The influence of communication climate, organizational identification, and burnout on real estate agents' turnover intention

Economic conditions are closely associated with real estate businesses. During a recession, real estate brokers have a high turnover rate, affecting the stable growth of the real estate brokerage industry. Moderate talent flow promotes healthy competition, but excessive turnover leads to imbalance and disarray. Thus, brokers’ turnover intention is an important academic and practical issue. Because individual data are nested in organizations, organizational effects on individuals may involve contextual effects, cross-level interactions, multilevel mediation, and moderation. When examining turnover intention, relationships between organizational and individual attributes must be considered. Organizationally, new brokers often face training needs, fierce competition, and peak resignation periods; job stress varies across offices. Communication climate shapes norms and can strengthen organizational identification through openness and participation, improving commitment to change and reducing turnover intention. Individual factors include heavy performance demands, overtime, and stress; when demands exceed capacity, stress and burnout rise, elevating turnover intention. Salary is also important given brokers’ performance-based pay. Prior work links job stress and burnout to turnover intention and indicates salary level as a determinant. This study focuses on Tainan City brokers within a volatile real estate cycle marked by supply–demand time lags, high professionalism, long payback periods, and income variability. Because industry actors cannot affect macrocycles, organizations can manage internal communication climate, job stress, burnout, and turnover intention to influence HR retention. Hierarchical linear modeling (HLM) appropriately analyzes nested data, capturing organizational and individual influences while avoiding independence violations. This study uses hierarchical linear moderated mediation modeling to examine whether communication climate (organizational variable) and job stress (individual variable) affect turnover intention via organizational identification and burnout, and whether communication climate moderates effects among job stress, burnout, and turnover intention. Findings aim to inform real estate managers’ intra-organizational management.

The review links communication climate, organizational identification, job stress, burnout, salary, and turnover intention, formulating hypotheses H1–H12. - Communication climate and organizational identification/turnover intention: Positive, participative climates foster identification and support for change, reducing turnover intention. Hypotheses: H1 Communication climate positively influences organizational identification. H2 Communication climate negatively influences turnover intention. H3 Organizational identification negatively influences turnover intention. - Job stress, burnout, salary, and turnover intention: Job stress reflects responses to job-related threats; prolonged stress leads to burnout (emotional exhaustion, depersonalization, reduced accomplishment). High job stress increases burnout and often turnover intention; however, in some public-sector contexts stable pay can retain staff despite stress. Salary and rewards generally reduce turnover intention. Hypotheses: H4 Job stress positively influences burnout. H5 Job stress positively influences turnover intention. H6 Burnout positively influences turnover intention. H7 Salary level negatively influences turnover intention. - Moderating and mediating roles of communication climate and burnout: Positive communication climates can weaken the stress→burnout link and help reduce turnover caused by stress and burnout. Hypotheses: H8 Communication climate moderates (weakens) the effect of job stress on burnout. H9 Communication climate moderates the effect of burnout on turnover intention. H10 Communication climate moderates the effect of job stress on turnover intention. - Multilevel mediation: Using HLM configurations (2→2→1 and 1→1→1), the study posits: H11 Communication climate influences turnover intention via organizational identification (2→2→1 mediation). H12 Job stress influences turnover intention via burnout (1→1→1 mediation).

Design and analysis: A hierarchical linear moderated mediation model (HLM) was employed to accommodate the nested structure (agents within branches) and to test multilevel mediation and moderation simultaneously. The modeling followed Baron and Kenny’s mediation steps and included cross-level interactions for moderation. Two mediation configurations were specified: (1) 2→2→1 mediation: Communication climate (Level 2) → Organizational identification (Level 2) → Turnover intention (Level 1). (2) 1→1→1 mediation: Job stress (Level 1) → Burnout (Level 1) → Turnover intention (Level 1). Communication climate (Level 2) was modeled as a moderator of Level-1 slopes (stress→burnout; stress→turnover; burnout→turnover). Models: Null models for turnover intention and burnout established significant between-branch variance. Step 1 tested effects of communication climate (L2) and job stress (L1) on turnover intention. Step 2 tested effects of job stress on burnout (with communication climate moderating the slope) and OLS-estimated communication climate→organizational identification (both L2). Step 3 tested the full model with mediators and moderators: turnover intention regressed on job stress, burnout, salary level, tenure (L1), and on communication climate, organizational identification, business model (L2), with random slopes for stress and burnout moderated by communication climate. Control variables included branch business model (direct sales vs franchise) at Level 2 and individual tenure at Level 1. Measurement: All constructs were measured on five-point Likert scales (1=strongly disagree to 5=strongly agree). Scales and sources: Communication climate (8 items: effective communication skills, precise communication; Madlock & Kennedy-Lightsey, 2010; Dewan & Myatt, 2008). Organizational identification (8 items: situated identification, in-depth structural identification; Rousseau, 1998; Dutton et al., 1994; Yang et al., 2014; Wu & Yu, 2021). Job stress (3 items; Parker & DeCotiis, 1983). Burnout (9 items: resource conservation, person–environment fit, expectation–reality discrepancy; Hobfoll & Freedy, 1993; Schaufeli & Buunk, 2003; Tung, 2019). Salary level (6 items: financial and nonfinancial compensation; Gerhart & Meiyu, 2015). Turnover intention (4 items; Li et al., 2017; Nadiri & Tanova, 2010). Sampling and data collection: Real estate brokerage chains in Tainan City (Taiwan Realty, Yung Ching Realty, H&B Housing, CTBC Real Estate, Sinyi Realty) across six districts were surveyed in person (Oct 1–Nov 30, 2020). 870 questionnaires were distributed to 87 branches; 615 were returned from 77 branches. After excluding branches with fewer than two responses and incomplete questionnaires, 558 valid responses from 73 branches remained (effective response rate 64.1%). Sample demographics: 55.4% male; mean age 38 (range 20–74); tenure varied, with 37.6% >5 years; education predominantly university/4-year technical college (54.7%). Aggregation tests: Communication climate and organizational identification were aggregated to Level 2 after assessing within-group agreement and reliability. ICC1 for both constructs was 0.99, and ICC2 was 0.99, indicating appropriateness and reliability of aggregation. Software/estimation: HLM was used for multilevel models with random intercepts and slopes where appropriate; OLS used for L2→L2 path (communication climate→organizational identification). Business model (direct sales=1, franchise=0) and tenure were included as controls.

- Null models indicated significant between-branch variance: turnover intention between-branch variance 0.228 (p<0.01), ICC≈0.241; burnout between-branch variance 0.073 (p<0.01), ICC≈0.161, justifying HLM. - Model 1 (baseline): Communication climate negatively related to turnover intention (coef = −0.784, p<0.01). Job stress positively related to turnover intention (coef = 0.212, p<0.01). - Model 2 (mediators): Job stress positively related to burnout (coef = 0.322, p<0.01), supporting H4. Communication climate did not significantly moderate stress→burnout (coef = −0.001, ns), not supporting H8. Communication climate positively related to organizational identification via OLS (coef = 0.542, p<0.01), supporting H1. - Model 3 (full moderated mediation): • Job stress→turnover intention direct effect not significant (coef = 0.018, ns), not supporting H5. • Communication climate moderated stress→turnover intention (interaction coef = −0.404, p<0.01), supporting H10; a more positive communication climate weakened the positive effect of stress on turnover intention. • Burnout positively related to turnover intention (coef = 0.568, p<0.01), supporting H6; communication climate did not significantly moderate burnout→turnover intention (interaction coef = 0.172, ns), not supporting H9. • Salary level negatively related to turnover intention (coef = −0.236, p<0.01), supporting H7 (higher salary, lower turnover intention). • Organizational identification negatively related to turnover intention (coef = −1.059, p<0.01), supporting H3. • Communication climate’s direct effect on turnover intention not significant in the full model (coef = −0.209, ns), not supporting H2. - Mediation: Communication climate influenced turnover intention via organizational identification (H11 supported). Job stress influenced turnover intention via burnout (H12 supported). Overall, H1, H3, H4, H6, H7, H10, H11, H12 were supported; H2, H5, H8, H9 were not supported.

The study addressed how organizational-level communication climate and individual-level job stress, burnout, and salary shape real estate agents’ turnover intention within a multilevel framework. Findings indicate that a supportive, participative communication climate strengthens organizational identification, which in turn reduces turnover intention (supporting the indirect pathway). While the communication climate’s direct effect on turnover intention dissipated in the full model, its role remained pivotal through enhancing identification and buffering the stress→turnover pathway. Job stress increased burnout, and burnout robustly predicted turnover intention, demonstrating that burnout is a central psychological mechanism translating stress into withdrawal cognitions. Salary level reduced turnover intention, underscoring the importance of both financial and nonfinancial compensation. Communication climate did not moderate stress→burnout nor burnout→turnover, suggesting that once burnout manifests, climate alone may be insufficient to alter its effect on turnover; however, climate did mitigate the effect of stress on turnover. These results refine theory by highlighting the primacy of mediating processes (identification, burnout) over direct effects and by specifying where communication climate exerts influence in multilevel systems.

This paper contributes by integrating organizational (communication climate, organizational identification) and individual (job stress, burnout, salary) factors in a hierarchical linear moderated mediation model to explain real estate agents’ turnover intention. Key contributions include: (1) demonstrating that communication climate lowers turnover intention indirectly via organizational identification; (2) establishing burnout as the mediator linking job stress to turnover intention; and (3) showing that communication climate moderates the stress→turnover link. Practically, management should cultivate an open, supportive communication climate to strengthen identification; monitor and reduce job stressors and burnout; and design competitive compensation packages to retain agents. Future research should explore additional organizational levers (e.g., appraisal systems, reward/punishment schemes), test moderation by salary level on stress/burnout effects, and examine different agent segments (e.g., presale vs pre-owned housing) and demographic subgroups to assess generalizability.

- Sample limited to real estate brokers dealing mainly with pre-owned houses in Tainan City; results may differ for presale agents or other regions. - Potential heterogeneity by demographics (gender, tenure, education) not fully explored. - Cross-sectional survey limits causal inference beyond modeled pathways. - Additional organizational variables (e.g., performance appraisals, reward/punishment systems) and potential moderators (e.g., salary level) were not included in the moderation tests and warrant future study.