I wish I had a paid job or a business: a moderated effect of entrepreneurial potential on career choice regret

Abstract

Purpose

Despite the merit of extant studies on career decision regrets, they are not well integrated, are developed at different speeds and differ in focus. Specifically, they do not address an important question about the levels and antecedents of regret arising from choosing entrepreneurship instead of paid employment and vice versa. The authors adopted the regret regulation theory as foundation to examining the moderated effect of entrepreneurial potential (EP) on career choice regret (CCR) among employees and entrepreneurs.

Design/methodology/approach

The authors surveyed 721 employees and 724 entrepreneurs from a developing country and applied partial least squares-structural equation modelling (PLS-SEM) to test the hypotheses.

Findings

Employees regretted their career choice three times more when compared with entrepreneurs. However, the authors failed to conclude that the latter had three times better living conditions when compared with the former. EP negatively influenced the regret of being an entrepreneur in lieu of an employee while it positively influenced the regret of being an employee in lieu of an entrepreneur. The perceived opportunity cost of being a higher EP employee was three times greater when compared with that of being a lower EP entrepreneur. The effect of EP on CCR was mitigated or amplified by duration in the career, former career status, decision justifiability, and perceived environment's supportiveness.

Research limitations/implications

The design was cross-sectional, thus, the findings cannot be interpreted in the strict sense of causality.

Originality/value

The authors rely on an important yet often overlooked context of the choice between entrepreneurship and paid employment to test, clarify, and extend the regret regulation theory. The findings have novel human resource management and entrepreneurship policy implications.

Citation

Ephrem, A.N. and Murimbika, M. (2023), "I wish I had a paid job or a business: a moderated effect of entrepreneurial potential on career choice regret", Career Development International, Vol. 28 No. 5, pp. 518-535. https://doi.org/10.1108/CDI-09-2022-0249

1. Introduction

Career choice is socially and economically important (Hsu et al., 2019) but one of the life domains, after romance and education, in which individuals experience severe regrets (Budjanovcanin and Woodrow, 2022). Regret is beneficial from a learning experience perspective (Hwanwoo and Sturm, 2017; Pieters and Zeelenberg, 2007). However, intense career choice regret (CCR) might decrease the well-being of individuals, their self-image, work commitment, and productivity (Inman and Zeelenberg, 2002). Scholars largely focused on regret and anticipated regret of choosing one job position or an organization over another (Budjanovcanin and Woodrow, 2022; Reb et al., 2018; Willits and Franco-Watkins, 2021). Entrepreneurs' regrets about daily task decisions have also been examined (Arora et al., 2013; Markman et al., 2005). Few studies have measured entrepreneur's regret either as a self-blame for not having a regular job (Hsu et al., 2019); a self-blame for business failure (Quach et al., 2021); a disappointment for delayed entrepreneurial pursuit (Ahmed and Talat, 2021); or an anticipated misery that hinders entrepreneurial choice (Neneh, 2019).

These studies have merit; however, they do not address an important question about the levels and antecedents of regret arising from choosing entrepreneurship in lieu of paid employment and vice versa. Addressing this gap is critical for a better understanding of widely used regret regulation theory and evaluating the attractiveness of entrepreneurship versus paid employment. We examined the effect of entrepreneurial potential (EP) on CCR among employees and entrepreneurs. EP is the extent to which an individual exhibits the competencies, behaviors, and mindsets necessary to successfully start and sustain a business (Ephrem and Murimbika, 2023). Extant literature offers adequate evidence that most of the dimensions of EP are important drivers of career success among entrepreneurs and employees, albeit a few mixed findings (Ephrem and Murimbika, 2023; Spurk et al., 2019). However, there is a need to complement rather than substitute career success with CCR. This approach is consistent with the regret regulation theory that postulates (see second proposition) that individuals care not only about the outcome of the chosen option but also about what might have been the outcome of the best foregone option (Zeelenberg and Pieters, 2007). Within the regret regulation theory debates, arguably, an important question that remains unaddressed is the EP implication of being an entrepreneur versus being an employee for CCR.

This paper addressed this gap by focusing on multiple moderators (decision justifiability, duration in the career, former career status, and perceived environment supportiveness); thus acknowledging that the effect of EP on CCR depends on many contingencies. We show that decision justifiability is one of the most researched propositions of the regret regulation theory (McQueen, 2017); however, confusion remains on when does it mitigate regret. We also consider duration in the career and former career status, arguing that the effect of EP on CCR operates through career outcome and its cognitive appraisal which depend on time and previous choices, respectively. While we agree with the regret regulation theory that regret is a self-blame negative emotion (Pieters and Zeelenberg, 2007), we include environment supportiveness as a moderator. We argue that the outcome and cognitive processes that cause regret are subject to the level of environmental support towards the chosen option.

We do not only test but also clarify and extend the regret regulation theory. More specifically, we make the following contributions. First, by focusing on EP as a driver of the outcome that generates CCR, the paper clarifies the reasons why what must be regulated is not only regret, but also, and more importantly, the outcome that produces regret. Second, we provide empirical evidence that the drivers of the outcomes that generate regret have varying effects over time and depend on previous choices. Third, given that the effect of EP on CCR was positive among employees and remained so after controlling for the moderating effect of decision justifiability, we discuss when decision justifiability mitigates regret, when it does not and urge the regret scholars to reconsider the decision justifiability assumption. Fourth, we avail empirical evidence showing that the role of the environment should be reconsidered in the regret regulation theory. Subsequently, we reveal that a friendly business environment increases the opportunity cost associated with entrepreneurial choice for higher EP individuals. Thus, a policy contribution is made by showing that the collective goal of governments to enable the business environment does not necessarily correspond to entrepreneurs' goal of maximizing profits. An effective policy must balance entrepreneurship and paid employment promotion, or else, policy makers will also experience regret.

2. Theoretical framework and hypotheses development

2.1 Effect of entrepreneurial potential on career choice regret

Intuitively and consistent with its definition, EP is a positive driver of success in entrepreneurship (Ephrem and Murimbika, 2023). EP is also a positive driver of career success in paid employment. Employee's EP is an antecedent of both intrapreneurship and corporate entrepreneurship (Glinyanova et al., 2021). Therefore, it might be viewed as a strategic resource enabling organizational competitive advantage (Glinyanova et al., 2021). In return, higher EP employees are easily promoted.

Since EP drives career success of both entrepreneurs and employees, it is arguably critical to address an often-overlooked question, in the regret regulation theory debates, about which regret regulation strategy between being an employee or an entrepreneur is most effective for individuals with higher versus lower EP.

We postulated that EP has a positive effect on employees' CCR and a negative effect on entrepreneurs' CCR. Following the agency theory (Jensen and Meckling, 1976), it is neither certain that the EP of employees would be fairly acknowledged by the employers nor that higher EP employees would dedicate themselves to the organizations as if they were their own businesses. Some organizations have rules, systems, politics, and cultures that might buffer the employee's EP by implying low tolerance for positive deviant behaviors, such as intrapreneurship (Glinyanova et al., 2021). Therefore, higher EP employees could experience regret because of the escalation of entrepreneurial inaction, which diminishes EP. Importantly, higher EP individuals are satisfied not only by paid employment hedonic rewards but also care about the eudemonic attributions of entrepreneurship (Glinyanova et al., 2021).They might realize that they deserve their positions, thus bow less to hierarchical bosses. However, in some developing countries where toxic bosses may want to be revered, such an employee attitude can be interpreted as arrogance (Milosevic et al., 2020). As a result, more EP stifles the career progression of employees, increases their CCR, especially when the latent norm is mediocrity and when incompetent candidates are promoted and support each other. Therefore,

H1.

EP has a (i) negative effect on CCR among entrepreneurs and (ii) a positive effect on CCR among employees.

2.2 Moderators of the effect of entrepreneurial potential on career choice regret

Conceptually, decision justifiability is one of the most researched propositions of the regret regulation theory (McQueen, 2017; Zeelenberg and Pieters, 2007). However, except in a few cases, such as Inman and Zeelenberg (2002) who reported its attenuating role in the feeling of regret, it has rarely been tested on the ground. Therefore, question remains whether it works in real life and under which conditions. We rely on the seventh proposition of the regret regulation theory (see Pieters and Zeelenberg, 2007) to posit that decision justifiability increases the negative effect of EP on CCR among entrepreneurs and decreases the positive effect of EP on CCR among employees. Arguably, if an individual had no other viable option than being an entrepreneur or an employee, then regret should not occur regardless of the outcome and EP level. Therefore,

H2.

Decision justifiability (i) positively moderates the effect of EP on CCR among entrepreneurs and (ii) negatively moderates the effect of EP on CCR among employees.

The sixth proposition of the regret regulation theory assumes that individuals' tendency to anticipate regret positively depends on the expectation that they would learn about the outcomes of their choices quickly (see Pieters and Zeelenberg, 2007). However, the theory does not go further to elaborate how actual regret depends on the time from decision formation to the point of outcome awareness. Arguably, because the effect of EP on career choice outcomes depends on time, the latter has the potential to moderate the effect of EP on CCR.

We acknowledge that individuals who have just got a chance of being employed or starting a business might not report CCR, and that at a given age, the individuals become less productive (Zhao et al., 2021). However, in general, the individual's productivity and outcome in the chosen career could increase with time because of experience. Thus, the negative effect of EP on CCR among entrepreneurs might be stronger in the long term, while the positive effect of EP on CCR among employees might be weaker in the long term. Therefore,

H3.

Duration in the career (i) positively moderates the effect of EP on CCR among entrepreneurs, and (ii) negatively moderates the effect of EP on CCR among employees.

In this study, the environment refers to the organization and the business ecosystem in the case of employees and entrepreneurs, respectively. The regret regulation theory (see Pieters and Zeelenberg, 2007) does not acknowledge environmental supportiveness among the drivers of regret. Scholars seem to concur with the argument that regret is a self-blame negative emotion (Pieters and Zeelenberg, 2007; Zeelenberg and Pieters, 2007). However, the outcome that causes regret could be driven by the lack of environmental support in the chosen option. In this case, individuals might blame themselves for the choice of an option that may turn out to be linked to a bad environment. While in general the effect of EP on CCR might be positive among employees, its magnitude can decrease when organizations are supportive. Within the same geographical area, some entrepreneurs might be better supported and achieve better outcomes than others, something that can amplify the negative effect of EP on CCR. Therefore,

H4.

Perceived environment supportiveness (i) positively moderates the effect of EP on CCR among entrepreneurs and (ii) negatively moderates the effect of EP on CCR among employees.

The effect of job satisfaction on employee turnover has been researched (Choi et al., 2022). However, the reverse relationship is not equally addressed. Instead of focusing on satisfaction as an outcome construct, we focused on CCR which encompasses both the satisfaction with the chosen career and the imagined satisfaction with the foregone option. We examined former career status as a moderator of the effect of EP on CCR. Thus, we shed light on whether career switching is an effective regret regulation strategy for either employees or entrepreneurs regardless of or with consideration of their respective EP. This focus built from the ninth proposition of the regret regulation theory (see Pieters and Zeelenberg, 2007), which states that the intensity of regret depends on the information available about what would have happened if the individual made a different choice. One of the best ways to collect information about the foregone option would consist of trying it at some point in the past. We posited that individuals are rational; thus they often have a bad experience with the options they abandoned. Subsequently, the negative effect of EP on entrepreneurs' regrets could be greater among former employees whereas the opportunity cost associated with an employee's higher EP could be lower among former entrepreneurs. Therefore,

H5.

The effect of EP on CCR is positively (i) and negatively (ii) moderated by having tried the foregone option in the past, among entrepreneurs and employees, respectively.

The hypotheses were summarized in the conceptual model (see Figure 1).

3. Methods

3.1 Design, context, population, sampling, and data collection procedure

We did not adopt a longitudinal but a cross-sectional design, as we were interested in examining whether the hypothesized effects, which were yet to be established in the literature, could be supported at a specific point in time. Fortunately, the constructs are cumulative of past events. Through a step-by-step process, the changes in EP gradually influence CCR.

Data were collected on randomly selected entrepreneurs and employees in the Democratic Republic of the Congo (DRC). There is high reluctance by individuals to pursue entrepreneurship in the DRC, despite remarkable business opportunities coupled with high rates of unemployment and underemployment (Ephrem and Murimbika, 2023). Institutional failures and natural resources' curse in this context potentially harm the entrepreneurial spirit. However, we argued that the starting point to addressing this conundrum is to investigate whether individuals who have higher EP indeed launch businesses and report lower CCR when compared with employees. Responding to this question is important because it accounts for the conditions and extent to which entrepreneurship might be regarded as a credible career path against unemployment; otherwise, it is a better option than inactivity.

Individuals who owned a business and a paid job concurrently were excluded from participating in this study. There was no prior information on the best foregone option between full-time entrepreneurship and full-time paid employment against which hybrid entrepreneurs could be asked to compare themselves.

We secured a sample of 724 entrepreneurs and 721 employees which was higher than the a priori G* power minimum sample size of 652 in each case. The latter was computed by considering a linear multiple regression (fixed model, single regression coefficient, two-tailed test), the effect size of 0.02, the coefficient α of 0.05, the strict power of 0.95, and tested predictors equal to 11 (Franz et al., 2009).

The study considered all the employees regardless of the industry, the kind of contracts, and hierarchical positions. Entrepreneurs were neither considered in the strict sense of value creation and business formality nor the number of employees. The aim was to comply with the characteristics of paid employment and entrepreneurship in the surveyed country. We controlled for career formality, coded 1 for employees who held formal jobs or entrepreneurs who held formal businesses and 0 otherwise.

Thirty research assistants were involved in the online data collection through Qualtrics. They spread the questionnaire to respondents through a link circulated on social media platforms, and made the follow-ups by prompting the participants to complete the survey where partial responses were recorded.

63.3% of the surveyed employees had formal contracts. Although the businesses were small (the average number of employees per business = 11), nearly two-third of entrepreneurs were operating formal businesses. 56.6% of employees were working for SMEs, followed by the public sector and NGOs, which employed 25 and 17.5% of the labor force, respectively. Women were less represented in paid employment (about 59 women for 100 men) and even more so in entrepreneurship (44 women for 100 men). Most respondents had a bachelor's degree; however, employees were more educated compared to entrepreneurs. The average age was 38 and 35 years for entrepreneurs and employees, respectively.

3.2 Measurements

The constructs were measured on ordinal seven-point Likert scales. Only EP in both samples and perceived environment supportiveness in the sample of entrepreneurs, were measured formatively. Full items for all the constructs are provided as supplementary material. We did not develop new measures; rather, we adopted or adapted existing measures selected based on their proven psychometric properties, their suitability to our research's goal, and their validation in a developing country context. In the data analysis section, we reflect on the adaptations we made to existing scales and their benefits in line with mitigating common method bias threats. The adaptation of scales also consisted of rewording some items so that they could better reflect the research's goal.

CCR was measured as self-blame emotion occurring when an employee imagines that entrepreneurship would have been the better career option and vice versa for an entrepreneur. Five items adapted from Hsu et al. (2019) were used to measure CCR. Sample of them for entrepreneurs was: “the choice of entrepreneurship instead of paid employment is one of the biggest mistakes of my professional life”. Similar wording was used for employees, mutatis mutandis. We adopted an EP scale from Ephrem and Murimbika (2023), which was made of thirty-four items, one of them being: “how often do people commend you on original ideas and tangible outputs?” Decision justifiability was measured by three items adapted from Inman and Zeelenberg (2002), sample of which for entrepreneurs was: “I did not have any other option than being an entrepreneur”. Similar wording was used for employees, mutatis mutandis. The nine items, sample of which was: “DRC is a good place for someone to invest and conduct businesses”, used to measure perceived environment supportiveness (PES) for entrepreneurs were adapted from Mujahid et al. (2019). For employees, three PES items, sample of them was: “the organization for which I work supports me in several ways”, were adapted from Tumwesigye (2010).

Duration in the career was measured by the number of years the respondent had spent in entrepreneurship or in paid employment. Former career status was coded 1 if an entrepreneur (or an employee) was a former employee (or entrepreneur) and 0 otherwise.

We controlled for variables that could confound the hypothesized main effect by inflating or deflating the estimates (Masood et al., 2023). Educational level was measured on an ordinal scale ranging from 0 (below the state diploma) to 5 (PhD). Entrepreneurship education attendance was measured by a single item ‘’How often do you participate in entrepreneurial training?’’ and answers ranged from ‘’1-never’’ to ‘’7-always’. To measure age, the respondents reported how old they were. Sex was coded 1 for males and 0 for females. Entrepreneurial family background captured whether at least one of the parents is or was an entrepreneur, then coded 1 and 0 otherwise.

3.3 Data analysis techniques

3.3.1 Preliminary analyses

The constructs were measured subjectively, on the same questionnaire, and from the same respondent. They were prone to common method bias (CMB) (Masood et al., 2023). We undertook several procedural remedies to offset the CMB effects. We avoided neutrally worded scale points, even though the scales did indeed have a medium point corresponding to four. The neutral point potentially encourages laziness as some respondents would arbitrary and consistently select neutrality to satisfy the need to participate in the research rather than making the necessary effort to reflect on the questions (Mackenzie and Podsakoff, 2012). It also breaks the order of the answers, thus these can no longer be interpreted in terms of higher versus lower or more versus less. We included more than one response format (“totally agree” versus “totally disagree”, “always” versus “never”, “totally true” versus “totally false”). We used a cover message to separate questions of different constructs (Tehseen et al., 2017). We assured the respondents that we would report the findings anonymously. Thus, we boosted participants' confidence not to over–or under–represent themselves in responses. Prior to data collection, we piloted the questionnaire on thirty employees and thirty entrepreneurs to test ease of understanding and quickness to respond, and revised problematic items (Tehseen et al., 2017). We were fortunate that the majority of respondents held a bachelor's degree, implying they could reflect on the questions they responded to. Except for EP, the scales were short to minimize respondents' fatigue, which otherwise could lead to CMB (Mackenzie and Podsakoff, 2012).

We examined the effectiveness of these remedies by Harman's test (Tehseen et al., 2017). We entered all the constructs' items within a single Principal Component Analysis with “none rotation method”. The first factor captured 11.44 and 14.43% of the variance in the data for entrepreneurs (15 factors loaded) and employees (13 factors loaded) respectively. Thus, CMB was not of concern since no single factor emerged with at least 50% of the variance (Masood et al., 2023; Tehseen et al., 2017).

3.3.2 PLS-SEM to estimate the structural model

Although PLS-SEM and covariance-based (CB)-SEM are both effective, we preferred the former. We were interested in both explanatory and predictive power of the model; and we did not aim at merely testing but also extending the regret regulation theory (Sarstedt et al., 2022). Unlike CB-SEM, which supports factor measurement models, PLS-SEM supports both factor and composite models (Dash and Paul, 2021); further justifying our choice. We applied bootstrapping with 5,000 samples, a technique credited with improving the quality of PLS-SEM outputs (Sarstedt et al., 2022).

4. Findings

4.1 Reliability and validity of latent constructs

4.1.1 Reflectively measured constructs

All the outer loadings of reflectively measured constructs were significantly different from zero and were higher than the threshold of 0.708 (Sarstedt et al., 2022). Subsequently, the AVE of each construct was higher than the threshold of 0.5 (see Table 1), fulfilling convergent validity (Sarstedt et al., 2022). The constructs were internally consistent (see Table 1 in which rho_A are higher than or closer to 0.7). All the Heterotrait-Monotrait (HTMT) ratios were lower than a more strict threshold of 0.85 (Sarstedt et al., 2022) on each sample; thus discriminant validity was fulfilled.

4.1.2 Formatively measured constructs

We relied on redundancy analysis to test convergent validity of formative constructs. For each construct, we run a PLS-SEM regression in which the formative construct was the exogeneous variable and its best alternative single measure item was the endogenous variable (Hair et al., 2020). The best measure of perceived environment supportiveness on the entrepreneur sample was “The DRC is a good place for someone to invest and conduct businesses”. The nearly closer to the best alternative measure of EP was “I trust my ability to succeed as an entrepreneur or initiator of new things”. In any case, convergent validity was fulfilled because the estimate was significant (Sarstedt et al., 2022). None of the indicators of formative constructs were redundant. They scored a variance inflation factor (VIF) value lower than three (Hair et al., 2020). The outer weights were significant for all formative constructs.

4.2 Descriptive statistics

We reported but did not interpret the average because the ordinal scale is not continuous, lacks a meaningful zero point, and tells nothing about the interval between the responses (Tóth et al., 2020). For each latent construct, the number of times the respondents agreed (scale point equal to or higher than five) was computed and divided by the total number of responses recorded. The descriptive statistics were summarized in Table 1.

In 48.3 and 46.9% of cases, entrepreneurs and employees agreed that they had the EP, respectively. In 47.2 and 15.3% of cases, employees and entrepreneurs regretted their career choices, respectively. The difference between the two groups was significant. On average, entrepreneurs and employees had been in their chosen career for 6 and 7.5 years, respectively. The proportion of former employees (43.8%) among entrepreneurs was significantly higher when compared with the proportion of former entrepreneurs (28.5%) among employees. The decision to start a business was justifiable in 71.8% of cases. The decision to hold a paid job was justifiable in 42.3% of cases. Organizations were significantly more supportive (54.3%) of their employees when compared with the perceived business ecosystem supportiveness by entrepreneurs (43.3%).

The coefficients of correlation (see Table 1) supported the hypothesized effects, except for perceived environmental supportiveness which was positively and significantly correlated with CCR among entrepreneurs.

4.3 PLS-SEM results

The findings in Table 2 supported the hypothesis H1. However, EP caused three times more regret about being an employee when compared with the level of rejoice it provided for being an entrepreneur (see effect sizes-f²). Decision justifiability positively moderated the effect of EP on CCR among entrepreneurs and employees. This moderating effect was significant among the former, therefore supporting the hypothesis H2 for entrepreneurs only. Duration in the career negatively and significantly moderated the effect of EP on CCR among employees and entrepreneurs. Thus, the hypothesis H3 was supported for employees only. PES negatively, but not significantly, moderated the positive effect of EP on CCR among employees; however, it negatively and significantly moderated the negative effect of EP on CCR among entrepreneurs. Thus, the hypothesis H4 was not supported. The hypothesis H5 was fully supported.

The PLS-SEM literature does not adequately elaborate on how to interpret fit statistics, and there has been an increasing call discouraging their use for model assessment (Sarstedt et al., 2022). We relied on the VIF, the explanatory and predictive power to assess the structural model.Except for duration in the career and former career status in the case of entrepreneurs, the VIF values were lower than 3, indicating that multicollinearity was not a concern. The two collinear variables were not dropped because their inclusion was important in making comparisons with employees. The model's explanatory power was equal to 69.2 and 34% for entrepreneurs and employees, respectively (see Table 2). All the PLS-SEM indicators yielded lower prediction errors on both samples (see Table 3) compared to the errors of a naïve linear model; thus the model had a high predictive power (Sarstedt et al., 2022).

5. Discussion

5.1 Theoretical implications

5.1.1 Career choice regret levels

While previous studies reported that entrepreneurs would experience regret (Hsu et al., 2019; Markman et al., 2005; Quach et al., 2021), because of the escalation of commitment to their ventures coupled with higher failure rates (Oliver, 2018), this research showed different trends. Both employees and entrepreneurs reported lower CCR, although the CCR of the former was three times more when compared with the latter. We neither conclude that employees and entrepreneurs were better off nor that the latter had three times better living conditions when compared with the former. Lower regret might be explained by career choice justifiability, regret aversion and satisficing attitude (Pieters and Zeelenberg, 2007). Several factors might justify the difference in CCR levels between employees and entrepreneurs. First is the ease with which employees could engage in upward social comparison with successful entrepreneurs whose wealth is public and easily noticeable compared to the wealth of successful employees. Second, employees and entrepreneurs are intrinsically different, in that the former are more likely to complain about their situation whereas the latter are passionate about their choices. Entrepreneurs seem to view an opportunity in every situation including learning from poor outcomes, they are satisfied not only with hedonic attributions such as monetary rewards, but also with eudemonic attributions such as the joy of being their own bosses (Wiklund et al., 2019).Third, there seems to be a myopic admiration of entrepreneurship or a collective underestimation of its challenges (Brownell et al., 2021; Dodd et al., 2013). Fourth, the ease justifiability of entrepreneurial choice due to its accessibility when compared with paid employment, and fifth, the ease of reversing paid employment choice both matter in regret occurrence. For this last reason, this study did not support the regret regulation theory's suggestion that decision irreversibility promotes more regret (Pieters and Zeelenberg, 2007). Rather, it supported Gilbert and Ebert (2002) who argued that reversibility generates counterfactual thinking or the ease of reassessing one's decision, which in most cases promotes regret as a switching justification.

5.1.2 Effect of entrepreneurial potential on career choice regret

By focusing on EP as a driver of the outcome that generates CCR and consistent with the remaining issues in the regret regulation theory (Pieters and Zeelenberg, 2007), we contribute by availing evidence that what individuals regulate or must regulate is not only regret, but also, and most importantly, the outcome that produces regret. Arguably, when individuals focus on the expected outcome of their choices in relation to the expected outcomes of foregone options rather than focusing on regrets, they are more likely to consider the true drivers of regret (which in this case are drivers of outcomes), thus they make informed decisions. Regret is meaningless unless associated with the process and antecedents that generate the choice outcomes. Individuals are regret averse (Zeelenberg and Pieters, 2007); thus, they might fail to anticipate regret when they are making the choice but could find it easier to collect the information on the antecedents of the expected outcomes of various options. In so doing, both effectiveness and efficiency of regret management increase because individuals can reduce their tendency to unnecessarily pay regret insurance. Therefore, outcome maximization and regret minimization strategies must complement each other.

5.1.3 Moderators of the effect of EP on career choice regret

Decision justifiability was a most preferred regret regulation strategy by entrepreneurs compared to employees, although it was less effective for the former. We contribute to a better understanding of the regret regulation theory by arguing that the more a regret management strategy is chosen by individuals, as was the case for decision justifiability among entrepreneurs, the less it is discriminative and the less effective it is. The attenuating effect of decision justifiability on regret, often reported in previous studies (Inman and Zeelenberg, 2002; McQueen, 2017), was not supported for higher EP employees. Our finding contributes to a better understanding of the regret regulation theory by showing that the effect of decision justifiability on regret depends on the characteristics of individuals, and on the severity of regret. When the regret is extremely intense, as was the case for higher EP employees, individuals' disappointment distances them from considering the conditions in which they made their choices. Thus, decision justifiability decreases regret, contingent to the level of disappointment about the outcome of the chosen option in relation to the imagined outcome of the foregone option.

This study clarifies the role of time and experience with the foregone option in the experience of regret by showing that they are not only drivers of regret as suggested by the regret regulation theory (Pieters and Zeelenberg, 2007), but that they are also conditions on which the effects of the antecedents of regret depend. In the long term, higher EP employees reduced their CCR either because their potential was finally rewarded by organizations, or they realized it was too late to launch and grow businesses or they simply gave up on hope for any alternative option. In contrast, the negative effect of EP on CCR among entrepreneurs decreased with time, which could be explained by unfulfilled expectations within necessity entrepreneurship or the escalation of entrepreneurial commitment (Oliver, 2018). Having tried the alternative option, mitigated the positive effect of EP on CCR among employees and amplified the negative effect of EP on CCR among entrepreneurs. Arguably, career switching is an effective regret regulation strategy for both employees and entrepreneurs. However, even though the outcome in the second career option is poor, individuals might avoid regrets. Individuals have no guarantee that switching to previous career would not increase regrets, which in turn could lead to an endless cycle of regrets. Besides, switching potentially reduces regret as the individual resigns to their fate confidence that they have exercised alternate choices.

While this research agreed with the regret regulation theory that regret is a self-blame negative emotion, it makes a novel contribution, showing that regret depends on the environment. If both career environments were supportive to the same extent and were constant, individuals could blame themselves for making the choice of a wrong environment. However, as the environments constantly change and provide varying support, individuals might blame the environment, thus amplifying or leading to self-blame. Supportiveness of the chosen environment could implicitly prevent individuals from thinking about what might have happened in alternative options, thus, potentially reducing the room for regret.

5.2 Practical implications

If we consider the levels of CCR and the reasons discussed for their difference between employees and entrepreneurs, it follows that in countries with high unemployment rates, entrepreneurship should be carefully encouraged, not because it necessarily provides more monetary benefits when compared with paid employment, but rather for its non-monetary rewards and for being an alternative to inactivity.

It is a bad choice to be an entrepreneur when one has lower EP, and it is worse for one to be an employee with higher EP. Our research is consistent with studies that highlighted the challenges of promoting corporate entrepreneurship and business sustainability, both of which positively depend on employee EP, in developing countries (Tariq et al., 2015). Arguably, higher EP employees are likely to leave their organizations to start their businesses or quiet quite, not necessarily because of poor work conditions, but because of the eudemonic rewards of entrepreneurship. Therefore, there is a need for human resource managers to examine the preference of their employees between entrepreneurship and paid employment and the reasons behind such a preference, so that they avoid misallocating resources and attention to employees who cannot stay in the organization, no matter what. Alternatively, identifying higher EP employees who are passionate about entrepreneurship and aligning their compensations to the non-monetary benefits of entrepreneurship is a human resource management action to reduce turnover and ‘’locked at the job’’ phenomenon (Feenstra-Verschure et al., 2023).

Enabling the business environment increased CCR among higher EP entrepreneurs. Unlike studies that emphasized the relevance of friendly business ecosystems in promoting entrepreneurship (Mujahid et al., 2019), we suggest and consistent with Shane (2009) that the collective goal of governments to enable the business environment, might not correspond with the goal of entrepreneurs of maximizing profits. When the business environment is enabled, many individuals pursue entrepreneurship, therefore, the number of available business opportunities per individual decreases followed by an increase in competition which decreases the profit at the individual's level by buffering the EP dividend. Therefore, there is a need for balanced policy that promotes both entrepreneurship and paid employment.

6. Limitations and opportunities for new research

First, we did not examine what was regretted by entrepreneurs for not being employees or by employees for not being entrepreneurs. Examining the advantages of entrepreneurship that are most attractive to employees and vice versa is a credible future research question. Second, our speculative discussion on the reasons for an attenuating effect of experience about the foregone option on CCR suggests that more research is needed on how the outcome of the chosen option versus the imagined outcome of the foregone option justifies the intensity of regret in a cycle of choices. It might be that for the second time choice, regret depends more on the first-time regret than on the outcome of the chosen versus the imagined outcome of foregone options, and so for the third time choice. Third, for relevant reasons, we modified scales. While we are not afraid to have altered content validity, especially because we avoided cascading adaptation, thrived to keep the operational definition of each construct, and though we reported acceptable validity metrics, we must acknowledge that the overall validity of modified scales might have been affected (Heggestad et al., 2019), but hopefully not significantly. Fourth, we included control variables and moderators to mitigate endogeneity. However, there are other factors that might confound the main relationship. Moreover, if a valid instrument was available, it could be used to address the bidirectional relationship between EP and CCR. Thus, we cannot interpret the findings in the strict sense of causality. The doubt is exacerbated by a cross-sectional design, yet the phenomena investigated are dynamic. While this weakness is somehow compensated by large sample size (Schreurs et al., 2022), it is important that future studies adopt longitudinal or quasi-experimental designs to replicate our model, perhaps in a context different from the choice of entrepreneurship and paid employment.

7. Conclusion

Within the regret regulation theory debates, we stimulated the conversation which we hope will continue on the role of the environment, experience with the foregone option, time, and decision justifiability in the experience of regret. We illustrated the usefulness and practicality of regulating, not only regret, but also the outcomes that generate regret. We discussed the difference in the levels of CCR between employees and entrepreneurs; thus, we conclude by suggesting that, in countries with high unemployment rates, entrepreneurship should be carefully encouraged, not because it necessarily provides more hedonic benefits when compared with paid employment, but for its eudemonic rewards and for being an alternative to inactivity. The government's effort to enable the entrepreneurial ecosystem is collectively beneficial in the short run. However, in the long run, it discourages entrepreneurship by reducing the EP dividend for entrepreneurs and by hindering corporate entrepreneurship, thus increasing entrepreneurial choice regret. A most effective policy should thrive to balance entrepreneurship and paid employment promotion.

Figures

Figure 1

Conceptual framework of the research

Table 1

Correlation coefficients and descriptive statistics of the variables for employees (entrepreneurs)

	1	2	3	4	5	6
Career Formality (1)	1
Sex (2)	0.026 (0.003)	1
Level of education (3)	0.208** (−0.010)	−0.024 (0.048)	1
Duration in the career (4)	0.192** (−0.017)	0.159** (0.001)	−0.027 (0.004)	1
Former career status (5)	1.053 (−0.003)	0.033 (−0.049)	0.118** (−0.011)	0.096* (0.906**)	1
EEA (6)	0.072 (−0.006)	−0.007 (0.046)	0.127** (0.160**)	−0.102** (0.091)	0.053 (0.119**)	1
Age (7)	0.137** (0.040)	0.084* (0.025)	0.030 (−0.006)	0.300** (−0.051)	0.127** (−0.025)	0.059 (−0.037)
EFB (8)	0.019 (0.005)	−0.060 (0.044)	0.039 (0.090*)	−0.090* (−0.132**)	0.101** (−0.113**)	0.110** (0.001)
Career choice regret (9)	−0.166** (0.003)	0.028 (−0.101**)	−0.097** (−0.015)	−0.377** (−0.756**)	−0.321** (−0.751**)	0.106** (−0.172**)
Entrepreneurial potential (10)	0.021 (−0.045)	0.066 (−0.037)	0.010 (0.155**)	0.088* (0.408**)	−0.086* (0.423**)	0.026 (0.054)
Decision justifiability (11)	−0.007 (−0.037)	−0.030 (0.209**)	−0.017 (−0.056)	0.028 (0.130**)	0.104** (0.126**)	0.007 (0.127**)
Environment supportiveness (12)	0.036 (0.045)	0.004 (0.076*)	0.086* (−0.075*)	0.075* (−0.148**)	−0.020 (−0.147**)	0.092* (−0.102**)
Mean				5.991 (7.466)
SD				6.19 (6.37)
AVE
rho_A
Prop (X ≥ 5)
Z-STAT for mean difference test				0.362
Z-STAT for proportion difference test

	7	8	9	10	11	12
Career Formality (1)
Sex (2)
Level of education (3)
Duration in the career (4)
Former career status (5)
EEA (6)
Age (7)	1
EFB (8)	−0.030 (−0.022)	1
Career choice regret (9)	−0.121** (−0.025)	0.039 (0.067)	1
Entrepreneurial potential (10)	0.022 (−0.041)	−0.038 (−0.093*)	0.472** (−0.352**)	1
Decision justifiability (11)	0.065 (0.038)	0.077* (−0.149**)	−0.180** (−0.209**)	−0.036 (0.109**)	1
Environment supportiveness (12)	0.050 (0.057)	−0.009 (−0.051)	−0.107** (0.078*)	0.060 (−0.160**)	−0.136** (−0.022)	1
Mean			4.298 (2.960)	4.37 (4.66)	4.106 (6.141)	4.537 (4.074)
SD			1.737 (1.337)	1.123 (1.037)	1.451 (1.479)	1.283 (1.641)
AVE			0.737 (0.756)		0.634 (0.686)	0.627
rho_A			0.914 (0.920)		0.761 (0.777)	0.784
Prop (X ≥ 5)			0.472 (0.153)	0.469 (0.483)	0.423 (0.718)	0.543 (0.433)
Z-STAT for mean difference test			36.969	3.61	2.035	13.644
Z-STAT for proportion difference test			29.035	2.65	16.805	8.890

Note(s): ** Correlation is significant at 1%

*Correlation is significant at 5%

The figures in () correspond to entrepreneurs

EEA = Entrepreneurship education attendance rate

EFB = Entrepreneurial family background

SD=Standard deviation

Prop (X ≥ 5) reads as the proportion of the total cases (responses) for which individuals agreed, at least, with the questions they responded to for a specific latent construct. Italic values indicate difference of mean or of proportion between employees and entrepreneurs is significant at 5 % when the associated z-stat is greater than 1.96

Source(s): Compilation in SPSS 20

Table 2

PLS-SEM results

Constructs/Variables	Employees			Entrepreneurs
	Estimate	f2	VIF	Estimate	f2	VIF
Effect of exogenous variables
Decision justifiability (DJ)	−0.151***	0.033	1.064	−0.076***	0.015	1.267
Duration in the career (DCR)	−0.320***	0.134	1.187	−0.411***	0.099	5.683
Entrepreneurial potential	0.310**	0.089	1.683	−0.183***	0.027	4.098
Former career status (FCS)	−0.590***	0.102	1.060	−0.616***	0.044	5.751
Perceived environment supportiveness (PES)	−0.087**	0.010	1.115	0.086***	0.017	1.450
Moderating (interaction) effects
DCR	−0.112***	0.017	1.022	−0.073**	0.005	3.415
DJ	0.029	0.002	1.061	0.076***	0.018	1.404
FCS	−0.270***	0.023	1.577	0.173*	0.005	4.035
PES	−0.023	0.001	1.073	−0.127***	0.040	1.737
Effect of control variables
Sex	0.143**	0.007	1.035	−0.195***	0.025	1.094
Age	0.012	0.000	1.139	−0.059**	0.011	1.022
Level of education	−0.064	0.006	1.087	0.002	0.000	1.088
Entrepreneurial family background	0.034	0.002	1.050	−0.026	0.002	1.069
Entrepreneurship education attendance	0.041	0.002	1.160	−0.037	0.004	1.124
Job (business) formality	−0.082***	0.009	1.103	−0.022	0.000	1.010
R-square (p-value)	0.354 (0.000)			0.698 (0.000)
Adjusted R-square	0.340			0.692

Note(s): ***Path estimate is significant at 1%

**Path estimate is significant at 5%

*Path estimate is significant at 10%

Source(s): Compilation in SmartPLS 4

Table 3

Predictive capacity of the model

Employees					Entrepreneurs
Indicators	PLS-SEM		RMSE/LM(b)	(a)-(b)	PLS-SEM		RMSE/LM(b)	(a)-(b)
	RMSE (a)	Q2_predict			RMSE (a)	Q2_predict
CCR1	1.616	0.276	1.684	−0.068	0.960	0.427	1.039	−0.079
CCR2	1.527	0.222	1.597	−0.070	1.003	0.371	1.100	−0.097
CCR3	1.545	0.235	1.622	−0.077	0.897	0.411	1.008	−0.111
CCR4	1.491	0.231	1.566	−0.075	0.916	0.435	1.031	−0.115
CCR5	1.519	0.215	1.618	−0.099	0.847	0.425	0.938	−0.091

Source(s): Compilation in SmartPLS 4.0

Table A1

Scales

Entrepreneurial potential (EP) dimensions and items Adopted from Ephrem and Murimbika (2023)

Proactive Innovativeness‒Response format: Never (1) to Always (7)

1. How often has it happened to you to initiate and implement new projects that give satisfaction to people around you? 2. How often has it happened to you to design something that did not exist in your environment? 3. How often do people commend you on original ideas and tangible outputs? 4. How often do you find yourself looking for new ways to serve people? 5. How often are you the first to look for solutions to the problems around you?

Social skill‒ Response format: Totally false (1) to Totally true (7)

6. I am patient to listen and respond to others when they are talking 7. Most people are always willing to associate with me for the pursuit of a common goal 8. When I have no other choice than lying, I do, to achieve my purpose (reverse code) 9. My partners and friends rarely disappoint me 10. When I ask for something from other people, I always get it

Entrepreneurial behaviors prone to cognitive and heuristic biases‒Response format: Completely disagree (1) to Completely agree (7)

11. My first impression of things and situations is always right 12. I am attached to whatever I initiate as the mum is attached to her baby 13. When I am part of a team for which I am not among the best performers, I feel disturbed 14. I always make sure I have exhausted all the alternatives before I give up on a failing project, relationship, or deal 15. I love being an entrepreneur 16. I trust my ability to succeed as an entrepreneur or initiator of new things 17. I am good at managing situations involving pressure and complexity 18. Failure of any of my projects disturbs and disappoints me a lot 19. I easily worry about unfortunate and bad circumstances (reverse coded) 20. I would describe myself as someone who does what most of my friends, neighbors, and popular persons do (reverse coded)

Calculated risk-taking propensity‒Response format: Totally false (1) to Totally true (7)

21. I would describe myself as someone who is good at assessing the risks involved in any kind of project 22. I can recall several situations I benefited from recently for which people around me could not see anything good than chao 23. When something seems interesting to me, I choose to do it regardless of the risk involved 24. I recall several occasions I took big risks when I knew I could benefit

Financial literacy‒Response format: Completely disagree (1) to Completely agree (7)

25. I save part of my earnings regularly 26. I know how to look for money 27. I am always willing to look for productive assets in which I can invest when I have money in hands 28. When I have money, I always know what to do about it

Bricolage‒ Response format: Completely disagree (1) to Completely agree (7)

29. I always make decisions based on what I am seeing, what I have 30. When something is interesting to me, I immediately do it 31. There are several projects I have not yet launched because of resource constraints (reverse coded)

Management skill‒Response format: Totally false (1) to Totally true (7)

32. I would describe myself as someone who spends time wisely 33. All my projects and activities are planned, structured, and continuously controlled 34. My projects and activities are always budgeted

Career choice regret-entrepreneurs (similar items were used for the employee sample mutatis mutandis) Response format: Totally disagree (1) to Totally agree (7) Adapted from Hsu et al. (2019)

1. The choice of entrepreneurship instead of the paid employment is one of the biggest mistakes of my professional life 2. When I think of what I would have achieved in the paid employment, I blame myself for being an entrepreneur 3. If I were given a new career choice opportunity, I would not hesitate to choose the paid employment 4. If I were an employee, I would be living a better life compared to my current situation 5. If I were asked to choose for my beloved ones, I would recommend them the paid employment instead of entrepreneurship

Decision justifiability-entrepreneurs (similar items were used for the employee sample mutatis mutandis) Response format: Totally false (1) to Totally true (7) Adapted from Inman and Zeelenberg (2002)

1. I did not have any other option than being an entrepreneur 2. There was no organization that could hire me before I started-up my business 3. I could not decide otherwise than being an entrepreneur given the conditions in which I found myself at that moment

Perceived environment supportiveness-entrepreneur sample Response format: Totally disagree (1) to Totally agree (7) Adapted from Mujahid et al. (2019)

1. In my entrepreneurial pursuit, I am always supported by different partners, friends, and family members 2. The local community cares about me as an entrepreneur 3. The Government supports my business in several ways 4. My employees care about my business and support me 5. Banks and other financial institutions have put in place good loan policies that benefit me 6. My customers are kind 7. The competition among entrepreneurs is fair and regulated in my country 8. DRC is a good place for someone to invest and conduct businesses 9. My suppliers are kind

Perceived environment (organizational) supportiveness-employee sample Response format: Totally disagree (1) to Totally agree (7) Adapted from Tumwesigye (2010)

1. At the workplace, I am helped by my colleagues whenever I am in need 2. The organization for which I work cares about me 3. The organization for which I work supports me in several ways

Funding: This research received no funding

Conflicts of interest: The authors declare that they have no conflict of interest