Cross-Sectional Study Links Perceived Wage Gaps to Nurse Turnover Intention

A cross-sectional study of 792 hospital nurses in South Korea found that perceived wage gaps were associated with turnover intention mainly through burnout, perceived organizational fairness and organizational commitment, rather than through a direct statistical link. Nurses reported a mean current annual wage of 46.92 million South Korean won, while the annual wage they considered appropriate averaged 53.50 million won. The resulting mean wage-discrepancy ratio was 1.14, indicating that the average participant regarded an annual wage about 14% above her current pay as appropriate. Structural equation modeling estimated a significant indirect association between wage discrepancy and turnover intention through effort–reward imbalance, organizational justice, burnout and organizational commitment.

The data were collected in May and June 2026 through a nationwide registered online research panel. Eligible participants were nurses with at least three months of clinical experience who worked in tertiary hospitals, general hospitals or hospitals in metropolitan and nonmetropolitan regions. Nurses working in clinics and nurse managers at or above the level of head nurse were excluded. Recruitment quotas were based on six strata formed by crossing three hospital types with two geographic regions, with an overall target of approximately 800 participants. Of 800 people who completed the survey, eight were excluded because their responses were incomplete or failed predefined data-quality criteria, leaving 792 nurses in the analysis. Because recruitment used an existing nonprobability panel rather than random selection from a national sampling frame, the researchers cautioned that the sample should not be considered nationally representative.

Participants were relatively young and predominantly female, with a mean age of 29.78 years; 99.1% were women. Most were staff nurses, and 80.9% worked shifts. The sample was divided almost evenly between metropolitan and nonmetropolitan areas, while 38.8% worked in tertiary hospitals, 36.1% in general hospitals and 25.1% in hospitals. Compensation and employment conditions were the most frequently reported factor influencing participants’ choice of first hospital after graduation, cited by 47.6%. Meanwhile, 75.4% agreed that wage level affected their turnover intention, and 40.2% reported previous turnover experience. These responses show that pay was salient to the surveyed nurses, but the study did not determine whether stated intentions led to subsequent departures.

Annual wages varied by institution type and location. After adjustment for clinical experience, nurses in metropolitan hospitals had significantly higher current annual wages than those in nonmetropolitan hospitals. The raw mean was 48.24 million won in metropolitan areas and 45.58 million won in nonmetropolitan areas. Wages were highest in tertiary hospitals, followed by general hospitals and hospitals, with significant differences among the three types in both metropolitan and nonmetropolitan strata. The highest raw mean annual wage was 54.49 million won for nurses in metropolitan tertiary hospitals, while the lowest was 39.06 million won among nurses in nonmetropolitan hospitals. The metropolitan–nonmetropolitan difference remained significant in tertiary and general hospitals after clinical experience was controlled, but not in smaller hospitals. These are cross-sectional geographic and institutional contrasts rather than evidence that wage differences increased or decreased over time.

Wage discrepancy was measured through two self-reported figures: each nurse’s current annual wage and the annual wage she considered appropriate. The ratio was calculated by dividing perceived appropriate pay by current pay. A ratio of 1.00 represented exact agreement, a value above 1.00 meant that the nurse expected a higher wage, and a value below 1.00 meant the current wage exceeded that expectation. The measure therefore captures perceived compensation adequacy rather than an objective determination of underpayment or a standardized comparison with market rates. It may also reflect a nurse’s assessment of workload, clinical responsibility, experience and professional expertise, making it useful for examining reward adequacy but unsuitable as a direct measure of inequitable pay.

The structural model linked the discrepancy to two first-stage evaluations of work and rewards. Greater wage discrepancy was positively associated with effort–reward imbalance, the perceived failure of rewards such as pay, recognition, job security and career opportunities to match invested effort, with a standardized coefficient of 0.303. It was negatively associated with organizational justice, with a coefficient of −0.281. Both associations had p-values below 0.001. Effort–reward imbalance was positively associated with burnout, with a coefficient of 0.276, while organizational justice was negatively associated with burnout, with a stronger coefficient of −0.591. Organizational justice encompassed perceived fairness in reward distribution, decision-making procedures and interpersonal treatment. Together, effort–reward imbalance and organizational justice accounted for 63.8% of the modeled variance in burnout.

Organizational justice and burnout were in turn associated with organizational commitment. Perceived justice was positively associated with commitment, with a coefficient of 0.662, whereas burnout was negatively associated with it, with a coefficient of −0.226. Together, these variables accounted for 69.1% of the modeled variance in commitment. Burnout had the strongest direct association with turnover intention, with a standardized coefficient of 0.667, while organizational commitment had a significant negative association, with a coefficient of −0.245. Burnout and commitment together accounted for 76.2% of the variance in turnover intention within the sample. That figure is an in-sample model statistic and should not be interpreted as demonstrated predictive accuracy, proof of causation or a claim that the model can reliably identify which individual nurses will leave.

Decomposition of the modeled effects showed how wage perceptions were associated with intended turnover. The direct coefficient from wage discrepancy to turnover intention was 0.005 and was not statistically significant, with p = 0.833. In contrast, the combined indirect effect through effort–reward imbalance, organizational justice, burnout and organizational commitment was 0.226, with p = 0.001 and a bootstrap 95% confidence interval from 0.176 to 0.279. The total effect was 0.231, also with p = 0.001. Organizational justice and effort–reward imbalance were associated with turnover intention only indirectly through burnout and organizational commitment. A competing model that removed the nonsignificant direct wage-discrepancy path had slightly lower Akaike and Bayesian information criteria, but the chi-squared difference test was not significant and the remaining coefficients were nearly unchanged. The researchers retained the prespecified model to display both direct and indirect pathways without claiming that it had superior empirical fit.

Several measurement and model checks supported the analysis, although the structural model’s fit statistics were mixed. Confirmatory factor analysis produced standardized factor loadings of 0.885 to 0.953 for organizational justice and 0.796 to 0.830 for turnover intention. Composite reliability was 0.909 for organizational justice and 0.840 for turnover intention, while the corresponding average variance extracted values were 0.769 and 0.567. The measurement model had a comparative fit index of 0.979, a Tucker–Lewis index of 0.965 and a standardized root mean square residual of 0.029. For the structural model, the comparative fit index was 0.961 and the standardized root mean square residual was 0.043, but the root mean square error of approximation was 0.093 and the chi-squared-to-degrees-of-freedom ratio was 7.82, which the authors characterized as less than optimal. Indirect effects were tested with bias-corrected bootstrapping using 5,000 resamples and 95% confidence intervals, and geographic region was included as a control variable.

The findings identify organizational and psychological correlates of turnover intention, but they do not establish a causal sequence. All variables were measured at the same time, so the analysis cannot show that perceived wage inadequacy preceded burnout, that burnout weakened commitment, or that changing any of these factors would reduce actual departures. Reverse and reciprocal relationships remain possible: nurses already planning to leave may judge compensation and organizational fairness less favorably, while burned-out nurses may perceive less reward or justice. The outcome was turnover intention rather than observed turnover. Self-reported wages and expectations, self-selection into the online panel and potential clustering of nurses within hospitals that was not explicitly modeled also limit generalizability. The evidence therefore supports a structural association between perceived wage gaps and intended turnover in this South Korean sample, not a causal claim that pay alone determines whether nurses remain in their jobs.

Subject of Research: Perceived wage discrepancy, organizational factors and turnover intention among South Korean hospital nurses

Article Title: A structural pathway from wage discrepancy to turnover intention among hospital nurses: the mediating roles of effort–reward imbalance, organizational justice, burnout, and organizational commitment

Article References: Yu, M. (2026). A structural pathway from wage discrepancy to turnover intention among hospital nurses: the mediating roles of effort–reward imbalance, organizational justice, burnout, and organizational commitment. Frontiers in Public Health, 14, Article 1966344. https://doi.org/10.3389/fpubh.2026.1966344

Image Credits: AI Generated

DOI: 10.3389/fpubh.2026.1966344

Keywords: nurse retention, perceived wage discrepancy, turnover intention, burnout, organizational justice, effort–reward imbalance, organizational commitment, hospital nurses, South Korea, structural equation modeling, structural, pathway