Entrepreneurs' stressors and well-being: A recovery perspective and diary study

The study addresses how entrepreneurs maintain well-being despite highly stressful, uncertain, and complex work contexts. It integrates the challenge–hindrance stressor framework with the stressor–detachment model to examine how cognitive (challenge) and emotional (hindrance) demands affect entrepreneurs’ well-being via recovery processes, especially psychological detachment during non-work time. The research highlights the need to understand short-term, within-person stress and recovery mechanisms for entrepreneurs, a domain dominated by between-person, cross-sectional studies.

Prior work shows entrepreneurs’ well-being relates to opportunity recognition, persistence, and firm performance, yet entrepreneurs face high cognitive and emotional demands. Traditional stressor–strain models often presume all stressors harm well-being, but emerging evidence suggests some stressors (e.g., long hours, complex tasks) can be appraised as challenges with potential benefits. The challenge–hindrance framework distinguishes demands that offer growth (challenge) from those that constrain goals (hindrance). Recovery research (stressor–detachment model) posits stressors spill over into non-work time, impeding psychological detachment and harming well-being. Perseverative thinking captures this lack of detachment, with problem-solving pondering (more neutral/constructive) and affective rumination (negative, worry-laden) as distinct forms. Entrepreneurship studies have rarely examined these mechanisms, especially at the day-to-day, within-person level. The authors develop hypotheses: H1 (cognitive demands positively relate to well-being), H2 (emotional demands negatively relate to well-being), H3 (cognitive demands’ positive effect is mediated by problem-solving pondering), H4 (emotional demands’ negative effect is mediated by affective rumination), tested at both within- and between-person levels.

Design: 12-day diary study with evening telephone interviews and next-morning surveys; multilevel structural equation modeling to separate within-person (day-to-day fluctuations) and between-person (average differences across entrepreneurs) effects. Sample: 55 entrepreneurs (“everyday” owner-managers/self-employed) primarily in Saxony, Germany. After exclusions for non-working/sick/missing data and requiring ≥3 lagged day-pairs per participant, the final dataset comprised 386 lagged day-pair observations (mean 7.02 per entrepreneur; range 3–9). Procedure: Each evening (post-work), trained assistants conducted ~5-minute standardized interviews to assess daily stressors and perseverative thinking. Each following morning, participants reported well-being upon waking. Actiwatch wrist devices objectively recorded sleep from 18:00 on the first Monday to 18:00 on the last Friday. Measures (5-point Likert scales unless noted):

• Well-being (WHO-5 adapted to momentary assessment; Cronbach’s α=0.87).
• Challenge stressors: daily cognitive demands (DISQ; 2 items; α=0.67), e.g., high concentration/precision, mentally taxing work.
• Hindrance stressors: daily emotional demands (DISQ; 2 items; α=0.60), e.g., emotionally draining work, display-rule dissonance.
• Mediators: problem-solving pondering (WRRQ; 5 items; α=0.81) and work-related affective rumination (WRRQ; 5 items; α=0.83). Controls: Between-person—age, gender, entrepreneurial experience (tenure leading firm), firm performance (4-item satisfaction index, α=0.90). Within-person—time (day sequence), previous-day morning well-being (autocorrelation), objective sleep efficiency (actigraphy; % of time asleep over time in bed) at both levels. Validity and analysis: CFAs supported discriminant validity among constructs (e.g., 5-factor model fit superior to single-factor; CFI≈0.87; RMSEA≈0.08). Multilevel SEM in STATA v14 used mean-centering at relevant levels to disentangle within- and between-person effects, full maximum likelihood, fixed slopes (random intercepts). ICCs indicated substantial within-person variance: well-being 51.08%, cognitive demands 71.25%, emotional demands 64.74%, problem-solving pondering 52.93%, affective rumination 67.25%. Model comparisons used −2 log-likelihood tests; mediation assessed via multilevel indirect effects (Preacher et al., 2010, 2011). Robustness checks included curvilinear tests, excluding evenings not in recovery mode, and expanded models allowing cross-paths (e.g., emotional demands to problem-solving pondering).

• Data: 386 lagged day-pairs from 55 entrepreneurs.
• Between-person direct effects (Model 2): • Cognitive demands positively predicted well-being (β=0.50, p=0.001), supporting H1a. • Emotional demands negatively predicted well-being (β=−0.34, p=0.005), supporting H2a.
• Within-person direct effects (Model 2): • No significant direct effects of cognitive demands (β=−0.01, p=0.847) or emotional demands (β=−0.03, p=0.587) on next-morning well-being; H1b and H2b not supported.
• Mediation via perseverative thinking (Table 3): • Cognitive demands → problem-solving pondering → well-being (within-person): a=0.12, p=0.012; b=−0.16, p=0.001; indirect effect ab=−0.02, p=0.047. Contrary to H3b’s expected positive mediation, the indirect effect was negative: higher day-specific cognitive demands increased evening pondering, which reduced next-morning well-being. No significant between-person mediation (H3a not supported). • Emotional demands → affective rumination → well-being (within-person): a=0.30, p<0.001; b=−0.13, p=0.036; indirect effect ab=−0.04, p=0.044, supporting H4b. No significant between-person mediation (H4a not supported; b=−0.31, p=0.098; ab=−0.16, p=0.099).
• Additional robustness finding: Emotional (hindrance) demands also increased problem-solving pondering (within-person a=0.12, p=0.014; between-person a=0.11, p=0.047); the indirect effect via pondering on well-being was negative and significant within-person (B=−0.02, p=0.050).
• Variance explained: Predictors accounted for 21% of between-person variance and 8% of within-person variance in well-being beyond controls.
• Results were robust to objective sleep efficiency controls; curvilinear tests of cognitive demands were non-significant.

Findings clarify how different stressors and recovery processes shape entrepreneurs’ well-being. Over time (between-person), cognitively challenging work relates to higher well-being, consistent with growth and goal attainment appraisals. However, day-to-day spikes in both cognitive (challenge) and emotional (hindrance) demands impede psychological detachment via increased perseverative thinking in the evening, leading to lower next-morning well-being. This highlights a recovery overload mechanism unique to entrepreneurs’ sustained high demands: on particularly demanding days, even problem-solving pondering becomes draining rather than resource-gaining. Emotional demands emerge as critical hindrance stressors, exerting negative effects via affective rumination. The study underscores the importance of distinguishing within-person (short-term) from between-person (aggregated) processes; mediation via impaired detachment appeared only at the within-person level, indicating micro-foundational recovery mechanisms can be masked in cross-sectional designs. Practical implications include prioritizing strategies that facilitate complete psychological detachment after work (e.g., absorbing leisure, mindfulness, exercise, socializing) and developing emotion regulation and conflict management capabilities to mitigate hindrance stressors.

The paper integrates the challenge–hindrance stressor framework with the stressor–detachment model to explain entrepreneurs’ well-being. It demonstrates that cognitive demands can be beneficial across entrepreneurs, while day-level increases in both cognitive and emotional demands undermine recovery through perseverative thinking, reducing next-morning well-being. The multilevel diary approach reveals micro-foundational recovery processes and cautions against generalizing across analysis levels. Future research should develop time-sensitive theory that simultaneously examines hedonic and eudaimonic well-being, investigate trait moderators (e.g., positive affect, self-efficacy, psychological capital, passion), and map entrepreneurs’ recovery strategies and work designs (e.g., breaks, weekends, vacations), combining subjective and objective health indicators.

• Self-reported measures for stressors, perseverative thinking, and well-being (typical in diary studies), limiting objectivity; future work should triangulate with partner/spouse reports and objective health metrics.
• Convenience sample of predominantly German entrepreneurs from Saxony; younger and more male than representative samples; generalizability may be constrained.
• Two-item scales for cognitive and emotional demands yield moderate reliability (α=0.67 and α=0.60); conservative estimates may attenuate effects.
• Lagged day-pair design supports directional inference, but reverse or longer-term recursive processes may exist (e.g., well-being influencing stressor appraisal), warranting extended time horizons.
• Although sleep efficiency was controlled, other unmeasured contextual factors may influence recovery and well-being.