1. Introduction
In Korea, policies for start-up support policy have been proposed as an alternative for solving problems of low economic growth and unemployment. Interest in entrepreneurship is also increasing in the field of culture and arts. Currently, about 40% of artists work as freelancers and self-employed workers [
1]. Similar trends are seen in other countries. In Australia, 80% of professional artists are self-employed [
2]. Based on the Current Population Survey between 2003 and 2015, Woronkowicz and Noonan [
3] found that many artists in the United States were freelancers or self-employed, unlike other industries; 37.4% of artists were self-employed while only 13.1% professional workers were self-employed overall. Artists prefer self-employment because they wish to work independently, and lack opportunities in the labor market [
3]. However, studies that examine artists as entrepreneurs are scarce. Moreover, few studies examine entrepreneurship as an alternative to employment, although entrepreneurship is considered an alternative solution for the lack of job opportunities.
We consider that push situations, such as low wages and lack of job opportunities in the labor market, have pressured artists in the field of culture and arts into choosing between employment and self-employment. Previous research on entrepreneurship focused on entrepreneurial motives [
4,
5,
6]. However, these studies cannot explain why individuals choose self-employment instead of other alternatives, or identify the factors influencing the decision-making process that drives entrepreneurial intention. According to Gohmann and Fernandez [
7], individuals can choose between self-employment, employment and unemployment by considering potential income and cost. Douglas and Shepherd [
8] studied how people’s attitudes to salary, independence, risk and work effort are related to career choice. Those who want greater independence and more economic opportunities prefer self-employment to employment. In addition to internal factors, external factors such as economic conditions also affect job choice. The entrepreneurial intention in culture and arts is the result of choosing between employment and self-employment, considering the internal and the external environment. To understand the career choice of cultural and artistic majors, we should study their employability, attitude towards careers, and psychological factors. The motivation for employment is determined by occupational working conditions and by internal and external environmental factors, which lead to predicted satisfaction from occupation. In the research model, we attempt to investigate the effect of employment inconsistency on working conditions, perceived personal competence related to employment, and employability variables, considering the internal and external environment and expected satisfaction with occupation.
This study investigates the factors that influence an individual to choose self-employment when choosing between employment and self-employment. This study examines the effect that job conditions and employability have on job satisfaction. Based on the theory of planned behavior (TPB), this study examines the effect that job satisfaction, the desirability of results, and self-efficacy have on entrepreneurial intentions. In addition, it examines whether heuristic factors influence the intentions of entrepreneurs in culture or arts, because the theory of planned behavior based on rational judgment weakly considers how psychological factors bear on the decision-making process.
4. Results
Partial least squares (PLS) [
49] was used to estimate the model. The PLS methodology for the measurement of structural equation modeling (SEM) was used for its reliability, validity and hypotheses tests. SEM enables the simultaneous examination of both the path (structural) and factor (measurement) in one model. PLS combines a factor analysis with linear regressions, and makes only minimal assumptions, with the goal of variance explanation. PLS is known to be suitable for analyzing relatively small samples of data. The analysis was run in two stages: firstly, to assess the validity and reliability of the measurement model, and secondly, to assess the structural model.
Before analyzing the data, exploratory factor analysis was conducted using principle component analysis (varimax rotation). The survey instruments and item loadings are given in
Table 3. A total of 48 items were analyzed for 10 factors. Factor loadings were 0.40 or more, and the eigenvalue was set to 1 or more. As the result of the factor analysis, the remaining question items, with the exceptions of some items, were reduced to 10 items with a cumulative variance value of 67.94%. The KMO (Keiser–Meyer–Olkin) value was 0.867 (
p < 0.001). The factor loadings for measurement items on the intended constructs were at least 0.450. Further, Cronbach’s alpha was greater than 0.6, which indicates that the scale had good internal consistency.
Table 4 reports Cronbach’s alpha, composite reliability, communality and average variance extracted (AVE). Construct reliability was assessed using composite reliability and a 0.7 benchmark was suggested [
50]. In addition, convergent validity was evaluated using the AVE measure. The recommended AVE-score and communality level was 0.50 [
51]. All composite reliabilities were over 0.7 and AVE and communality were above 0.50. Moreover, Cronbach’s alpha was greater than 0.8.
Discriminant validity was assessed. The square root of the values of the AVE should be greater than the correlation coefficient between the construct and other model constructs [
49].
Table 5 lists the correlation matrix with correlations among constructs and the square root of AVE on the diagonal. The square root of AVE was over the value of the correlations, confirming discriminant validity.
A bootstrap method was used to evaluate the path coefficient because PLS cannot show the significance of the path coefficient nor the confidence level. A structural model was tested using the loadings, the significance of the path coefficients and the
R2 value. A bootstrapping technique (500 re-samples) was applied to estimate the standardized path coefficients and
R2 values. The results are presented in
Figure 2 and
Table 6. The model explains 51.6% of the variance in entrepreneurial intention.
H1 states that employability positively affects job satisfaction. Major accomplishment (β = 0.186, t = 3.660), university brand (β = 0.135, t = 2.395) and labor market (β = 0.134, t = 2.613) were significant predictors of job satisfaction, however, confidence of employment (β = 0.049, t = 0.833) was not a significant predictor of job satisfaction. So, this hypothesis was partially supported. The gaps in work conditions significantly affected job satisfaction (β = −0.419, t = 8.091). The paths from job satisfaction to outcome expectation (β = −0.135, t = 2.297), self-efficacy (β = −0.162, t = 2.867) and overconfidence (β = 0.306, t = 5.151) were significant. The results imply that Hypotheses 2, 3a, 3b and 4 should be accepted. Paths from outcome expectations (β = 0.430, t = 8.692) and self-efficacy (β = 0.369, t = 7.675) to entrepreneurial intention were significant, indicating that H3c and H3d are supported. The paths from overconfidence (β = 0.024, t = 0.536) to entrepreneurial intention and from overconfidence (β = 0.059, t = 1.057) to outcome expectations were not significant, which indicates that H5a and H5b should be rejected. The path from overconfidence (β = 0.164, t = 2.425) to self-efficacy was significant, in support of H5c. Lastly, all control variables for entrepreneurial intention were not significant predictors: gender (β = −0.017, t = 0.391), training (β = −0.063, t = 1.542), majors (β = 0.033, t = 0.798).
We tested for mediation effects of outcome expectations and self-efficacy in the relationship between job satisfaction and entrepreneurial intention [
52]. We first estimated a model containing only the direct effects of outcome expectations on entrepreneurial intention. We tested whether the impact of the direct effect declined with the inclusion of an indirect effect through the mediator and the paths. As shown in
Table 7, outcome expectations and self-efficacy partially mediate the relationship between job satisfaction and entrepreneurial intention. We also applied the z statistic [
53] to confirm the mediation effects. For both mediator models, the
z value exceeded 1.96 (
p < 0.05). These results suggest that outcome expectation and self-efficacy had a mediating effect on the relationship between job satisfaction and entrepreneurial intention.
In addition, we analyzed the effect sizes (
f2) of outcome expectations and self-efficacy as mediators on the variables in the model. The effect size (
f2) was computed by noting the change in
R2 when a specific construct was eliminated from the model [
49]. If an exogenous construct strongly contributes to explaining an endogenous construct, the difference in
R2 between the variable-included model and the excluded model should be high. This difference leads to a high effect size. As
Table 7 shows, the
f2 of the outcome expectation and the self-efficacy variables exceeded 0.5 and had large mediating effects (Effect size—small = 0.02, medium = 0.15 and large = 0.35).
5. Discussion and Conclusions
The purpose of this study was to investigate the effects that expected job satisfaction has on entrepreneurial intention through self-efficacy and outcome expectations in the field of culture and arts. Expected job satisfaction was affected by perceived employability and expectation gaps in working conditions. In addition, we examined the influence of overconfidence, a heuristic factor, on the decision-making process of entrepreneurial intention.
Firstly, perceived employability had a positive effect on job satisfaction. The more that culture and arts majors experienced significant accomplishments and felt that their college brands were renowned, and the more job prospects they had, the higher was their expected job satisfaction. This finding is in line with studies by De Cuyper et al. [
24] and Gamboa et al. [
25] that perceived employability has a positive impact on job satisfaction.
Secondly, in the context of the results of Jusoh et al. [
26], the bigger the discrepancy in the expectation of working conditions (wage, work time, independence, autonomy and stability), the lower the expected job satisfaction. This finding means that the expected utility of employment is perceived to be low.
Thirdly, this study examined whether job satisfaction affects entrepreneurial intention with mediating effects from outcome expectations and self-efficacy. As suggested by Ajzen’s [
22] theory of planned behavior, outcome expectations and self-efficacy had a positive effect on entrepreneurial intentions. Further, expected levels of work satisfaction negatively affected the intention to start a business. This finding is in line with previous work [
27,
28,
29] that shows a negative relationship between job satisfaction and entrepreneurial intention. This result can be explained from various viewpoints. From the career-choice perspective, individuals who are satisfied with their job have a desire to maintain their job [
30] because doing so might provide higher expected utility than self-employment. From prospect theory [
54], if individuals’ criteria are not met, it is recognized as a loss. When the loss is perceived to be greater than the profit, the individual will pursue risk-seeking behavior. Low job satisfaction can be interpreted as a loss, which may encourage risk-seeking behavior. People take risky initiatives such as starting a business instead of choosing a stable job.
Finally, we examined the influence of overconfidence, which is a heuristic factor, on entrepreneurial intention. Job satisfaction had a positive effect on overconfidence. Job satisfaction is a positive emotion and can be interpreted as a success or a reward for self-determination. According to previous studies, overconfidence is strengthened when people have positive emotions and experience compensation or success [
40,
43]. Unlike previous studies [
34,
39] in which overconfidence leads to optimistic results in entrepreneurial intentions, no statistically significant effect was seen. It might be difficult for artists to consider positive outcome expectations for starting a business in the cultural arts field, because those who have experienced job dissatisfaction or lack of job prospects may start up a business for subsistence, and not for seeking opportunities for personal development. However, overconfidence affects self-efficacy positively. This result aligns with previous research findings that entrepreneurial intention may be affected by overconfidence [
34,
37,
38].
In terms of the relation between job satisfaction and self-efficacy, job satisfaction negatively influenced self-efficacy. However, job satisfaction positively influenced overconfidence, and overconfidence positively affected self-efficacy. That is, job satisfaction positively influenced self-efficacy through overconfidence. This contradictory result can be explained by cognitive bias and regulatory focus. Firstly, overconfidence is a cognitive bias [
34]. It is reasonable for rational judgment on self-efficacy for start-ups to be lowered because job satisfaction plays a role in lowering intentions to launch a start-up. As a bias, overconfidence affected by job satisfaction might make self-efficacy higher. Secondly, people change behavior with regulatory focus [
55]. People with a promotion focus want to start a business, and people with a prevention focus want to maintain their jobs [
56]. From this point of view, when people expect job satisfaction, those with promotion focus might follow a positive path from job satisfaction though overconfidence to self-efficacy, and those with prevention focus might follow a negative path from job satisfaction to self-efficacy.
This study has important implications for academics, practitioners and policy-makers. Firstly, we researched career choices and entrepreneurial intention in the field of culture and arts. Over the past several years, the culture and arts industry has continued to grow and attracted much attention from the national government owing to its high employment inducement effect. In addition, the emphasis on creativity during the Fourth Industrial Revolution suggests that the competence of culture and art majors and employees can be extended to other industries. Therefore, it is necessary to study culture and arts majors and their entrepreneurial intentions; however, prior studies were insufficient. This study contributes to the study of culture and arts by examining the career choices of culture and arts majors and the variables that influence them. Secondly, the purpose of this study is to investigate the effect of the internal and external environment on job satisfaction, and job satisfaction on career choice, by combining the social cognitive career theory and the theory of planned behavior, which explains entrepreneurial intention. While the social cognitive career theory focuses on career interest and choice process, this study found that alternative choices are possible. In addition, by including expected job satisfaction, job condition and perceived employability, our study covered the weakness of entrepreneurial intention based on the theory of planned behavior. Previous studies of entrepreneurial intention do not consider the push factors but only the pull factors of entrepreneurial activity. Finally, this study shows that psychological factors such as overconfidence influence the decision-making processes.
This study’s findings indicate that when an entrepreneurship education plan is established for the field of culture and arts, opportunistic entrepreneurship and subsistence entrepreneurship may occur, and the entrepreneurial intention can result from career choice. Policymakers should build social-security systems for subsistence entrepreneurship. This would encourage artists to become entrepreneurs.
The results of this study might be hampered due to the use of limited samples. To generalize this study’s findings, we need larger and more heterogeneous samples. Another limitation is that our study relies on analyzing limited impact factors of entrepreneurial intention. It is possible that other impact factors were not considered. Further, it is necessary to identify effective strategies or support policies for entrepreneurship.