Self-employed and stressed out? The impact of stress and stress management on entrepreneurs' mental health and performance

Entrepreneurs substantially contribute to economic growth and social stability, yet Germany has experienced a declining start-up rate since 2012. Self-employment is frequently associated with elevated stress levels stemming from financial uncertainty, time and deadline pressure, high workload, and complex decision-making responsibilities. Such stressors adversely impact entrepreneurs’ mental health and, by extension, business performance. Recent literature emphasizes that entrepreneurs’ well-being is directly linked to performance and success, making effective stress management crucial. This study investigates how work demands (quantitative and cognitive) and coping strategies (work extension, presenteeism, proactive coping, and positive reframing) relate to self-employed individuals’ mental exhaustion. The research is guided by: RQ1: To what extent does work stress affect the mental health of the self-employed? RQ2: To what extent do health-endangering coping strategies affect the relationship between workload and well-being? RQ3: To what extent do health-promoting strategies affect the relationship between workload and well-being?

Grounded in the Job Demand-Resources (JD-R) model, work demands (e.g., time pressure, cognitive load) are key predictors of mental exhaustion, whereas work resources support goal achievement and buffer stress. The JD-R posits two processes: high demands lead to exhaustion, and lack of resources diminishes motivation and engagement; resources can buffer the demands-exhaustion link. Extensions include personal resources (self-efficacy, optimism) and job crafting, suggesting proactive coping and reframing as intrapsychic strategies. Self-employed individuals often face high workloads, competitive pressure, responsibility, role strain, social isolation, and stakeholder-related stressors. Chronic stress can result in physiological, psychological, and behavioral consequences (e.g., burnout, anxiety, somatic complaints). Despite stressors, self-employment may offer autonomy, competence, control, self-realization, self-efficacy, meaning, and intrinsic motivation as protective resources. Coping among entrepreneurs tends to be more active and problem-focused (e.g., cause analysis, planning, seeking support) compared to avoidance; emotion-focused strategies like self-management and self-leadership are also used. Maladaptive coping such as work intensification and presenteeism can constitute self-endangerment: actions aimed at problem-focused coping but harmful to health. Based on this literature, the study hypothesizes: H11: Work demands increase mental exhaustion; H12: Work extension mediates the demands–exhaustion relationship; H13: Presenteeism mediates the demands–exhaustion relationship; H14: Positive reframing moderates (reduces) the demands–exhaustion link; H15: Proactive coping moderates (reduces) the demands–exhaustion link.

Design: Explanatory, quantitative pilot study testing a theory-derived model among self-employed persons in Germany. Data were collected via a fully standardized online questionnaire (TIVIAN) distributed primarily through LinkedIn and partner organizations between July–August 2023. A qualitative pretest (n=7) informed the estimated completion time (~10 minutes). Participants provided informed consent, and confidentiality and data protection were ensured. Sampling: Non-probabilistic, cross-sectional convenience sample (self-selection). Of 839 link accesses, 130 completed; after data cleaning and filters excluding non-employed/non-self-employed, n=117 remained. Sample characteristics: 49.6% male, 50.4% female; 50.4% with employees, 49.6% solo self-employed; mean age 42.26 years (SD=11.10; range 21–66); mean self-employment experience 11.95 years (SD=10.61). Measures: Work demands assessed by COPSOQ (German) scales for Quantitative Demands and Cognitive Demands; mental health by personal burnout/exhaustion (CBI) translated in COPSOQ; health-promoting coping by Stress and Coping Inventory (SCI) scales for Proactive Coping and Positive Reframing; maladaptive coping by Krause et al. instrument (Work Extension, Presenteeism); control variable Autonomy by the Scale for basic needs satisfaction in the workplace (German, 6 items, 5-point Likert). Reliability: Cronbach’s alpha values (with modified response formats) ranged from 0.72 to 0.92 across scales. Descriptive means (1–6 Likert): Quantitative Demands M=3.77 (SD=0.98, α=0.88); Cognitive Demands M=4.74 (SD=0.64, α=0.72); Demands Index M=4.29 (SD=0.64, α=0.83); Positive Reframing M=3.86 (SD=0.92, α=0.78); Proactive Coping M=3.97 (SD=1.00, α=0.87); Work Extension M=3.69 (SD=0.98, α=0.78); Presenteeism M=2.45 (SD=1.35, α=0.92); Mental Exhaustion M=3.22 (SD=0.96, α=0.89); Autonomy M=4.56 (SD=0.70, α=0.73). Analysis: IBM SPSS with Hayes PROCESS macro (bootstrapping). Hypothesis tests included hierarchical multiple linear regression (H11), mediation analyses (H12, H13), and moderation analyses (H14, H15). Significance level α<0.05 with 95% CIs. For H11, hierarchical regression controlled autonomy (Model 1), then added quantitative and cognitive demands (Model 2). For mediation, indirect effects with bias-corrected bootstrapped CIs. For moderation, interaction terms (centered variables) with autonomy as covariate.

• Reliability: All scales demonstrated acceptable to excellent internal consistency (α=0.72–0.92). • Descriptive: 88.0% often/always faced cognitive demands; 44.4% often/always faced quantitative demands. Mental exhaustion mean was 3.22; 16.2% reported often/always mentally exhausted. Presenteeism was relatively low (M=2.45). • Correlations: Autonomy inversely correlated with quantitative demands (r=-0.44, p<0.001) and mental exhaustion (r=-0.45, p<0.001). Presenteeism correlated positively with work experience (r=0.20, p=0.030). Proactive coping correlated negatively with demands (r=-0.22, p=0.016) and mental exhaustion (r=-0.31, p=0.001), and positively with autonomy (r=0.23, p=0.015). • H11 (Demands → Exhaustion): Quantitative demands significantly predicted mental exhaustion (Model 2: B=0.38, β=0.39, p<0.001), cognitive demands were non-significant (B=-0.03, β=-0.02, p=0.782). Model 2 overall: R²=0.33, adj R²=0.31, F(3,113)=18.16, p<0.001. Autonomy had a protective effect (Model 1: B=-0.62, β=-0.45, p<0.001; Model 2: B=-0.39, β=-0.29, p<0.001). • H12 (Work Extension mediation): Demands → Exhaustion overall effect B=0.44, β=0.30, p=0.01; Demands → Work Extension B=0.79, β=0.52, p<0.001; Work Extension → Exhaustion B=0.30, β=0.30, p=0.001. Indirect effect B=0.23, 95% CI [0.0866, 0.3944]; β=0.16, 95% CI [0.0570, 0.2616]. Direct effect after mediator non-significant (B=0.21, β=0.14, p=0.144). Full mediation supported. • H13 (Presenteeism mediation): Demands → Exhaustion overall effect B=0.44, β=0.30, p<0.001; Demands → Presenteeism B=0.71, β=0.34, p=0.001; Presenteeism → Exhaustion B=0.23, β=0.32, p=0.001. Indirect effect B=0.16, 95% CI [0.0478, 0.3069]; β=0.11, 95% CI [0.0316, 0.2024]. Direct effect remained significant (B=0.28, β=0.19, p=0.031). Partial mediation supported. • H14 (Reframing moderation): No significant interaction (product term B=-0.21, β=-0.12, p=0.131). Main effects: Demands positively predicted exhaustion (B≈0.47–0.48, β=0.32, p<0.001); Reframing negatively predicted exhaustion (B=-0.18, β≈-0.17 to -0.18, p≈0.026–0.034). • H15 (Proactive coping moderation): No significant interaction (product term B=-0.01, β=-0.00, p=0.968). Main effects: Demands positively predicted exhaustion (B=0.40, β=0.27, p=0.002); Proactive coping negatively predicted exhaustion (B=-0.17, β=0.18, p≈0.033–0.040).

Addressing RQ1, results confirm that higher perceived work demands—specifically quantitative demands such as workload and time pressure—are associated with greater mental exhaustion among self-employed individuals, aligning with JD-R predictions. Cognitive demands showed no direct effect, possibly because such demands may activate protective personal resources (e.g., self-efficacy, intrinsic motivation, meaning), attenuating their detrimental impact. Autonomy emerged as a strong protective resource, substantially reducing mental exhaustion. Regarding RQ2, self-endangering coping strategies fully (work extension) and partially (presenteeism) mediated the relationship between demands and exhaustion, indicating that high demands trigger maladaptive behaviors that harm mental health. This supports prior findings on self-endangerment in flexible work contexts. For RQ3, positive reframing and proactive coping did not moderate the demands–exhaustion relationship, yet both were associated with lower mental exhaustion as main effects; proactive coping was also inversely related to demands and positively linked to autonomy, suggesting it may reduce perceived demands and be more effectively deployed in autonomous environments. Overall, findings underscore the importance of reducing quantitative demands and discouraging self-endangering behaviors to protect entrepreneurs’ mental health.

This pilot study provides evidence that high quantitative job demands in self-employment are linked to increased mental exhaustion and that maladaptive coping behaviors—work extension and presenteeism—transmit part of this effect. Autonomy serves as a protective resource, while proactive coping and positive reframing are associated with lower mental exhaustion, though they did not moderate the demands–exhaustion link. The study contributes to entrepreneurship research by highlighting how specific coping strategies and job demands interact to affect self-employed individuals’ mental health and potential performance. Practically, entrepreneurship centers and support programs should include psychoeducation, stress management training, and awareness of self-endangering behaviors. Voluntary risk assessments, tailored mental health support, and health-promoting policies for solo self-employed contractors could help sustain well-being and business continuity. Future research should broaden the coping constructs assessed, incorporate entrepreneurial personality factors, examine JD-R dual pathways (resources, engagement, exhaustion), include industry classifications to identify vulnerable groups, and use longitudinal or mixed-methods designs to clarify causality and context-specific coping.

The non-probabilistic, self-selected online sample limits representativeness and generalizability; recruitment via specific networks may underrepresent certain groups. The cross-sectional design precludes causal inferences regarding the directionality of associations. Coping was assessed via a limited set of four strategies due to practical constraints, not capturing the full multidimensional construct. Modified response formats, while yielding acceptable reliability, may affect comparability with original instruments. Industry-specific data were not collected, potentially introducing bias. As a pilot study with modest sample size (n=117), findings should be interpreted cautiously and replicated in larger, diverse samples.