The Impact of Social and Cultural Norms, Government Programs and Digitalization as Entrepreneurial Environment Factors on Job and Career Satisfaction of Freelancers

Abstract

At the end of the last century, digital technology emergence enabled millions of people to compete globally by remotely offering their knowledge and skills. In addition, business processes are becoming fragmented into smaller components, so-called short-term projects. These work arrangements are often carried out by so-called independent professionals (contractors), better known as freelancers. A literature review has shown that the research topic of freelancing from an entrepreneurial perspective is relatively new and has its assumptions and gaps. Different stakeholders and institutions connect, mediate and manage the services of the entrepreneurial ecosystem to support entrepreneurs. As freelancers belong to the self-employed entrepreneurial category, they are engaged in business activities and need support from their environment. To contribute to this topic, we have analyzed the relationships between freelancers’ job and career satisfaction, digitalization and entrepreneurial ecosystem factors, with the aim of making policy recommendations. Regarding the entrepreneurial environment, we analyzed the cultural and social norms and government programs that provide support, based on the measurement instrument, developed in the Global Entrepreneurship Monitor (GEM) research. We surveyed 200 freelancers (respondents) in Slovenia. Using factor analysis and structural equation modelling, the empirical results suggest that digitalization and cultural and social norms are significant factors that promote the success of new careers as freelancers in terms of job and career satisfaction. The results show that both positively influence job and career satisfaction, while government programs indirectly influence the job and career satisfaction of freelancers. This explains how the entrepreneurial ecosystem, plays a special role in supporting freelancers on their career path.

1. Introduction

At the end of the 20th century, innovations have become more frequent and their effects increasingly disruptive [1]. The emergence of new technologies, such as digital platforms, artificial intelligence, clouds, the Internet of Things, wireless networks, etc., has changed the way companies work and create value. Starting a digital business and adopting digital technologies is called digital entrepreneurship [2]. The adoption of these digital technologies takes place under the terms of digital transformation and digitalization [3]. Accordingly, the term digital transformation refers to strategic transformations that are aimed at organizational changes implemented through digitalization projects with the aim of enabling significant business improvements, while digitalization represents a comprehensive socio-technical process and implies the integration of different technologies into aspects of daily social life [4]. They influence markets, present challenges to established companies and affect business strategies. In terms of market impact, one of the most important effects of digitalization is its impact on the labor market. Open innovation became a widespread process as companies wanted to respond effectively and quickly to changing resource configurations by drawing on expertise beyond the boundaries of the enterprise [5]. As a result, large and small companies developed their organizational structure to become more flexible, project-based and agile, relying increasingly on self-employed workers [6]. In this way, digitalization has enabled more knowledge-based and decentralized production, i.e., providing services [7,8]. Many activities and work can be easily performed remotely (online) through the introduction of digital infrastructures [3]. Digitalization has enabled many workers to perform their tasks in a more efficient and fragmented manner, creating new jobs and improving old jobs in the labor market [9]. This new business model that has emerged as a result of the digital transformation is called the gig economy [10]. Gig work involves the performance of very short-term tasks or projects within the framework of specific activities carried out by individual self-employed workers and usually coordinated via online platforms or applications [11]. These self-employed individuals who perform short-term activities are usually referred to as independent professionals (contractors) or freelancers [12].

They offer many advantages to both startups and established companies, as they have the ability to select from the pool the most suitable talents for the tasks they need done. They help reduce the fixed costs of employment, save time (as they require a low level of supervision to do the work), and streamline companies’ organizational processes [13,14,15]. Companies may require new forms of flexibility or specific types of skills, while workers may be interested in improving their work-life balance, i.e., their well-being [16]. According to [17], freelancers belong to the entrepreneurial category of self-employment. They are enablers of entrepreneurship because they facilitate the application of derisking strategies, reduce financial constraints, increase entrepreneurial agility and productivity and facilitate market entry [14]. Recently, both the benefits of digitalization and freelancers have been increasingly recognized as a source of entrepreneurship by researchers, policy makers and industry.

In particular, the COVID-19 pandemic drove the trend toward remote working, as companies prepared for the measures and lockdowns and asked their employees to work from home. For many freelancers, working from home is part of their domain. This puts freelancers a step ahead of traditional workers, as they were used to working this way and already had an efficient workplace. This trend towards remote working has raised many questions about the workforce and workplace of the future. One of the questions is whether more (foreign) freelancers will be hired in the future. Data from the Online Labour Index, an economic indicator that measures the supply and demand of online freelancers in different countries and occupations in real time, show that since the period of COVID-19, the number of posted tasks has first increased and then sharply decreased [18]. According to the study [19], the increase could be explained by the fact that since the outbreak of COVID-19, many full-time workers began using online labor platforms to replace on-site jobs. The next decrease came from the fact that companies affected by the pandemic had to eliminate many tasks that were not considered so essential [20]. With the effects of the pandemic still being felt, there is no clear indication of how demand for freelancers will evolve in the short term. In the long term, freelance work will likely continue to drive remote work. However, recent fluctuations in online freelancing show how much freelance work still depends on non-digital economy [21].

In 2015 the Digital Single Market strategy was adopted by European Commission [22]. It comprises ten policy priorities and is based on three policy pillars: (a) improving access to digital goods and services, (b) an environment where digital networks and services can prosper and (c) digital as a driver for growth. It aims to become a global role model for the digital economy to help developing countries transition to the digital economy and develop and promote digital standards internationally [23]. Furthermore, in 2016, Slovenia introduced Digital Slovenia Strategy 2020. The vision of the strategy was to accelerate the progress of the digital society, to take advantage of the development opportunities of ICT and the Internet and to become an advanced digital society and the reference environment for the application of innovative approaches to the use of digital technologies at the system level [24]. In addition, in 2018, there were 122,000 self-employed in Slovenia, the highest number since the start of Labour Force Survey, with this trend continuing in 2019 [25].

In [14], the authors raise the question of job and career satisfaction, whether freelance work enables more fulfilment through project work, or whether employees with a longer-term and “more familiar” connection to the company lead to more fulfilled employees? Accordingly, we study digitalization as a phenomenon that affects the labor market and society as a whole. Therefore, our research examines perceptions of digital support in Slovenia and the influence and relevance of government programs as well as cultural and social norms that ultimately affect the job and career satisfaction of freelancers. Our research question reads:

Is the job and career satisfaction of freelancers in Slovenia related to their perception of digital technologies support, government programs as well as cultural and social norms?

This paper aims to answer the above research question by focusing on the impact of digitalization at the national level. A literature review and an empirical analysis have been carried out and the research is based on the following assumptions.

First, while Slovenia and the other EU Member States are implementing national strategies for digital transformation, the paper provides an insight into the freelancers in Slovenia and their perception of certain aspects of external environmental factors that may be useful for assessing current digital policies and future policy proposals and decisions. Although the media and science are interested in examining the challenges of freelance work, there is no academic work in this field in Slovenia. Moreover, there is no official register of freelancers in Slovenia. Accordingly, the purpose of the research is to expand the literature on entrepreneurship concerning freelancers, to ensure better perception and awareness of new trends and effects of digitalization on the labor market and to promote freelance work in the national context based on empirical evidence.

Secondly, job and career satisfaction is examined from the perspective of the external environment. Ref. [26] suggested that job satisfaction and performance depend on the correspondence between the individual’s personality and environmental factors. They can be assessed as strengths/weaknesses or opportunities/threats in an individual’s career or be relevant in terms of their perceived contribution to their entrepreneurial venture’s success or failure [27]. Research has therefore analyzed the influence of the perception of relevant aspects of the entrepreneurial ecosystem (government programs, social and cultural norms) and digital support on the job and career satisfaction of freelancers.

In the case of the data obtained, some of the constructs act both as independent and dependent variables. The test of the conceptual model is based on the application of exploratory and confirmatory factor analysis and the modeling of structural equations.

With regard to the structure of the paper, the background of the literature and the development of the hypotheses are presented in Section 2. Furthermore, data, methods and analyses are described in Section 3. Section 4 presents the results, while the final Section 5 presents the discussion with concluding thoughts.

2. Background Literature and Hypotheses Development

In 2013, based on an intelligent career framework [28] and a protean career model [29], the freelancer career driver model [30] was developed, which is the representative basis for freelancer career research. The protean career model represents a career orientation in which the key success criteria are subjective, while the intelligent career model was developed for intelligent firms and is appropriate for careers of freelancers as they sell their knowledge and skills (expertise). The model lacks the extended external component. For further research, the authors suggest that more work needs to be done on the relationship between the environment and freelancer’s career success. Therefore, we turn to the approach of [31] to explore how the environmental context affects the performance of self-employed workers. According to [32,33], the experiences of self-employed workers are shaped by the entrepreneurial climate of a country, including legal, institutional and cultural factors. Accordingly, we share a similar point with the study of [16], in which the authors examine the psychological well-being of different types of self-employed individuals at the macro level, based on their entrepreneurial ecosystem.

As already mentioned at the beginning, the introduction of new technologies is turning the jobs into smaller tasks for global and digital production [34], contributing to the growth of freelancing. Current technological transformations are not only affecting production systems, but also society as a whole. The mix of skills required to perform in modern societies will continue to evolve with the development of technology-enabled work environments, leading to the need for future generations of workers to develop digital literacy and lifelong learning skills at an early age [35]. A study by the World Economic Forum Future of Jobs [36] also shows that by 2022, 38% of businesses worldwide expect to add new, productivity-enhancing roles to their workforce. More than a quarter expect automation to create new roles in their business. In the future, companies will increasingly use contractors who work on a task-specific basis, and they also intend to use labor more flexibly by deploying remote personnel beyond physical offices and decentralizing operations [37]. The evolution towards a digital society is not so much about getting people to use technology, but rather how technology affects and changes people’s lives. Thus, in an e-inclusion perspective, individuals and their social and economic behavior rather than companies and production activities become the relevant policy objectives and the subject of analysis [38]. In line with the Introduction and the arguments above, we propose the first hypothesis:

Hypothesis 1.1:

The impact of the freelancers’ perception of the digital technologies’ support on the perceived government programs is positive and significant.

Hypothesis 1.2:

The impact of the freelancers’ perception of the digital technologies’ support on the perceived cultural and social norms is positive and significant.

Hypothesis 1.3:

The impact of the freelancers’ perception of the digital technologies’ support on the perceived job and career satisfaction is positive and significant.

To achieve this goal and strengthen the context of entrepreneurship and innovation [39,40], it is proposed to focus on the actors that play a role in the entrepreneurship environment [41,42,43,44]. Accordingly, government policy is certainly one of the most important factors that can turn entrepreneurial intentions into actual business activity. In modern market economies, government policies (programs, measures, initiatives) play an important role in shaping entrepreneurial culture, i.e, the conditions under which entrepreneurial potential and intentions can be turned into reality (entrepreneurial activities) [45]. The Global Entrepreneurship Monitor also highlights the government as a key factor in promoting and creating new businesses through specific programs and initiatives [46]. In addition, [47] it points out how government programs not only encourage and create new businesses, but also shape the culture of entrepreneurship, in which entrepreneurial values and norms such as proactivity, risk-taking, acceptance of failure, openness to new ideas, individualism, autonomy and fulfillment are collectively valued [48]. Accordingly, we propose the following hypotheses:

Hypothesis 2:

The impact of the freelancers’ perception of the government programs on the perceived job and career satisfaction is positive and significant.

Hypothesis 3:

The impact of the freelancers’ perception of the government programs on the perceived culture and social norms is positive and significant.

The external environment component is also supported by social cognitive career theory, which states that the decision-making process related to career development is influenced by both personal and environmental/contextual elements [49,50]. Although individual characteristics of entrepreneurs may be important predictors of entrepreneurial ventures, studies examining environmental effects clearly suggest that the influence of external factors should also be considered. The interaction between an individual’s perception of an opportunity and his or her perceived ability to seize that opportunity in a particular context leads to processes of business creation and development [51]. This concept is also supported in the literature by [52,53,54,55,56], who explain the significant influence of certain external factors on entrepreneurship [40]. Scholars have found that the reasons for individuals’ participation in the gig economy is additional income, followed by flexibility and autonomy, as well as interaction with clients and interest in entrepreneurship [57,58,59,60]. The research suggests that autonomy is associated with greater job satisfaction [61,62,63,64,65,66]. In particular, the autonomy that results from self-employment contributes significantly to job satisfaction, which could explain why people continue to be self-employed despite unsatisfactory economic returns [66,67,68,69,70]. Furthermore, [66] also underline the results indicating that self-employed individuals generally have more control aspects of their work compared to employees and therefore are likely to have a higher degree of autonomy in their work [71,72,73,74]. In addition, [75] suggests that the phenomenon of entrepreneurship develops in such a way that culture promotes entrepreneurial potential, which may increase personal wealth and income (society’s focus on entrepreneurship through behavior, language beliefs and customs) [76]. In this way, in the context of freelance work, knowing someone who has started to work as a freelancer can increase their awareness and sharpen their understanding of the costs and benefits involved. Getting to know other freelancers means getting in touch with role models and mentors, understanding the motivational factors or success factors and establishing connections to relevant actors. Moreover, seeing the advantages of freelance work can also point to innovation potential and demonstrates the ability to recognize such opportunities. Considering whether it is easy to go freelance can be an indication of how people perceive the environment that makes this type of entrepreneurial activity possible or restricts it. Social norms have rarely been empirically tested in the literature on entrepreneurship [77]. However, there are studies on how the values and beliefs of different cultures/groups influence the creation of new entrepreneurial ventures [78,79]. Social or material support for a person’s most important goals is likely to promote satisfaction [80,81]. This is supported by [82], whose studies also investigate effects based on social-cognitive career theory, in which individuals’ cognition influences their behavior and is influenced by their environment. The author finds out how social support contributes positively not only to the career self-efficiency, but also to career development and career beliefs [83]. Based on the above arguments, we propose the fourth hypothesis:

Hypothesis 4:

The impact of the freelancers’ perception of cultural and social norms on the perceived job and career satisfaction is positive and significant.

3. Materials and Methods

3.1. Conceptual Model

By combining the arguments from Section 2, based on the theoretical background, we proposed the model to explain the effects of the perceived government programs, cultural and social norms in the context of digitization on the job and career satisfaction of freelancers. Figure 1 presents the research model.

The factors included in the model indicate the following:

➢

Perceived digital technologies support—the level of perceived national support for the creation of a more favorable environmental infrastructure for a faster and more coordinated development of the information society and for the equal participation of Slovenian stakeholders in the global digital space.

➢

Perceived government programs—the extent of the perceived presence and quality of the government programs that create the conditions for the development of freelance activity.

➢

Perceived cultural and social norms—the perceived extent to which cultural and social norms encourage or discourage freelance activity, leading to new businesses or activities that can potentially increase personal wealth and income.

➢

Perceived job and career satisfaction—the perceived extent to which a pleasant emotional state results from the assessment of one’s career achievements and the assessment of one’s job values.

To test the reliability of the conceptual model, the empirical analysis was divided into two parts. In the first part, the factor analysis was performed for the model variables. In the second part of the empirical analysis the hypotheses were tested by structural equation modelling. In the following section the sample data are presented, followed by the results.

3.2. Sample Data

The empirical study was conducted on a sample of Slovenian freelancers. As Slovenian freelancers are not officially registered, the respondents were randomly selected from a list of self-employed persons from the database Bisnode Gvin [84], channels of online labor platforms, and online groups of self-employed persons. In order to obtain a representative sample of freelancers, we limited our research to those respondents who fall into the entrepreneurial category of self-employed with zero employees and who work remotely from home, or in coworking spaces or anywhere else where they have access to the necessary resources and who worked in the last 12 months at least with 2 clients. The original survey had 342 responses. After certain restrictions were made for the purposes of research, we came up with 200 valid responses. A total of 137 responses were collected using CATI (computer assisted telephone interviewing) technology and 63 responses were collected using CAWI (computer assisted web interviewing) technology. Data collection lasted from the end of July, 2020, to middle of September, 2020.

Considering the place of work, more than half of the surveyed freelancers (57.5%) work remotely from home (regardless of the situation during the crisis of COVID-19), 40% of the respondent work in the coworking spaces, while only 2.5% work anywhere if they have access to the necessary resources.

In terms of gender, the sample includes more male freelancers (59%) than female freelancers (41%). In terms of age, most freelancers are found in the age group 35–44 (30%), followed by the age groups 45–54 (27%), 55–64 (18.5%) and 25–34 (17.5%), while the fewest are found in the age groups 65 and older (5.5%) and 18–24 (1.5%).

Regarding the type of industry, according to the Standard Classification of Activities in Slovenia, most freelancers participate in professional, scientific and technical activities (54.5%) and information and communication activities (30%). They also appear to be active in manufacturing (4.5%), education (4%), real estate (2.5%), wholesale and retail trade, repair of motor vehicles and motorcycles (1.5%), arts, entertainment and recreation (1.5%), administration and support services (1%) and finance and insurance (0.5%).

In addition, 40.9% of the freelancers surveyed use an online work platform to look for work.

3.3. Measures

When designing the measurement instrument, we used the theoretical starting points and research work of other authors to design a survey for data collection. The individual factors were measured by multiple statements on a 7-point Likert scale [85] (1 = strong disagreement; 7 = strong agreement), in which the respondents expressed their agreement or disagreement with the defined statements. The survey contained closed questions. For the job and career satisfaction (dependent variable in the model) the adjusted 7-point Likert scale by [86] was used, which is based on the studies of [87,88,89,90,91]. The measurement scale contains 6 items. To measure independent variables or the perceived impact of government support through government programs and cultural and social norms, the measurement scales of GEM National Expert Survey were used [76]. The items were adapted to freelance activity.

Finally, in order to measure support for digital technology in line with the scientific work of [92,93] and the Strategy of the Republic of Slovenia Digital Slovenia 2020 [24], a 5-item questionnaire on a 7-point Likert scale was created in which individual freelancers expressed their perception of support for digital technology in the country. All measurement scales are presented in Section 3.6. As mentioned above, due to the process of adaptation to our field of research and due to translation procedures, some words in the measurement scales may differ slightly.

3.4. Methodology

The conceptual model was tested with the method of factor analysis and structural equation modelling (SEM).

Due to the multidimensional variables contained in the conceptual research model, factor analysis is used to reduce many individual items (contained in survey research) to a smaller number of dimensions (factors) [94]. The aim was to determine the smallest number of linear combinations of measured variables by explaining the total variance of the data as far as possible. Factor analysis is a study of the relationships between variables and the extraction of a new set of variables (less than the measured variables) representing the underlying/latent variables [95]. In this paper, two types of factor analysis are used: exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) [96]. EFA helps to determine how the factor structure looks like according to the responses of the participant and is essential to determine the underlying constructs for a set of measured variables [97]. In the absence of research on the underlying structure of the impact of digitalization on government programs, cultural and social norms and the job and career satisfaction of freelancers, exploratory analysis was chosen before CFA.

In addition, CFA is used to verify the factor structure of an observed set of variables. In addition, the aim of CFA is to determine the ability of a predefined factor structure or to determine the extent to which the observed variables that are assumed in EFA to measure a single factor actually fit into an observed data set [86].

Structural equation modelling (SEM) was used to test the research hypotheses. SEM is a statistical methodology that takes a confirmatory approach to the analysis of a given theory [98].

SEM models consist of two main components [99]:

➢

the measurement model, which indicates the number of factors, the relationship of the various indicators to the factors, and the relationships between the indicator errors (CFA model);

➢

the structural model, indicating how the various factors are related (e.g., direct or indirect effects, no relationship).

As mentioned above, CFA is used as a precursor to SEM, which specifies an acceptable measurement model, before the structural relationships between latent variables are estimated and interpreted.

For data analysis, EFA was performed with the statistics program IBM SPSS Statistic 22, while CFA and structural equation modeling was performed with the statistics program WarpPLS 7.0.

3.5. Sample Adequacy

In the first part of the data processing the EFA was used. In order to assess the adequacy and acceptability of the data, the criteria listed in

Table 1. General rules for the adequacy and acceptability of data.

Criterion	Level of Acceptance	Authors
Bartlett’s Test of Sphericity (BTS)	Chi-Square; p < 0.05	Field, 2005 [100]
Kaiser–Meyer–Olkin Measure of Sampling Adequacy (KMO)	>0.5	Kaiser, 1974 [101]
Communalities values	>0.5	Field, 2009 [102]
Factor loadings	>0.5	Janssens et al., 2008 [96]
Total variance explained	60% (in some cases even less)	Hair et al., 2014 [103] Streiner, 1994 [104]
Cronbach’s Alpha	>0.6	Janssens et al., 2008 [96]

were taken into consideration in the design of the EFA.

The analysis of the interdependence among the variables of the job and career satisfaction construct and in the case of the constructs of digital technology support, government programs and cultural and social norms, shows the appropriateness of the use of exploratory factor analysis, which is confirmed by BTS (p < 0.05) and KMO (KMO > 0.5) in

Table 2. BTS and KMO of the Model Constructs.

Construct	Kaiser–Meyer–Olkin Measure of Sampling Adequacy	Bartlett’s Test of Sphericity; Chi-Square; p Value
Job and Career Satisfaction (JCS)	0.833	0.000
Digital Technology Support (DTS)	0.664	0.000
Government Programs (GP)	0.854	0.000
Cultural and Social Norms (CSN)	0.826	0.000

.

3.6. Data Processing Method—Exploratory Factor Analysis (EFA)

Table 3. EFA results for the job and career satisfaction (JCS) construct.

Variable Label	Variable	Communalities	Factor Loadings Factor JCS
q1_JCS	I am satisfied with the success I have achieved in my career.	0.714	0.845
q2_JCS	I am satisfied with the way I feel about my job as a whole.	0.793	0.890
q3_JCS	I am satisfied with the opportunities to use my abilities on the job.	0.572	0.756
q4_JCS	I am satisfied with the progress I have made towards meeting my goals for my overall daily life.	0.622	0.789
q5_JCS	I am satisfied with the progress I have made towards meeting my goals for the development of new skills.	0.689	0.829
q6_JCS	I am satisfied with the support I receive from my clients.	0.549	0.741
Number of Items		6
Total Variance Explained for Construct		65.60
Cronbach’s Alpha for Construct		0.893

Note: Principal component analysis with varimax rotation and Kaiser normalization.

,

Table 4. EFA results for the digital technology support (DTS) construct.

Variable Label	Variable	Communalities	Factor Loadings Factor 1 Digitalization (DT)	Factor Loadings Factor 2 Digital Skills Acquisition Need (DSAN)
q1_DTS	In my country, there are systems that allow citizens to access and use electronic services in a secure and reliable way.	0.569	0.692
q2_DTS	In my country, people are well informed about the opportunities in the digital labour market.	0.537	0.719
q3_DTS	In my country, there are programs that support regional coworking spaces.	0.560	0.729
q4_DTS	In my country, the education sector should give more priority to the acquisition of digital skills.	0.907		0.952
q5_DTS	Government support is sufficient for my country to be considered a digitized country.	0.690	0.782
Number of Items of Construct		5
Total Variance Explained for Construct		65.252
Cronbach’s Alpha for Construct		0.608

Note: Principal component analysis with varimax rotation and Kaiser normalization.

,

Table 5. EFA results for the government programs (GP) construct.

Variable Label	Variable	Communalities	Factor Loadings Factor GP
q2_GP	Science parks and business incubators provide effective support for freelance activity.	0.576	0.759
q3_GP	There is an adequate number of government programs for freelance activity.	0.778	0.882
q4_GP	The people working for government agencies are competent and effective in supporting freelance activity.	0.713	0.844
q5_GP	Almost anyone who needs help from a government program for a freelance activity can find what they need.	0.791	0.890
q6_GP	Government programs aimed at supporting freelance activity are effective.	0.784	0.885
Number of Items		5
Total Variance Explained for Construct		72.839
Cronbach’s Alpha for Construct		0.906

Note: Principal component analysis with varimax rotation and Kaiser normalization.

and

Table 6. EFA results for the cultural and social norms (CSN) construct.

Variable Label	Variable	Communalities	Factor Loadings Factor CSN
q1_CSN	The national culture is highly supportive of individual success achieved through personal efforts.	0.702	0.838
q2_CSN	The national culture emphasizes self-sufficiency, autonomy, and personal initiative.	0.790	0.889
q3_CSN	The national culture encourages entrepreneurial risk-taking.	0.779	0.883
q4_CSN	The national culture encourages creativity and innovativeness.	0.819	0.905
q5_CSN	The national culture emphasizes the responsibility that individual (rather than the collective) has in managing his or her own life.	0.591	0.769
Number of Items		5
Total Variance Explained for Construct		73.616
Cronbach’s Alpha for Construct		0.908

Note: Principal component analysis with varimax rotation and Kaiser normalization.

below, show the EFA results for each construct (factor) of the conceptual model. The tables show the variables and their labels, the number of items, the communalities values, the factor loadings, the total variance explained and Cronbach’s alpha for the construct (factors).

3.6.1. Job and Career Satisfaction

Table 3 shows the EFA results for the construct of job and career satisfaction (JCS), based on the set of measured variables.

According to the results of the exploratory factor analysis for the construct of job and career satisfaction, the values of the communalities for all items are high enough, and the factor loadings are also large enough (>0.5). Based on the EFA, the one factor that defines the construct of job and career satisfaction is formed. The factor given also explains the sufficient extent of the total variance (65.60%). The result is consistent with a measurement scale developed by [96], in which job and career satisfaction is treated as one factor. Cronbach’s alpha coefficient for the factor (JCS) is 0.893. Since Cronbach’s alpha coefficient is above the required minimum value of 0.60, it can be said that the dimension of job and career satisfaction with the one factor formed was measured with sufficient reliability.

3.6.2. Digital Technology Support

Table 4 shows the EFA results for the construct of digital technology support (DTS), based on the set of measured variables.

According to the EFA results for the DTS construct, the values of the communalities for all items as well as the factor loadings correspond to the criteria (>0.5). On the basis of the EFA, two factors based on the set of measured variables were formed: Digitalization (DT) and Digital Skills Acquisition Need (DSAN). The construct explains the sufficient extent of the total variance (65.252%) and Cronbach’s alpha coefficient is 0.608, indicating that the DTS dimension was measured with sufficient reliability.

For contextual reasons and for the purposes of our research, which focuses on the perception of the current environment and the ecosystem in which the freelancers work, the factor Digital Skills Acquisition Need is excluded for further analysis.

3.6.3. Government Programs

With regard to the construct government programs (GP), we excluded varibale q1_GP from further consideration, based on the analysis of communalities. Specifically, the communality of variable q1_GP (a wide range of government assistance for freelance activity can be obtained through contact with a single agency), was less than 0.5. For all other variables, the values of the communalities and the factor loadings meet the criteria (>0.5), so it was not necessary to exclude another variable from further consideration. Table 5 shows EFA results for the construct of government programs (GP), based on the set of measured variables.

Based on the EFA results, only one factor defines the construct GP. The factor formed also explains the sufficient extent of the total variance (72.839%). The result is consistent with a measurement scale developed by Global Entrepreneurship Monitor, which treats government programs as one factor. Cronbach’s alpha coefficient for the factor GP is 0.906, which is above the required minimum value (0.60) and indicates that the dimension GP was measured with sufficient reliability.

3.6.4. Cultural and Social Norms

The final table in this section, Table 6, presents the EFA results for the construct of cultural and social norms (CSN), based on the set of measured variables.

According to the results of the EFA, the one factor that defines the cultural and social norm construct is formed. The factor formed also explains the sufficient extent of the total variance (73.616%). The result is also consistent with a measurement scale developed by Global Entrepreneurship Monitor, which treats cultural and social norms as one factor. Cronbach’s alpha coefficient for the factor CSN is 0.908, (above the required minimum value (0.60)), which indicate that the CSN dimension was measured with sufficient reliability.

In the following section, the variables that form individual factors based on the EFA were used as the basis for implementing the CFA for the structural equation model.

3.7. Structural Equation Modelling (SEM)

In order to obtain the final results from SEM, the CFA was first carried out. In conducting the CFA, we followed the main criteria presented in

Table 7. Criteria for implementation of confirmatory factor analysis (CFA).

Criterion	Level of Acceptance
Indicator Loadings	>0.50
Indicator weight	positive
Statistical significance of the indicator loading and indicator weight	p < 0.05
Variance inflation factor	≤5; ≤10.0
Effect size	≥0.02

Source: developed according to [105,106].

and the results are shown in

Table 8. CFA results.

Construct	Variable	Mean	SD	Indicator Loading	Indicator Weight	p Value	VIF	Effect Size
JCS	q1_JCS	5.75	1.100	0.836	0.215	<0.001	3.019	0.180
	q2_JCS	5.92	1.113	0.882	0.227	<0.001	3.900	0.201
	q3_JCS	5.85	1.267	0.746	0.192	0.003	1.831	0.143
	q4_JCS	5.84	1.123	0.782	0.201	0.002	2.259	0.157
	q5_JCS	5.68	1.144	0.829	0.214	<0.001	2.333	0.177
	q6_JCS	5.88	1.068	0.742	0.191	0.003	1.690	0.142
DT	q1_DTS	4.40	1.537	0.686	0.330	<0.001	1.246	0.227
	q2_DTS	3.55	1.504	0.728	0.350	<0.001	1.306	0.255
	q3_DTS	4.33	1.437	0.699	0.337	<0.001	1.273	0.235
	q5_DTS	3.78	1.540	0.767	0.369	<0.001	1.389	0.283
GP	q2_GP	3.40	1.484	0.698	0.207	0.001	1.495	0.145
	q3_GP	2.76	1.469	0.840	0.249	<0.001	2.229	0.209
	q4_GP	2.52	1.313	0.829	0.246	<0.001	2.434	0.204
	q5_GP	2.59	1.362	0.880	0.261	<0.001	3.026	0.230
	q6_GP	2.65	1.325	0.845	0.251	<0.001	2.398	0.212
CSN	q1_CSN	3.07	1.703	0.820	0.227	<0.001	2.645	0.186
	q2_CSN	3.16	1.578	0.879	0.243	<0.001	3.268	0.213
	q3_CSN	3.01	1.569	0.884	0.244	<0.001	3.605	0.216
	q4_CSN	3.22	1.585	0.895	0.247	<0.001	3.857	0.221
	q5_CSN	3.40	1.702	0.771	0.213	<0.001	1.894	0.164

.

According to the CFA results in Table 8, all variables meet the above criteria. It was not necessary to exclude any of the observed variables from the CFA analysis. The indicator loadings are greater than the required condition (>0.5) and their values range from 0.686 to 0.895, indicating their appropriateness. The analysis of the indicator weights of the observed variables also showed that the weights of all these variables are positive and statistically significant (p < 0.05). The variance inflation factor (VIF) cheques the multicollinearity between independent variables. A rule of thumb is that VIFs of 3.3 or lower do not suggest multicollinearity in the model. A more conservative recommendation is that VIFs should be lower than 5, and a more relaxed criterion is that it should not be higher than 10 [105,107,108]. Within CFA, we also consider the size of the influence of the observed variables on the factors. The threshold of the effect of the observed variable is 0.02, if the value is below 0.02, the exclusion of such an observed variable from further analysis should be considered [109]. The results suggest that that all variables included reach the appropriate level according to the given criteria.

Table 9. Reliability and validity indicators of the formed factors.

Construct	R-Square	Adjusted R-Square	Composite Reliability	Cronbach’s Alpha	AVE	VIF	Q2
JCS	0.054	0.040	0.916	0.890	0.647	1.041	0.059
DT	-	-	0.812	0.691	0.519	1.556
GP	0.339	0.336	0.911	0.877	0.674	1.863	0.344
CSN	0.347	0.340	0.929	0.904	0.724	1.526	0.351

shows the reliability and validity indicators of the factors formed. According to the adjusted coefficient of determination (adjusted R-square) for the factor formed GP (government programs), we can see that 33.6% of the variance is explained by the perceived digitalization, while just over 34% of the variance of cultural and social norms is explained by the perceived government programs and the digitalization. 4% of the variance of JCS (job and career satisfaction) is explained by perceived government programs, digitalization and cultural and social norms. The result is to be expected, since the job and career success of a freelancer depends on a variety of factors, including external factors (environment) and internal factors, i.e., motivation, personality traits. In addition, all factors formed in the conceptual model meet the conditions regarding the measurement reliability measured by the Composite Reliability indicator (>0.7) and Cronbach’s alpha coefficient (>50%). Furthermore, the average extracted variance for formed constructs (generated factors) is between 51.9% and 72.4%. All factors in the model fulfill the condition of average variance extracted (AVE > 0.5), which indicates that the measurement instrument had good convergent validity that the individual statements (questions within the measurement scale) were understood by the respondents as it was planned according to the questionnaire design. Furthermore, the collinearity of the factor formed with the VIF indicator was investigated. As mentioned before, according to [96], the recommended limit value is 3.3. All factors of the model fulfill this condition, which indicates that multicollinearity is unproblematic (VIF indicators are between 1.041 and 1.863). On the basis of the results, we can confirm that the formulated factors of the conceptual model are appropriate with regard to the indicators of reliability and validity.

Table 10. Correlation matrix between constructs of the SEM model.

Factor	GP	CSN	DT	JCS
GP	(0.821)
CSN	0.570 ***	(0.851)
DT	0.581 ***	0.429 ***	(0.721)
JCS	0.381 **	0.168 **	0.167 **	(0.804)

Note: correlations are significant at *** p ≤ 0.01; ** ≤ 0.05.

shows the correlation matrix between the constructs (formed factors) included in the model. The values of the square root of the AVE coefficient are shown on the diagonal of the correlation matrix [106]. The values of the square root of the AVE coefficient in our case are in all cases higher than the correlation coefficients (r), which confirms that the conceptual model corresponds to the discriminant validity (discriminant validity shows the unidimensionality between constructs (each observed construct has a weak relationship to constructs other than the one to which it belongs) [110]). Furthermore, the existing statistical characteristic of correlation coefficients confirms the nomological validity (nomological validity check the extent to which the predictions of the conceptual model formed and the constructs associated with it are theoretically grounded; to meet this condition, the correlation coefficients of the constructs are expected to be statistically significant [111]) of the model. In addition, most of the relationships between constructs are statistically significant and positive. The strongest link is between GP (government programs) and DT (digitalization) (r = 0.581, p < 0.01). Although, the weakest link is between GP (Government Program) and JSC (job and career satisfaction), it is significant (r = 0.167, p < 0.05).

Finally, the quality and validity of the model’s indicators, which are presented in

Table 11. Model fit and quality indices, including latent variables (factors).

Criterion	General Rule for Acceptable Fit If Data Are Continuous	Value
Average path coefficient (APC)	acceptable if p < 0.05	APC = 0.252, p < 0.001
Average R-squared (ARS)	acceptable if p < 0.05	ARS = 0.247, p < 0.001
Average adjusted R-squared (AARS)	acceptable if p < 0.05	AARS = 0.239, p < 0.001
Average block VIF (AVIF)	acceptable if ≤ 5, ideally ≤ 3.3	AVIF = 1.553
Average full collinearity VIF (AFVIF)	acceptable if ≤ 5, ideally ≤ 3.3	AFVIF = 1.497
Tenenhaus GoF (GoF)	small ≥ 0.1, medium ≥ 0.25, large ≥ 0.36	GoF = 0.398
Simpson’s paradox ratio (SPR)	acceptable if ≥ 0.7, ideally = 1	SPR = 0.833
R-squared contribution ratio (RSCR)	acceptable if ≥ 0.9, ideally = 1	RSCR = 0.999
Statistical suppression ratio (SSR)	acceptable if ≥ 0.7	SSR = 1.000
Nonlinear bivariate causality direction ratio (NLBCDR)	acceptable if ≥ 0.7	NLBCDR = 1.000

Source: developed according to [99].

, were also checked.

The results of the model indicators as a whole, indicate the adequacy of the model designed, and thus guide us to further use of SEM.

4. Results

After we had achieved a good fit in the measurement model, we added the structural path. The results are shown in Figure 2 and

Table 12. Hypothesis testing.

Hypothesis	Impact	Effect Size
H1.1	0.15 ***	0.339
H1.2	0.58 ***	0.191
H1.3	0.15 ***	0.041
H2	−0.01 n.s.	0.008
H3	0.49 ***	0.279
H4	0.13 **	0.025

Note: *** p ≤ 0.01; ** ≤ 0.05; n.s.—non-significant. Source: authors.

.

The standardized path coefficients show that the digitalization factor has a positive and a highly significant (p < 0.01) effect on government programs, cultural and social norms and job and career satisfaction, which confirms hypothesis H1.1, H1.2 and H1.3 In addition, government programs have a negative and not significant (p = 0.46) effect on job and career satisfaction, but they have a positive and significant (p < 0.01) influence on cultural and social norms, which confirms hypothesis H3. In addition, cultural and social norms have a positive and significant (p = 0.03) effect on job and career satisfaction, which confirms hypothesis H4.

Moreover, the results in Table 12, together with the impact factor and statistical significance, also show the effect size. As far as our results are concerned, the effect size is within the limits for all relationships (<0.02, according to [112]), except for hypothesis H2 (influence of government programs on job and career satisfaction). The effect size is too small and the impact is negative and not statistically significant. Consequently, we reject hypothesis H2.

5. Discussion

Under the influence of digital technologies, new work patterns are emerging in the labor market. Employment frameworks are moving away from the traditional standard model of full-time employment and adopting more flexible non-standard forms. Recent Gallup research has shown that 29% of US workers have an alternative work arrangement as their main occupation [113], while in Europe the number of online labor platforms has increased significantly in recent years. It is estimated that 11% of the adult population has ever used online platforms for some type of work service [114]. In order to make the most of digital technologies and drive digital progress, economies must be prepared to take over this transformation. The IMD World Digital Competitiveness (WDC) Ranking assesses the extent to which countries around the world are adopting digital technologies that impact government practices, business models and society. In the IMD ranking, Slovenia is in 31st place out of 63 countries in 2020 (one place higher than in 2019), while the leading countries are the United States, Singapore, Denmark and Sweden [115].

This indicates how national policies play an important role in preparing countries (through the formulation and implementation of programs and initiatives) to create value in the digital age. The level of development of an individual country has a major impact on the development of the business environment. Countries differ not only in their economic strength, but also culturally, politically, demographically, etc., so government policies must be adapted to the individual level of development of the national economy [76]. The digital strategy of the Republic of Slovenia was adopted for the years 2016 to 2020, and the subsequent one has not yet been introduced. Policy makers in Slovenia therefore need to recognize where their country currently stands and launch programs and plans that focus on the widespread use of digitalization. This includes adopting a broader ecosystem perspective.

The Digital Innovation Hub (DIH Slovenia) actively analyzes digital transformation in Slovenia. In 2016, the digital economy in Slovenia accounted for just over five percent of GDP and achieved only half a percent annual growth. As DIH Slovenia further reports, the share has been increasing since then, also boosted by the epidemic as an accelerator of digitalization. The growth potential for the share of the digital economy is considerable. It could increase by around 2.5 percentage points or 2.1 billion euros by 2025 [116].

It is therefore obvious that Slovenia has the potential to become the reference in the digital environment. Our research results show how the digitalization process has a positive impact on government programs as well as on the job and career satisfaction of freelancers who have adapted to the new working conditions that have emerged as a result. However, when looking at the results of the impact of government programs, they still do not significantly promote freelance activity in terms of job and career satisfaction. Although, according to the Slovenian experts, based on the results of GEM Slovenia in 2019, for the second year in a row, the most improvements were seen in the area of governmental entrepreneurial programs for SMEs and other incentives for entrepreneurship [76]. According to [117], the government needs to make more efforts to facilitate lifelong employability and to promote the transition from less stable forms of employment to more secure forms. The strong trend towards self-employment induced by digitalization also needs to be countered by a professional infrastructure that supports self-employment from the outset. More financial incentives for the training and further education of freelancers could improve their commitment and career progression.

With regard to the perception of cultural and social norms by freelancers in terms of job and career satisfaction, the results show positive and significant effects. Digitalization also has a positive and significant impact on social and cultural norms. This could be explained by the fact that digital technologies are used more frequently and awareness of them has increased, supported by Digital Economy and the Society Index (DESI), a complex index bringing together relevant indicators (connectivity, human capital, use of Internet services, integration of digital technology and digital publishing). The data for Slovenia 2020 show that Slovenia belongs to the group of countries with average performance (16th place out of 28 EU Member States). Based on pre-pandemic data, Slovenia improved its score in all five dimensions and made progress in integrating digital technologies [118]. DESI shows that Slovenia is on the level of the European average. However, current trends supported by our research also show room for improvement.

Moreover, as mentioned in the introduction, the COVID-19 pandemic has triggered a trend toward remote work, improved perceptions of freelancing as a career and a way of working, and raised many questions about the future workforce. At the same time, traditional jobs are becoming more unpredictable. There is more overlap between freelancers’ and traditional workers’ concerns about affordable health and welfare benefits, debt, and the ability to save. On the other hand, freelancers are taking the lead in learning new skills. The McKinsey Global Institute survey, states that 55% of freelancers across Europe and the US have attended training in the last six months to compete and survive in the marketplace, compared to 30% of non-freelancers (traditional employees) [119]. In this way, governments, companies, social institutions and other relevant stakeholders are responsible for developing an environment that ensures fair working conditions, social protection and even innovative tax systems for domestic and foreign market work. In doing so, they contribute to better working conditions for workers of all employment types and statuses, and ultimately to their job and career satisfaction, leading to higher productivity and welfare for society.

The research has several limitations that lead to recommendations for further research. First, the job and career satisfaction of freelancers in Slovenia is limited to the analysis of the effects of three factors of the external environment. In order to gain a broader insight and explanation of job and career satisfaction, the study should be extended to other factors of the entrepreneurial ecosystem (i.e., legislation, entrepreneurial education) and internal factors (i.e., motivation, social capital, human capital). The emphasis is also placed on the subjective outcomes, which could be extended with the objective ones (i.e., satisfaction with earnings). Secondly, the analysis is based on the adopted questionnaire. New scales for freelancers could be developed to provide a broader measurement of their characteristics and features. In addition, semi-structured interviews could be introduced to gain a deeper understanding of their career drivers and challenges. Third, our analysis is limited to freelancers. A comparison with other types of entrepreneurs or employees would give a better insight into their job and career satisfaction and the influence of the external environment. Furthermore, our research is limited to the perception of the external environment factors by freelancers, which could also be explored from a company perspective. Finally, there is still little literature on freelance activity in the national context and missing data on freelancers in Slovenia in general.

To conclude, the study highlights various aspects of freelancers from the entrepreneurial perspective in national contexts. One of the implications is that freelance work, like any entrepreneurial venture, is influenced by the national (entrepreneurial) ecosystem. The study points out the importance of government programs and social and cultural norms that need to be taken into account when explaining the specifics of freelance work. The results indicate that the environment, i.e., society, plays a special role in supporting freelancers in their work and career path. Policy makers should therefore make more efforts and take their needs into account when developing public policies to support this type of entrepreneurial venture. In this way, the challenges and benefits of freelance work should be further explored not only from a theoretical but also from an empirical perspective.