exchange theory (SET) is a powerful theory for understanding workplace behavior (Blau, 1964; Cropanzano & Mitchell, 2005). Many different views of SET have emerged over the years, theorists agree that social exchange involves interactions that create obligations (Cropanzano & Mitchell, 2005; Emerson, 1976). These interactions are often seen as interdependent and contingent on the actions of other individuals (Blau, 1964). Over time these interdependent interactions have the potential to create strong relationships, although these relationships are contingent on circumstance (Cropanzano & Mitchell, 2005).
Research suggests that SET includes at least six resources; love, status, information, money, goods and services (Cropanzano & Mitchell, 2005). SET assumes that relationships evolve over time, creating stronger bonds and mutual commitment. This is contingent on a set of exchange rules (Cropanzano & Mitchell, 2005). The first rule is “reciprocity”, which has many forms (e.g. normative or folk belief). This study will only look at research focused on organizational “reciprocity” (Cropanzano & Mitchell, 2005). The second rule of SET is the rule of “negotiations”, which allows parties to negotiate mutually beneficial arrangements (Cropanzano & Mitchell, 2005). Negotiations differ from reciprocity in the sense that they are more clear and quid pro quo (Cropanzano & Mitchell, 2005). Building on SET, this study proposesthat if gig workers perceive HR practices as supportive, gig workers will reciprocate these supportive feelings back to the platform, and as a result gig workers will be more
committed to the platform. Thus, this study will be looking at the relationship between perceived HR practices of platforms and organizational commitment.
H1: Perceived HR practices have a positive and direct relationship with organizational commitment.
20
2.4 HR and organizational commitment and the mediating role of POS
Perceived organizational support (POS) is affected by several antecedents of which HR practices is one influencing factor (Allen e.a., 2003; Whitener, 2001). Other influence factors are politics, justice, job conditions, trust, appraisal supervisor support and personality (Allen e.a., 2003; Whitener, 2001). Huselid (1995) argues that POS can help reduce turnover and increase commitment (Allen e.a., 2003). This study however explores the possibility of other HR practices being antecedents of POS such as training, selection, job design and participation as a system, which remains unexplored. HR practices that specifically imply that the organization values their workers, such as participation and training, will lead to higher levels of POS. Supportive HR practices are practices that indicate that the organization recognizes employee contributions. Supportive HR practices are discretionary in nature, meaning that the organization is not compelled to offer to practice to all employees. According to the organizational support theory, the development of POS is stimulated by workers habit to assign the organization humanlike characteristics (Eisenberger, Huntington, Hutchison, & Sowa, 1986). Positive or negative perceptions of the organization’s intent can lead to favourable or unfavourable treatment indicators (Rhoades & Eisenberger, 2002). Perceived organizational support can be defined as the availability and assurance of aid that the organization gives to employees to help them deal with challenging job situations (Rhoades & Eisenberger, 2002). Organizational rewards that are perceived as more genuine opposed to external constraints contribute more to POS (Eisenberger e.a., 1986). Examples of regulated organizational constraints are union negotiations or governmental health and safety restrictions.
Employees, however, are more interested in the organization’ commitment to them (Rhoades & Eisenberger, 2002). Employees that perceive being appreciated by the
organization are more likely to receive benefits, such as approval, respect, promotion and access to information (Rhoades & Eisenberger, 2002). This relation between employers and employees is based on mutual reciprocity. The organizational support theory argues that the organizational intent to reward increased work effort and fulfil employees social needs, employees develop global beliefs regarding the extent to which the organization values their contribution and cares about their well-being (Rhoades & Eisenberger, 2002).
Eisenberger et al. (1986) found that individuals with a strong exchange orientation are more likely to return a good deed, then individuals with low exchange orientation. In addition, the social exchange process can strengthen the POS relationship between employees and organizations with feelings of obligation (Eisenberger, Armeli, Rexwinkel, Lynch, &
21
Rhoades, 2001). Witt & Wilson (1990) found that social exchange process can improve the relationship between perceptions of income and organizational outcomes such as job satisfaction and commitment. For platforms the social exchange relationship means that if platforms take care of their gig workers, which results in mutual and beneficial effects such as effective work behavior and positive employee attitudes, gig workers are more likely to feel supported (Cropanzano & Mitchell, 2005). In addition, according to Gouldner (1960) the reciprocity norm assumes that individuals tend to feel obligated to help those who have supported them. Supported gig workers are more likely to form a relationship with their platform, thus improving mutual commitment. Therefore, this study proposes that perceived HR practices influence POS, and consequently greater levels of POS by gig workers is expected to lead to greater feelings of attachment and obligation to the platform (Allen e.a., 2003).
H2: Perceived HR practices have a positive relationship with organizational commitment mediated by perceived organizational support, so that higher levels of perceived
organizational support will lead to higher levels of organizational commitment.
2.5 HR and job satisfaction
Previous research has looked at the relationship between HR practices and job satisfaction and found a significant effect between the two variables (Huo & Boxall, 2018; Ileana Petrescu & Simmons, 2008; Khoreva & Wechtler, 2018; Kinnie e.a., 2005). Perceived HR practices can positively affect the employee and the work context in several ways, thus influencing job satisfaction. However, differences between the intended HR practices by HR and perceived HR practices by employees can exist and can affect the intended goal (Khilji & Wang, 2006).
Locke (1969, p. 316) defines job satisfaction as “the pleasurable emotional state
resulting from the appraisal of one’s job as achieving or facilitating the achievement of one’s job values”. Job satisfaction has been the topic of many empiric studies, but there still is complexity in what causes job satisfaction (Locke, 1969). This complexity resulted in different views on job satisfaction: the “intrinsic” view whereby the determinants of job satisfaction are attributed to the employee’s job, and the “extrinsic” view looks at the
employees’ mind as the cause of job satisfaction. Job satisfaction can also be determined and attributed to the interaction between the employee and the work environment. Agho, Mueller, & Price (1993) discuss three influencing factors that affect job satisfaction. These factors are job characteristics, environmental effects and personality variables. Given the complexity of job satisfaction, there are several types of job satisfaction measures (Scarpello & Campbell,
22
1983). A distinction can be made between a measure of overall job satisfaction and a measure of job facet satisfaction (Scarpello & Campbell, 1983). While job facet satisfaction measures can give insight in why individuals leave their organization, this study focuses on overall job satisfaction, because this study is interested in the overall levels of job satisfaction of gig workers working for platforms. The topic of job satisfaction in the gig economy may seem to lack relevance at first glance, considering gig workers are classified as self-employed, but there is substantial opportunity for researchers to increase our understanding of satisfaction as a concept, considering the growing number of individuals who earn their income via gig work (Kuhn, 2016). Platforms have significant control over gig workers compensation and working conditions (Kuhn, 2016). Sundararajan (2014) goes a step further and argues that platforms develop their own “cultures”. Sundararajan (2014) notes that some platforms like Airbnb invest a lot of resources in developing a platform culture, while other platforms fail to do so (e.g. Uber). Therefore, these platforms seem to have more similarities with traditional organizations than they like to admit. Reports also show that gig workers are satisfied with their autonomy, but have difficulties communicating with their platforms and issues about collecting payment (Huws e.a., 2017; Kuhn, 2016). Therefore, this study will explore the job satisfaction of gig workers. To explain the relationship between HR practices and job
satisfaction, this study will build upon the signaling theory (Spence, 1978). Researchers have applied the signaling theory to help explain the influence of information asymmetry in several different research fields (Connelly, Certo, Ireland, & Reutzel, 2011; Spence, 1978). One of these research fields is HRM, by exploring signals during the recruitment process of job applicants (Suazo, Martínez, & Sandoval, 2009). Information is of vital importance for the decision making process of individuals. Information asymmetry can arise when some parties have less information than other parties (Connelly e.a., 2011). Information can be separated into two types where asymmetry is especially crucial: information about intent and quality (Stiglitz, 2000). Generally, parties with less information make less well-informed decisions, thus favouring the party that has access to more information, however this is contingent on the intent and quality of the information, which dictates the severity of information asymmetry (Connelly e.a., 2011). Platforms can use the signaling theory to send positive signals to gig workers by deliberately focusing on positive organizational attributes, such as flexible working arrangements. Gig workers working for platforms tend to have flexible work
arrangements. A study by Baltes, Briggs, Huff, Wright, & Neuman (1999) found that flexible work arrangements were related to performance and job satisfaction. Based on the signaling theory, this study proposes that HR practices, such as flexible work arrangements of gig
23
workers, increase the attachment to the organization in the form of increased job satisfaction (Casper & Harris, 2008; Grover & Crooker, 1995).
H3: Perceived HR practices have a positive and direct relationship with job satisfaction.
2.6 HR and job satisfaction and the mediating role of POS
Previous research has found that POS has a positive relationship with job satisfaction (Allen e.a., 2003; Kurtessis e.a., 2017). Job satisfaction and POS are thus related, but they remain distinctive constructs (Eisenberger, Cummings, Armeli, & Lynch, 1997). This study proposed that perceived HR practices directly affected job satisfaction based on the signaling theory. However, some theorists argue that this relationship is mediated by feelings of
perceived organizational support (Eisenberger e.a., 1997). Gig workers experience higher levels of autonomy, but at the cost of employment securities of traditional work arrangements (Friedman, 2014). Thus, given that platforms have fewer obligations towards gig workers compared to traditional organizations and employees, it is more likely that gig workers
perceive actions of platforms as more voluntary. Based on SET, this study proposes that if gig workers feel that HR practices of the organization are voluntary and in the best interest of them, opposed to mandatory actions (e.g. contractual obligations), then they are perceived as more supportive, leading to higher job satisfaction. Therefore, this study argues that perceived organizational support of platforms partly mediates the relationship between perceived HR practices and job satisfaction.
H4: Perceived HR practices have a positive relationship with job satisfaction mediated by
perceived organizational support, so that higher levels of perceived organizational support will lead to higher levels of job satisfaction.
24 3. Method
3.1 Sample & Procedure
To test the hypotheses about the hypothesized relationships, a questionnaire was distributed. The questionnaire contained questions about demographic variables, control variables, screening, quality response items and the measured variables. The software program Qualtrics was used to create the questionnaire. The data was collected by handing out physical questionnaires to gig workers and Qualtrics. The questionnaire was online from the 19th of April until the 19th of May. Participants were instructed to fill in the survey. Participants were also informed that the data will only be used for the purposes of this study and will be treated anonymously. Four students in the Business Administration Master of the University of Amsterdam recruited participants in collaboration with Qualtrics. This research used multiple non-probability sampling methods, such as snowball sampling and convenience sampling and was cross-sectional and correlational in design. Convenience sampling was used to target people that fitted the scope of this research. For example, we handed out traditional paper questionnaires at locations where gig workers worked. Platforms were contacted and invited to participate in this study. In total, twenty-two platforms were contacted, of which two platforms agreed to collaborate with this study. Platforms that agreed to collaborate were asked to distribute the questionnaire internally. In addition, participants were asked in the questionnaire to distribute the survey to co-workers, making use of the snowball sampling method. Qualtrics as a service was used to distribute the questionnaire in the UK and the United States to increase the sample size of this study. Resulting in a combination of web-based and traditional paper questionnaires.
Traditional paper questionnaires were inserted in Qualtrics by the researchers. The paper questionnaires were marked with codes to ensure that the questionnaires could be traced back to the dataset. By using both traditional paper and web-based questionnaires, a large enough sample size was obtained. The use of non-probability sampling techniques in
combination with the collection of participants by four different researchers supported a more diverse sample. To motivate gig workers to participate in this study, an incentive was given. People participating in this study had a chance to win a €50,- voucher. The people that filled in the questionnaire had to be gig workers matching the scope of this research. Participants of the study were screened by asking the question if they fitted the “gig worker” category that is in line with this research. Those who were included in the study worked in the Netherlands, United Kingdom or the United States of America. People that participated could have multiple jobs or work in several sectors. The scales in the questionnaire were in English and
25
Dutch. The scales that were not originally in Dutch were translated using the back-translation method as described in the measures section. In total, 280 responses were collected in
Qualtrics. The data quality was checked by excluding cases on several quality questions. All respondents that failed on the first question, whether or not the respondent was suitable for this research, were excluded. A number of respondents failed this question and they did not meet the criteria of being a gig worker. The three quality check questions were also examined. However no cases needed to be excluded based on the quality check questions. Furthermore, several cases were removed based on the basis of completion rate and response duration. All cases that finished the survey within 5 minutes were excluded. After data cleaning, the dataset consisted of 181 respondents.
Of the questionnaires, 13,81% was collected as hardcopy and 86,19% was digital. The average age of respondents is 32.6 years old (SD=10.41). 63,54% of the respondents were male and 36,46% were female. The average “employment” of gig workers at organizations is 3,15 years (SD=4,80). An average of 46,94% of the participants in the survey worked in the USA, 17,13% in the UK and 35,91% in the Netherlands. The average gig workers worked around 24,01 hours each week ((SD=13,69). The work in European gig economy survey had a similar ratio of male and female in the Netherlands, consisting of a 60% male and 40% female ratio, with a sample size of 103 (Huws, Spencer, Syrdal, & Holts, 2017).
3.2 Analytical strategy
First, the data has been imported into IBM’s Statistical Package for Social Sciences (SPSS) 24.00. In SPSS the data was analysed. In addition, the statistical tool PROCESS as developed by Hayes, was used to analyse the mediation effect. A mediation analysis was conducted on Hayes’ model 4 (2007). The data was checked for errors by examining the frequencies. This resulted in no issues. The variables of perceived HR practices, POS and Job satisfaction had counter-indicative items, meaning that these items were phrased so that an agreement with the item represents a low level of the construct. Also, several questions had an “I don’t know” option which resembled a “99” in the data. All cases with “99” and “999” were also categorized as missing data. All the variables with counter-indicative items were recoded. A listwise correlation matrix was created showing the means, standard deviations and correlations. Reliability tests were performed to test the internal consistency of the variables perceived organizational support, organizational commitment and job satisfaction. The internal consistency of perceived HR practices was tested by combining the variables training, participation, and selection. Job design was removed because of severe reliability issues. Reliable Cronbach’s Alpha scores need to be 0.7 or higher (Nunnally & Bernstein,
26
1994). The hypotheses in the previous chapter were tested by using Hayes’ model 4 of
PROCESS. Finally, a path analysis with bootstrapped estimation of indirect effects was tested through PROCESS.
3.3 Measures
To measure and operationalize the independent variables, well-established scares were used. The following variables were used in this study: perceived HR practices and perceived organizational support as independent variables, job satisfaction and organizational
commitment as dependent variables. Perceived organizational support was used as a mediator variable between perceived HR practices and organizational commitment. Items that used measures that were not published in Dutch were translated to Dutch. This back-translating was double checked by different persons. Small adaptations were made to the original scale by changing the terminology from organizations, firms or companies to “platforms” to make it more suitable for gig workers. The questionnaire was available in English and Dutch.
Selection was measured with the validated 4 items of the scale of Sun et al. (2007) and
has a Cronbach’s alpha of 0.92. This study used an adapted version with only 3 questions (E.g. Great effort is taken to select the right person). The scale ranged from 1 to 5 (strongly disagree to strongly agree).
Training was measured with the adapted 16 items of Jiang et al. (2017) and has a
Cronbach’s alpha of 0.90. This study used an adapted version with only 3 questions (E.g. The platform continuously provides training programs.) The scale ranged from 1 to 5 (strongly disagree to strongly agree).
Participation was measured with the adapted 13 items of (Jiang e.a., 2017) and has a
Cronbach’s alpha of 0.88. This study will use an adapted version with only 4 questions (E.g. Employees are often asked to participate in work-related decisions.) The scale ranged from 1 to 5 (strongly disagree to strongly agree).
Perceived organizational support was measured with the short 8-item scale of
Eisenberger et al (1986) and has a Cronbach’s alpha of 0.79 (E.g. The platform values my contribution to its well-being.) The scale ranged from 1 to 7 (strongly disagree to strongly agree).
Job satisfaction was measured with the 3 items of Cammann et al. (1979) and has a
Cronbach’s alpha of 0.62 (E.g. All in all I am satisfied with my job.). By removing the reverse question, the Cronbach’s Alpha was improved to 0.87. The scale ranged from 1 to 5 (strongly disagree to strongly agree).
27
Organizational commitment was measured with the short 6-item scale of Mowday et
al. (1979) and has a Cronbach’s alpha of 0.91. (E.g. I really care about the fate of this platform.) The scale ranged from 1 to 6 (strongly disagree to strongly agree).
Job design was measured with 3 items of Bae & Lawler (2000) and has a negative
Cronbach’s Alpha score after recoding the reversed questions. Therefore, this variable violated reliability assumptions and was excluded from further analysis. (E.g. The platform provides broadly designed jobs requiring a variety of skills). The scale ranged from 1 to 6 (strongly disagree to strongly agree).
Perceived HR practices will be measured by averaging the items of training, selection,
and participation and has a Cronbach’s Alpha of 0.96. Thus, the combined practices of
training, selection, and participation were measured as one construct. The scale ranged from 1 to 5 on all items (strongly disagree to strongly agree).
Control variables. The control variables, gender, age, working hours per week,
working country, contract hours, payment form, platform (different organizations) and income source were all evaluated in a correlation matrix. After evaluating all control variables, a second correlation matrix was created with the purpose of including only the high number of significant correlations. For several control variables, such as platform, working country, contract hours and payment form, dummy variables were created. The platform with significant correlations with the main variables was renamed to “Organization X” to ensure anonymity. All control variables that had significant correlations with the main variables in this study were included for further analysis.
28 4. Results
4.1 Results of the correlation matrix
In the result section we take a look at the correlations, descriptive statistics and the results of model 4 of PROCESS.
Further examination of the correlation table leads to several significant correlations. All main variables had significant correlations with each other. According to Cohen's (1992) rules of thumb a correlation of 0,1 is a small effect, 0,3 is a medium effect and 0,5 is a strong effect. Perceived HR practices had a strong and significant effect (p < 0.01) on perceived organizational support and organizational commitment. The correlation between perceived HR practices and job satisfaction had a medium effect, but was still significant (p < 0.01). Perceived organizational support had a strong and significant (p < 0.01) correlation with job satisfaction and organizational commitment. As shown in the correlation matrix, several control variables were significantly correlated to the main variables in this research. All significant correlated control variables were age, gender, working hours per week, working country, payment form and “organization X” and therefore were controlled for in the analysis.
Table 1: Descriptive Statistics and listwise correlations of all Variables in this Study
M SD 1 2 3 4 5 6 7 8 9 10 11 12 13 14
1. Perceived HR practices 3,48 1,20 1 (.96)
2. Perceived organizational support 4,45 1,22 .516** 1 (.79)
3. Job satisfaction 3,78 0,82 .304** .534** 1 (.87)
4. Organizational commitment 3,72 0,94 .797** .601** .493** 1 (.91)
5. Age 32,57 10,40 .296** 0,14 .227** .335** 1
6. Gender 0,37 0,48 0,135 0,08 0,05 .222** 0,08 1
7. Working hours per week (platform) 24,05 13,64 .350** 0,11 0,07 .260** .186* -0,02 1
8. USA (working country) 0,48 0,50 .494** .216** .159* .494** .445** .188* .249** 1
9. NL (working country) 0,35 0,48 -.682** -.234** -0,1 -.625** -.438** -.244** -.246** -.705** 1
10. Zero Hour (contract) 0,30 0,46 -.223** -.189* -.158* -.224** -0,13 -0,01 -0,06 -.195** .148* 1
11. Set Hours (contract) 0,34 0,48 .197** 0,13 .177* .159* 0,01 0,05 .154* -0,09 -0 -.466** 1
12. Task (payment form) 0,55 0,50 .255** 0,04 -0,1 .178* 0,13 0,05 0 .359** -.458** -.167* -.170* 1
13. Hours (payment form) 0,42 0,49 -.310** -0,05 0,02 -.242** -.173* -0,12 -0,04 -.412** .514** .188* .197** -.933** 1
14. Organization "X" 0,46 0,50 .501** .161* 0,04 .464** .254** 0,03 .259** .590** -.625** -.166* -0 .434** -.398** 1
