Abstract
BACKGROUND:
The negative consequences of job loss on financial and psychological quality of life have been well documented.
OBJECTIVE:
This study evaluated the level of unemployed individuals’ embitterment and the mechanism by which negative life experiences increase embitterment by lowering the belief in a just world (BJW).
METHODS:
A survey was administered to 1,074 unemployed Korean adults who visited a regional Center for Employment and Welfare. Question items included the Korean version of posttraumatic embitterment disorder (PTED) self-rating scale, the BJW (personal and general) scale, negative life events, and sociodemographic characteristics.
RESULTS:
Only 45.9% of the participants were categorized as being in a ‘normal state,’ indicating that many of the unemployed were emotionally embittered. There was a positive direct effect of an increase in negative life events on embitterment. Negative life events also significantly and indirectly affected embitterment through personal BJW.
CONCLUSIONS:
Embitterment was prevalent among the unemployed participants and embitterment was a reactive emotion following personal downgrading, not so much related to conditions in the world at large. Thus, we recommend developing a screening program to detect extreme embitterment and an interventional program to help people better cope with emotional stress due to job loss. Efforts should be made to restore their violated expectation and trust that they would be treated in a just and fair way.
Keywords
Introduction
Detrimental consequences of job loss, both on financial and psychological quality of life, have been well documented. Job loss can be more impactful than divorce or separation and may result in mental health problems [1]. Embitterment can be a major negative psychological consequence as well [2]. Embitterment is a complex emotional state that is triggered after experiencing negative life events and is characterized as a nagging, burning feeling of “being let down, insulted, or being a loser, and of being revengeful but helpless” [3]. This emotional state can be pathogenic and may lead to an adjustment disorder termed post-traumatic embitterment disorder (PTED), which manifests as prolonged and disabling embitterment reactions and symptoms [4].
Negative life events, such as divorce, bullying, procedural injustice, and over-controlling supervision in the workplace [5, 6], can cause many different psychological reactions. Life events are risk factors, as they may be or may not be associated with injustice or downgrading. Though life events do not directly lead to embitterment, they can cause cases of embitterment when the events are associated with injustice or downgrading. For unemployed individuals who have experienced job loss in the past, the impact of experiencing negative life events on embitterment could be worse when they experienced them as injust, since job loss is one of the top-quadrant traumatic experiences that cause stress [7].
The “belief in a just world (BJW)” refers to a belief that “everyone gets what they deserve and deserves what they get” in life [8]. Quoting “the core pathogenetic mechanism of PTED is thus the violation of basic beliefs [4], ” Dalbert argued that embitterment could only be fully understood when a basic need to believe in justice is taken into account [8]. This construct from justice psychology suggests that embitterment may be higher if people with strong BJW experience injust events. Also, embitterment in response to negative events can be reduced in the presence of a stronger BJW since people who have a stronger BJW are likely to accept a negative event as a deserved one. Prior research that explored embitterment dynamics in a South Korean population via structural equation modeling [9] found that BJW had the most powerful indirect influence on embitterment. In this context, we posit that BJW can mediate the association between negative life events and embitterment in unemployed individuals.
Studies have distinguished between two spheres of justice: belief in a personal just world (personal BJW), in which one is usually treated fairly, and the belief in a general just world (general BJW), in which people in general get what they deserve [10–12]. These two constructs have different meanings and people tend to endorse the personal BJW more strongly than the general BJW. Prior research has indicated that personal BJW is positively associated with life satisfaction, self-esteem, or positive mood, and negatively predicts stress, depression, or negative mood [10–12]. General BJW is associated with victim derogation, punishment of offenders, harshness with the poor, and defense of the status quo [13–15]. Dalbert argued that personal BJW is more relevant to the context of embitterment [8]. Thus, the current study examined how personal BJW affects embitterment in unemployed individuals by considering both spheres of BJW.
Significant influential factors of individual responses to job loss include gender, age, education level, household income, previous occupational work types, unemployment route (involuntary job loss), duration of unemployment, personality, and economic difficulty [16–21]. These variables were evaluated as sociodemographic characteristics in the current study to identify a high embitterment group and were controlled for when multiple mediation analyses were conducted.
The aim of the present study was to identify the level of embitterment of unemployed individuals surveyed and to examine the influence of negative life events on embitterment by examining the potential mediation effect of two spheres of BJW using a multiple mediation model. We hypothesized that negative life event experience is positively linked to embitterment, general BJW and personal BJW mediate this link, and the magnitude of the mediating effect of general BJW and personal BJW differs.
Methods
Data collection
Data were collected at the regional Center for Employment and Welfare in Ansan-si, South Korea. Participants were unemployed adults from 19 to 65 years of age who attended unemployment benefit education. They were asked to anonymously complete a structured survey questionnaire. For quality assurance, all of the incomplete data were excluded and the collected data were modified again at the data coding stage. A total of 1,074 data points were used for the statistical analysis
Measures
Embitterment
The Korean version of the posttraumatic embitterment disorder (PTED) self-rating scale [22, 23] of the original 19 items was used to measure the embitterment of the unemployed using a 5-point Likert scale (“not true at all” to “extremely true”). Those with an average score below 1.6 were considered to be in a normal state, those with an average score above 1.6 and below 2.5 were considered to be in a chronic state of embitterment, and those with an average score of 2.5 or higher were considered to have a clinically significant intensity of reactive embitterment [22]. The 19 items included statements such as “That hurt my feelings and caused considerable embitterment,” “That I see as very unjust and unfair,” “That causes me to be extremely upset when I am reminded of it,” “That triggers me to harbor thoughts of revenge,” and “For which I blame and am angry with myself”. The inter-item reliability was excellent (Cronbach’s α= 0.96).
Negative life events experience
To measure the intensity of experiences associated with negative life events, 11 of 16 items in a negative life event list originally developed by Brugha [24] were used. The last five items were added based on recent social issues frequently covered by the news media in South Korea [25]. The questions included “being insulted seriously by or receiving an unfair treatment from those in an intimate relationship,” “experiencing a financial crisis,” “having an intimate relationship broken,” “being insulted or treated unfairly at a medical, legal, or educational institution,” and others. A 3-point Likert scale was used wherein larger values indicated more frequent negative experiences (1 = “never experienced;” 2=“experienced one time;” 3 = “experienced more than two times”). The inter-item reliability was good (Cronbach’s α= 0.85).
Belief in a just world (BJW)
BJW was assessed using two measures: the belief in a just world in general and the belief in a personal just world [26]. Six items were used to evaluate general BJW, including “I believe that, by and large, people get what they deserve,” “I am confident that justice always prevails over injustice,” and “I am convinced that in the long run people will be compensated for injustices.” For personal BJW, seven items including “I believe that, by and large, I deserve what happens to me,” “I am usually treated fairly,” and “I believe that I usually get what I deserve” were asked. To determine respondents’ agreement with these statements, a 6-point Likert scale (1 = “strongly disagree”; 6 = “strongly agree”) was used. Both inter-item reliabilities were also excellent (Cronbach’s α= 0.84 and 0.89, respectively). The two average scores for personal and general BJW measures were used as the values for manifest variables to represent two spheres of BJW.
Sociodemographic characteristics
The sociodemographic characteristics of respondents measured were gender, age, marital status, education level, monthly household income, previous occupational work types (permanent/temporary employment), unemployment duration (shorter than 3 months/3 to 6 months/longer than 6 months), and unemployment route (involuntary/voluntary).
Statistical analysis
Data were analyzed using the Statistical Package for the Statistical Analysis System (SAS) 9.4. The SPSS version 25.0 was used for a multiple mediation analyses. First, descriptive statistics were calculated. Second, A Student’s t-test and one-way Analysis of Variance (ANOVA) were conducted to examine whether there were statistically significant group differences of embitterment in terms of sociodemographic characteristics, respectively. The Scheffe test was performed for the post-test. To verify multicollinearity among variables, Pearson’s correlation analysis was conducted and variance inflation (VIF) was checked. Finally, a multiple mediation model analysis was performed utilizing SPSS Process Macro [27], which can verify multiple individual mediation effects of variables simultaneously. Bootstrapping was used to examine the significance of the mediation effect and to determine whether the two spheres of BJW were significant between negative life event experiences and embitterment. Bootstrapping is an advanced method that does not assume normality of the sampling distribution [28]. Indirect effects and the 95% confidence intervals (CIs) were estimated using 10,000 bootstrap samples and were considered statistically significant when the confidence intervals did not include zero. Potential confounders (sociodemographic characteristics) were controlled for in the multiple mediation model.
Results
Sociodemographic characteristics
A total of 1,074 of the study cohort included 562 females (52.4%) and the mean age of participants was 47.1 years (range: 19–65 years; SD: 13.59). Most respondents were in their 60s and 50s (26.5%; 24.5%) followed by those in their 40s (17.8%), 20s (17.2%), and 30s (14.0%). More participants were married/partnered (57.8%) than unmarried (26.4%), divorced/separated (11.7%), or widowed (4.1%). The majority of participants (44.0%) had a high school education, 39.8% respondents had a college degree, and 13.42% had less than a middle school education. Only 2.8% of participants had a graduate school education.
The average monthly household income of the participants was 2.27 million won ($1,848), and the most common monthly household income was 2.01–3.60 million won ($1,630–$2,920; 38.2%). The next levels were 2.00 million won ($1,622) or less (31.2%), 3.61–5.40 million won ($2,928–$4,380; 18.3%), and 5.41 million won ($4,387) or more (12.3%), respectively. In terms of previous occupational types of work, 50.3% of participants indicated they had permanent employment and 49.7% indicated they had temporary employment. The majority of participants were unemployed for less than 3 months (80.6%), 15.6% were unemployed for 3–6 months, and 3.8% were unemployed for more than 6 months. Most participants (95.4%) were willing to keep working but were asked or forced to quit.
Embitterment
The average PTED score of unemployed individuals was 1.64 (SD = 0.70). Fewer than half of the respondents (45.9%) showed an average score below 1.6, indicating a normal state, 44.6% showed an average score above 1.6 and below 2.5, indicating a chronic state of embitterment, and 9.5% had an average score of 2.5 or higher, indicating a clinically significant intensity of reactive embitterment [22].
Based on the results from the PTED self-rating scale, 76.0% of respondents selected “partially true,” “very much true,” and “extremely true” when asked if they had feelings of embitterment, 69.6% had stressful situations to think about over and over again, and 72.7% were extremely upset when reminded of it. In addition, 66.3% had experienced a situation that led to a noticeable and persistent negative change in their mental well-being, and 65.2% had a situation viewed as very unjust and unfair.
Differences in the mean embitterment scores among groups according to sociodemographic characteristics were verified using a Student’s t-test or ANOVA for the identification of groups with high embitterment. The results showed that the differences in average embitterment were statistically significant based on marital status, education level, and monthly household income; however, occupational characteristics (previous type of occupational work, unemployment duration, and unemployment route) were not statistically significant. Embitterment for participants within the divorced and separated groups was found to be significantly higher than that for participants in the unmarried or married (partnered) groups. Those who had a lower education than middle school had significantly higher scores of embitterment than the college graduates. There were significant differences in embitterment depending on monthly household income: the lower the income group, the greater the embitterment. Gender differences were not statistically significant and age was significant in the ANOVA but not in Scheffe’s post-test.
Negative life event experiences and BJW
The average score of negative life event experiences was 1.29 (SD = 0.29), where 1 refers to “never experienced,” 2 refers to “experienced one time”, and 3 refers to “experienced more than two times”. The most commonly experienced negative life event except for job loss was “significant economic crisis” (M = 1.47; SD = 0.68), followed by “death of family or close friends” (M = 1.38; SD = 0.62), “broken relationship with those in close personal relations” (M = 1.38; SD = 0.59), and “bad things (disease, injury, or violence) happened to those with whom I have a close relationship” (M = 1.33; SD = 0.59). In addition, the personal BJW (M = 3.60; SD = 0.81) scores were higher than the general BJW scores (M = 3.35; SD = 0.86).
Correlation analysis
Table 1 shows a correlation matrix of all variables. The independent variable of experience of negative life events was negatively related to general BJW (r = –0.193; p < 0.001) and personal BJW (r = –0.298; p < 0.001) and positively related to embitterment (r = 0.464; p < 0.001). The experience of negative life events was positively related to gender, age, marital status, and unemployment duration, while negatively related to education level, monthly household income, and previous type of occupational work.
Correlation matrix of measured variables (N = 1,074)
Correlation matrix of measured variables (N = 1,074)
Note: *p < 0.05; **p < 0.01; ***p < 0.001
General BJW was positively related to personal BJW (r = 0.626; p < 0.001) and negatively related to embitterment (r = –0.285; p < 0.001), showing a significant relationship with marital status. Personal BJW was also negatively related to embitterment (r = –0.451; p < 0.001), showing statistical significance only with monthly household income. The dependent variable embitterment was positively related to age and marital status, while negatively related to education level and monthly household income. These results are in line with the directionality of the study hypothesis, and the correlation coefficient results showed that the multicollinearity among independent variables was not suspected.
The results of the multiple mediation model are shown in Table 3. Each model was statistically significant, including model 4, which applied all of the study variables and had a power of 57.4% (F = 24.62; p < 0.001). The total effect (c), which was the direct path coefficient from the predictor (negative life events) to embitterment (β= 0.46; p < 0.001) in the absence of mediators, was significant, indicating that those having experienced negative life events more intensely had a higher level of embitterment.
Difference between mediation effects (N = 1,074; 10,000 bootstraps)
Difference between mediation effects (N = 1,074; 10,000 bootstraps)
Summary of multiple mediation analysis (N = 1,074; 10,000 bootstraps)
Note: *p < 0.05, **p < 0.01; ***p < 0.001.
A multiple mediation model with two mediators and a direct path from negative life events to embitterment is shown in Fig. 1. Considering the effects of mediators and controlled variables (sociodemographic characteristics), negative life events negatively influenced both general (β= –0.21; p < 0.001) and personal BJW (β= –0.29; p < 0.001) and positively affected embitterment (β= 0.26; p < 0.001). These results indicate that greater negative life events leads to a higher level of embitterment and lowers both levels of BJW in the general society as well as personal BJW. The influence of negative life events on mediators was greater on the path to personal BJW than on the general BJW. Furthermore, only higher personal BJW (β= –0.33; p < 0.001) was significantly related to lower levels of embitterment; general BJW (β= –0.02; p = 0.616) was not significant. These findings suggest that personal BJW, or the belief that the world will be fair to me, is solely influential on embitterment, while general BJW and personal BJW work at the same time. The influence of controlled variables was not statistically significant.

Results of the multiple mediation model. Note: *p < 0.05; **p < 0.01; ***p <0.001. All of the parameter coefficients are standardized results.
The results from bootstrapping showed that negative life events significantly and indirectly affected embitterment through personal BJW alone, with an indirect effect coefficient estimate of 0.0145 (95% CI = 0.0100, 0.0197). General BJW was not statistically significant, with an indirect effect coefficient estimate of 0.0005 (95% CI = –0.0016, 0.0030). This included zero within the confidence interval (see Table 3). These results show that the mediation effect of general BJW was not significant and only the influence of belief that the world will be fair to an individual is significant. The magnitude difference of the mediating effect of both mediators was 0.0140 (95% CI = 0.0201 0.0082), indicating that personal BJW had a greater mediation effect than general BJW.
The purpose of the current study was to evaluate embitterment in unemployed individuals in South Korea and to examine the influence of negative life events on embitterment with the potential mediation effect of two spheres of BJW using a multiple mediation model. The results identified embitterment as a prevalent emotion, with only 45.9% of participants categorized to be in a normal state. The average PTED score in these participants was 1.64 (SD = 0.70), which is higher than a study of a German population [2] that assessed reactive embitterment in unemployed individuals (1.24; SD = 0.92). A relatively high rate of a chronic state of embitterment (44.6%) was found in the current study, which may be characteristic of embitterment in those who are unemployed. The proportion of participants experiencing a clinically significant intensity of reactive embitterment was 9.5%.
For the PTED self-rating scale single items, one of the top five embitterment situations in the current study was that it led to a noticeable and persistent negative change in mental well-being. These findings are consistent with those from a German study [2]. Thus, it is important to consider key mental health indicators, such as depression and suicidal ideation, when evaluating embitterment in unemployed individuals. The experience of justice, which was also included in the top five situations, impacts embitterment as well.
Previous studies have found that sociodemographic characteristics are significant influential factors on the effect of unemployment on individual behavior [16–21]. The current study showed significant differences in embitterment based on marital status, education level, and monthly household income. These significant influential factors could not only be considered as risk factors, but also as resources which make a difference in how people cope with certain life events and therefore mediate whether for example job loss is experienced as injust or not. Specifically, embitterment in divorced and separated individuals was significantly higher than in unmarried or married (partnered) groups. Those who received less education than middle school graduates had significantly higher rates of embitterment than college graduates. In addition, monthly household income differences were also significant, such that the lower the income, the higher the level of embitterment. Although previous studies have shown an effect of occupational characteristics, such as unemployment duration, on an unemployed individual’s reaction [20], the occupational characteristic variables assessed in this study (previous type of occupational work, unemployment duration, and unemployment route) were not significant.
This study is similar to others [3, 4] in that it provides evidentiary support that negative life events are a significant factor in embitterment, such that the more often an individual experiences negative life events, the higher the level of embitterment. The present study is also in agreement with previous findings showing that individuals experience higher embitterment under specific conditions, such as patients with chronic kidney disease [29] or rheumatic disease [30], workers in highly stressful conditions [31, 32], German nurses [33], and survivors of a ferry disaster [34]. Our study could highlight the importance of social experience that downgrades herself/himself in the experience of embitterment emotion.
The mediating effects of BJW on embitterment were also identified in this study, since embitterment is related to BJW [35]. A prior study that explored embitterment dynamics in a South Korean population using structural equation modeling [9] detected factors that indirectly influenced embitterment not distinguishing between the two spheres of BJW. BJW had the most powerful indirect influence on embitterment, and BJW itself was reduced or increased by negative life events or social support, respectively. In the current study, the path from negative life events to BJW to embitterment was statistically significant, suggesting that it represents a core mechanism of embitterment. Thus, we provide empirical evidence that repeated negative life events undermine BJW, which in turn heightens the level of embitterment.
Most importantly, the current study verified a multiple mediation model by considering both spheres of BJW simultaneously within the relationship of negative life events and embitterment. The result gives an empirical proof that embitterment is a reactive emotion following personal downgrading and not so much related to difficult conditions in the world at large. Prior studies [10–12] distinguished belief in a personal just world (personal BJW) and the belief in a general just world (general BJW), reporting that the two constructs have different meanings. And, the current result shows importance of personal BJW in the context of embitterment. In other words, it is not only the negative life events affect embitterment, but BJW explains the important mechanism between the life events and the embitterment. And the injustice here is not how the whole world is injust, but how the world treats injust each individual, and a reactive emotion from that is embitterment.
These findings can be explained by Dalbert, who argued that personal BJW is more relevant to the context of embitterment [8] regarding people who tend to endorse personal BJW more strongly than general BJW. This considers the self-serving bias in general whereby one’s own discrimination will be more strongly denied than discrimination of one’s group [36] and fairness, reasoning in particular that one’s own superior fairness will be emphasized [37, 38]. In a self-presentation experimental study by Alves and Correia [39], Portuguese participants used higher personal BJW than general BJW when they were asked to positively self-present. On the other hand, participants used even lower personal BJW than general BJW when they were asked to negatively self-present. This pattern was also identified by Ganglof and Duchon [40] who polled 504 employed and unemployed individuals. These results also can be explained by the concept of “context dependency”, meaning that what individuals say to convey various images of themselves and the way others judge them is in general highly context dependent [41]. This is because expressing high versus low personal BJW is more socially valued than general BJW [42]. The social value, or normative character, of the expression of general BJW seems weaker and perhaps ambivalent [43].
There were several limitations in this study. First, the study was cross-sectional in design. Longitudinal studies are needed to better understand the long-term prospective development of embitterment in unemployed individuals. In addition, the study result has possibility of a common method bias for the set of measurements came from the same source. Although some significant influencing factors on embitterment were examined, other aspects, such as self-efficacy, trust, and political orientation could affect embitterment related to job loss [9]. Personal subjective evaluation and the cause of unemployment have also been identified as significant factors [44–46]. Thus, further studies that consider those factors and utilize practical research methods are required to better understand embitterment in unemployed individuals.
Conclusion
Embitterment was found to be a prevalent emotion in unemployed participants surveyed in this sample of South Koreans. Multiple mediation model verification demonstrated that negative life events not only directly affect embitterment, but also indirectly affect embitterment through personal BJW when both spheres of BJW (personal and general) work together. The result implicates that embitterment is a reactive emotion following personal downgrading and not so much related to difficult conditions in the world at large.
Our research also shed light on practical implications. First, given the prevalence of embitterment in the unemployed in our study, an effort for screening serious embitterment prevention can be made. Also, our study indicated the importance of efforts to figure out the ways how a worker has been fired from her workplace in terms of fairness and justice, and if her experience were considered shameful and downgrading, a psychological counseling and support would need to be provided. Finally, our study can be used to persuade policy makers why more attention should be paid to cultivate mutual respect and fairness at workplace in order to prevent a vicious cycle that a worker’s job loss can result in embitterment once it is experienced with insulting and disrespect from others, it makes harder for her to find a new job timely because of her negative mental conditions, which in turn aggravates her life conditions, reducing her personal belief for change and hope.
Footnotes
Acknowledgments
The National Research Foundation of Korea Grant funded by the Korean Government (No. 21B20151213037) and the Center for Happiness Studies at Seoul National University supported this work.
Conflict of interest
None to report
