Unemployment, Inequality, and Occupational Stress: Mental Health Outcomes in Brazil (2012–2022)

4.1. Analysis of Historical Series The significant negative trend observed in the records of leaves related to severe stress reactions, with an average reduction of 629 cases per year, can be partially explained by changes in the work environment during the pandemic. The widespread adoption of remote work may have contributed to a decrease in stress associated with physical workplace factors, such as long commutes, direct exposure to supervisors, and interpersonal conflicts in corporate settings. However, it is important to consider that this reduction in records may also reflect practical difficulties in accessing health and social security systems during the pandemic, including the interruption of medical examinations at the National Institute of Social Security (INSS), which limited the formal diagnosis and official recognition of these conditions [9,10]. The absence of a significant trend in aggression cases, which remained relatively stable over time, can be explained by the fact that workplace violence dynamics may not have undergone relevant changes in terms of official records despite labor changes. Additionally, this category includes aggression committed by individuals external to the workplace, such as clients and service users, which may have remained stable even during the pandemic [19]. The significant reduction in cases of occupational exposure to risk factors, with an average annual decrease of 147 cases, directly reflects the impact of changes in working conditions during the pandemic. The transition to remote work temporarily eliminated exposure for many workers to hazardous workplace conditions, such as noise, chemical, biological, or ergonomic agents. Moreover, the shutdown of economic sectors, especially industries and in-person services, may have contributed to the decline in the registration of these illnesses. However, the interruption of medical examinations also played an important role, limiting the formal recognition of conditions related to in-person work environments during this period. The COVID-19 pandemic period brought substantial challenges to monitoring and recording occupational illnesses in Brazil. The temporary suspension of medical examinations by Social Security created a bottleneck in the official recognition of work-related illnesses. Furthermore, the shift to remote work profoundly altered traditional occupational exposures, shifting health risks to new contexts, such as musculoskeletal disorders associated with inadequate remote work setups, emotional stress from isolation, and difficulty reconciling professional and family demands in the same physical space. It is important to note that while the absolute numbers of occupational illnesses decreased during the analyzed period, these data do not necessarily reflect an improvement in working conditions but rather changes in the dynamics of recording and exposure. The gradual return to in-person activities may reveal a subsequent increase in the registration of these illnesses, especially in categories where risk factors remain high [1]. 4.2. Hypothesis Testing The Structural Equation Modeling (SEM) demonstrated the overall quality of fit, evidenced by the statistical indicators obtained. The CFI (1.00) and TLI (1.22) pointed to an ideal model fit, while the RMSEA (0.000) suggested that the model fits the data. The SRMR (0.017), in turn, indicated a low discrepancy between the observed and estimated data. Additionally, the chi-square test presented a p-value of 0.751, which reinforces the model’s adequacy by not rejecting the null hypothesis of good fit. Together, these results indicate that the proposed model is robust and provides a reliable basis for analyzing the formulated hypotheses. In our study, the backward relationships, where stress-related leave was modeled as a predictor of the socioeconomic indicators, yielded coefficients that were essentially zero. This result suggests that, within our sample and model specification, there is no evidence that stress-related leave exerts an influence on the unemployment rate, the Gini index, or the income rate. Furthermore, the standard errors for these paths were not estimable (reported as NA), indicating potential issues with model identification, possibly due to low variability or redundancy with other components of the model. These findings imply that the causal flow may predominantly operate from socioeconomic conditions toward work-related outcomes, rather than in the reverse direction. Future research should consider alternative model specifications or additional data to further explore and clarify the nature of these bidirectional relationships. Overall, the hypothesis analysis indicated a combination of confirmations and refutations, reflecting the complexity of the relationships between macroeconomic variables and different types of work-related leaves (Table 4). Hypotheses related to stress-related leaves (SL) showed greater consistency with initial expectations, highlighting the significant impact of Unemployment Rate (UR), Income Rate (IR), and adverse work conditions such as Aggression (AL) and Occupational Risk Exposure (OR). On the other hand, hypotheses concerning aggression-related leaves (AL) and occupational risk exposure leaves (OR) showed more divergent results, with some expected relationships not being confirmed, such as the role of UR and IR in reducing these outcomes. Notably, the Gini Index (GI) showed significant impact in some situations, such as in the increase of AL, but was not relevant for SL or OR, suggesting that social inequality may affect work indirectly or be mediated by other variables. These results underscore the importance of a multifactorial perspective when analyzing the work context, considering not only economic variables but also organizational and cultural factors that influence work dynamics [6,7,24]. Regarding stress-related leaves (SL), the results indicate significant relationships that support part of the initial hypotheses. The hypothesis that the Unemployment Rate (UR) increases stress-related leaves was confirmed, demonstrating a significant positive impact. This corroborates the literature that identifies unemployment as one of the main sources of stress, not only due to financial instability but also because of the psychological impact of social vulnerability situations, such as isolation and loss of occupational identity [8,17,18,19]. Conversely, the Gini Index (GI), despite being an important indicator of social inequality, showed no significant impact on stress-related leaves, refuting the initial hypothesis. This result suggests that while social inequality is relevant in various dimensions of health, its effects may be mediated by contextual variables such as working conditions, social support, and access to mental health services, which dilute its direct impact on occupational stress [22]. The Income Rate (IR), on the other hand, had a significant negative impact on stress-related leaves, confirming the hypothesis that better economic conditions can protect workers against stress by promoting greater stability and psychological security [6,24,26,42]. Stress-related leaves (SL) were significantly predicted by aggression-related leaves (AL) and occupational risk exposure (OR). The findings underscore the critical psychological impact of unsafe working conditions and workplace violence on stress-related outcomes [4,5,6]. Aggression (AL) showed a positive and significant relationship with SL, suggesting that experiences of interpersonal violence in the workplace contribute directly to elevated stress levels [2,4,12]. Similarly, occupational risk exposure (OR) emerged as the strongest predictor of SL, highlighting the importance of addressing unsafe environments to mitigate stress-related disorders. These results emphasize the interconnectedness of workplace conditions and stress-related outcomes, underlining the need for targeted interventions to promote safety and psychological well-being at work. These findings reinforce the need for targeted interventions to ensure safety and well-being in the workplace [2,4,5]. In terms of aggression-related leaves (AL), the results indicated interesting nuances. The hypothesis that the Unemployment Rate (UR) would increase aggression-related leaves was refuted, with UR showing a marginally negative impact. This finding can be explained by reduced exposure to work environments during periods of high unemployment, as interpersonal interactions decrease due to smaller teams or reduced activities in some organizations. Another possibility is that in unemployment contexts, workers who retain their jobs may adopt a more tolerant stance or avoid reporting aggression for fear of reprisals or losing their positions [19]. Conversely, the hypothesis that the Gini Index (GI) would increase aggression-related leaves was confirmed, showing that social inequality amplifies interpersonal tensions and conflicts, which can result in workplace aggression. Surprisingly, the Income Rate (IR) showed a significant positive impact on aggression-related leaves, refuting the initial hypothesis. This result may indicate that workers in higher-income positions have greater visibility and awareness of their rights, leading them to report aggression more frequently [21,22,23]. Alternatively, higher income levels may be associated with more competitive work environments, where interpersonal tensions and conflicts, including verbal aggression, are more common. In the context of occupational risk exposure leaves (OR), none of the hypotheses were confirmed. The Unemployment Rate (UR) had no significant impact, suggesting that high unemployment is not directly related to workplace safety conditions. This result may reflect the maintenance of activities in essential sectors such as construction and mining even in adverse economic scenarios [8]. The hypothesis that the Gini Index (GI) would increase risk exposure leaves was also refuted. This finding points to a more complex relationship between social inequality and unsafe working conditions, possibly mediated by legal regulations and public policies that mitigate the direct effects of inequality. Similarly, the Income Rate (IR) showed no significant impact, contrary to the expectation that better economic conditions would lead to greater investment in occupational safety. One possible explanation is that even in favorable economic contexts, some companies may prioritize profit over workplace safety, especially in sectors with high labor turnover [9]. However, in another context (e.g., European countries), cohort studies conducted in pre-pandemic periods suggest that more subtle or mediated reciprocal relationships might exist between work conditions and occupational stress [43,44,45]. Even though the present study did not detect significant effects of stress-related leaves on macroeconomic variables, previous findings indicate that deteriorating mental health among employees may, in some cases, prompt or intensify adverse psychosocial conditions at work. Furthermore, an analysis of cross-country differences showed that organizations with clearly defined risk-management policies, including designated roles for dealing with psychosocial stress, tend to have lower levels of psychosocial risk [46]. Such findings suggest that large increases in stress-related absences might encourage companies—particularly in well-regulated environments—to bolster preventive strategies, thereby creating a feedback loop where higher leave rates trigger improvements in health and safety measures. Thus, while the present results revealed minimal backward effects, other evidence implies that reciprocal influences could emerge more clearly under certain contexts, such as longer follow-up intervals, more detailed data on organizational practices, or when examining smaller units of analysis (e.g., at the company or sector level). When comparing the Brazilian and European contexts, it is important to note that structural and cultural differences may shape these potential feedback loops in distinct ways. In Brazil, where higher informality, underreporting, and economic inequality are prevalent, workers may be more hesitant to seek leaves due to fear of unemployment or job instability [7,17,21]. Additionally, insufficient oversight and limited access to formal occupational health services can mask the actual impact of psychosocial risks. Cultural and organizational factors—such as prioritizing productivity and short-term economic outcomes over employee well-being—may further undermine the implementation of mental health initiatives [11,16,23,24]. These findings emphasize the need for a closer look at the specificities of the Brazilian labor market and the interaction between economic, social, and organizational variables in fostering a safer and healthier work environment.