A manager can invite ideas, stay accessible and respond constructively, yet still make employees feel tense every time they have to speak to them. New evidence from more than 4,500 public servants suggests that contradiction matters.
Among Chinese tax officials, inclusive leadership was positively associated with employees reporting more innovative behaviour. But when direct interactions with supervisors felt psychologically pressuring, the strength of that relationship was roughly halved.
The findings add an important qualification to a popular management idea. Being open to employee input may help innovation, but the broader experience of approaching the same manager can determine how useful that openness actually feels.
A large survey inside a tightly regulated public service
Researchers Xiaoting Zhong and Guoxian Bao surveyed tax civil servants across a provincial tax administration system in China between September and October 2023. The final complete-case analysis included 4,511 employees working across general administration, tax collection, taxpayer services, tax inspection and information technology.
The setting is useful because innovation in tax administration does not happen in a free-form environment. Employees may identify better processes or service improvements, but they still work within formal rules, performance systems, hierarchical approval and accountability requirements.
Participants rated inclusive leadership, their psychological contract with the organisation, public service motivation and their own innovative behaviour. Innovative behaviour covered generating, promoting and implementing ideas rather than simply holding favourable attitudes toward change.
They also answered a single question about how psychologically pressured or nervous they felt when directly interacting with their immediate supervisor. The authors deliberately call this supervisory interaction pressure rather than a general measure of workload or performance pressure because it captures a much narrower experience.
The researchers first tested whether the main multi-item measures were empirically distinct. Their four-factor measurement model showed acceptable fit, including a comparative fit index of 0.954 and root mean square error of approximation of 0.065.
They then used sequential mediation and moderated mediation models, controlling for gender, age, education, years of service and organisational function. Indirect effects were evaluated using 5,000 bootstrap resamples.
Inclusive managers were associated with more innovative behaviour
The basic relationship was clear. Inclusive leadership was positively associated with innovative behaviour, with an unstandardised coefficient of 0.1469 after the proposed mediators and controls were included.
Inclusive leadership was also strongly associated with a more favourable psychological contract, meaning employees were more likely to perceive a reciprocal relationship in which the organisation recognised their contributions and fulfilled its obligations. That path had a coefficient of 0.8156.
Psychological contract was in turn positively associated with public service motivation, with a coefficient of 0.4090. Public service motivation was positively associated with innovative behaviour at 0.2109 after the other variables were considered.
This produced a statistically supported ordered indirect association. The pathway from inclusive leadership through psychological contract and then public service motivation to innovative behaviour was 0.0704, with a 95% bootstrap confidence interval from 0.0563 to 0.0848.
The total association between inclusive leadership and innovative behaviour was 0.3566. Of that total, 58.8% was represented by the combined indirect pathways included in the model, while 41.2% remained as the direct effect.
Those percentages should not be read as proof that leadership caused the psychological changes and those changes then caused innovation. The survey was cross-sectional, so the ordering is theoretically consistent with the data rather than demonstrated over time.
Pressure changed how much inclusive leadership appeared to help
The most practically interesting result appeared when the researchers examined supervisory interaction pressure. The interaction between inclusive leadership and this pressure was negative and statistically significant, with a coefficient of -0.0524.
At low supervisory interaction pressure, defined as one standard deviation below the mean, the relationship between inclusive leadership and innovative behaviour had a slope of 0.1958. At high pressure, one standard deviation above the mean, the slope fell to 0.0936.
In other words, inclusive leadership remained positively associated with innovation even when employees felt more pressure around their supervisors, but the relationship was less than half as strong. The supportive behaviour had not become harmful. Its apparent advantage had been substantially constrained.
A similar, though smaller, pattern appeared for psychological contract. The relationship between inclusive leadership and psychological contract declined from 0.8318 under low interaction pressure to 0.7857 under high pressure.
The data therefore capture a managerial tension that can easily be missed by leadership checklists. A supervisor may genuinely invite suggestions and remain available while simultaneously creating enough apprehension that employees become more cautious about using that access.
Not every proposed pressure effect was supported
The moderation results also contain an important negative finding. The researchers expected supervisory interaction pressure to weaken the indirect pathway running through psychological contract and public service motivation.
The point estimate moved in the expected negative direction, but the 95% bootstrap confidence interval included zero. The index of moderated mediation for the full sequential pathway was -0.0020, with a confidence interval from -0.0042 to 0.00004.
That means the study did not provide sufficient statistical evidence that pressure significantly altered the full indirect mechanism. The strongest evidence concerns the direct relationship between inclusive leadership and innovative behaviour, and the relationship between inclusive leadership and psychological contract.
Sensitivity analyses using different treatments of years of service and missing control data produced substantively similar estimates. In the multiple-imputation analysis, the leadership-by-pressure interaction for innovative behaviour remained negative at -0.0458.
Why the result matters for public organisations
Public organisations often ask employees to improve services while simultaneously enforcing strict accountability. That combination is unavoidable in areas such as taxation, licensing, policing, healthcare administration and regulatory enforcement, where experimentation cannot simply override legal or procedural requirements.
The findings suggest that the quality of managerial access matters alongside the formal invitation to contribute. An employee who expects a difficult or highly evaluative encounter may be less willing to bring an incomplete idea, admit uncertainty or ask for the support needed to move a proposal through formal channels.
For South African public institutions, the study should not be treated as evidence that the same coefficients would apply locally. The organisational, administrative and cultural context is different, and the research was conducted within one Chinese provincial tax system.
The underlying management problem is nevertheless recognisable. Public-sector leaders often need to be both supportive and evaluative. Separating developmental conversations from formal performance assessment, explaining the purpose of supervisory meetings and responding consistently to suggestions may help prevent an invitation to innovate from feeling merely symbolic.
The design sets clear limits on what can be claimed
The study has several important limitations. All focal variables came from the same self-report survey, so common method bias and self-presentation cannot be eliminated even though the researchers used procedural separation and statistical diagnostics.
The cross-sectional design also prevents causal conclusions. Innovative employees may perceive their supervisors more favourably, or some unmeasured organisational condition may influence leadership perceptions, motivation and innovation at the same time.
Supervisory interaction pressure was measured with a single item. It therefore cannot distinguish pressure caused by evaluation, communication style, authority, anticipated criticism or other features of the supervisor relationship.
The anonymous survey also did not identify shared supervisors or work units. The researchers could not estimate whether responses clustered within teams or adjust directly for employees reporting to the same manager.
Finally, innovative behaviour was self-reported. The measure captures employees saying that they generate, promote and implement ideas, not whether those ideas were adopted or produced measurable improvements in public services.
Even within those boundaries, the study adds a useful distinction. Inclusive leadership is not only about whether a manager says the door is open. It may also depend on what employees expect to feel when they actually walk through it.
Source Information
Study Title: When inclusive leadership meets supervisory interaction pressure: explaining innovative behavior among public employees
Authors: Xiaoting Zhong and Guoxian Bao
Journal: Frontiers in Psychology
Year: 2026
Published: 29 September 2026
Sample: 4,511 tax civil servants in a provincial tax administration system in China
DOI: 10.3389/fpsyg.2026.1960198
