Performance leadership improves when every critical role has a clear contract connecting purpose, expected outcomes, decision rights, evidence, capability requirements, and review. Annual ratings alone arrive too late and compress complex work into a label that offers little guidance for improvement.
A role contract does not eliminate judgment. It makes judgment more transparent, timely, and useful.
Public-sector reform is exposing a broader leadership problem
Vietnam's Ministry of Industry and Trade reported new efforts to standardize civil-service positions and modernize performance evaluation under Decrees 361/2025/NĐ-CP and 335/2025/NĐ-CP. The direction is to clarify functions, responsibilities, capability requirements, and outputs for each position while connecting recruitment, placement, training, use, and evaluation.
The underlying problem is common to public institutions and companies. Job descriptions describe activities, while annual reviews summarize impressions. Neither creates a strong operating link between what a role exists to achieve, what authority it has, what evidence matters, and how capability should improve.
Leadership needs a more active instrument.
What a role contract means
**A role contract is a living agreement that connects a position's purpose, priority outcomes, decision rights, operating guardrails, evidence standards, capability needs, and review-and-remedy cycle.**
It is not an employment contract and should not become a rigid task list. It clarifies the value a role must create and the conditions required to create it.
The contract should be stable enough to guide accountability and adaptable enough to reflect changing priorities, risks, and evidence.
Why annual ratings fail as a leadership system
They separate expectations from daily decisions
Employees may know their general duties but remain unclear about trade-offs. Should speed or accuracy dominate? Which risks can they accept? When should they escalate? What evidence is sufficient?
Without decision rights and guardrails, accountability becomes retrospective criticism.
They mix outcomes with visibility
Annual reviews are vulnerable to recency, hierarchy, presentation skill, and manager memory. Work that prevents failure or improves a system may be less visible than a dramatic individual contribution.
Evidence should be designed into the role rather than reconstructed at year-end.
They identify gaps after the cost has accumulated
If capability needs are reviewed only annually, training becomes a reward, requirement, or generic course catalog. It is not connected to a current performance gap.
The Ministry's direction toward training based on position and execution capability reflects a better principle: learn in order to perform better, then verify whether performance improved.
A six-part role contract
1. Purpose and public or customer value
State why the role exists and who benefits. Avoid describing the department. Define the value the position is expected to create or protect.
Purpose helps people make coherent choices when detailed instructions do not cover the situation.
2. Priority outcomes
Define a small number of results for the review period. Include quality, timeliness, service, risk, and learning where relevant. Outcomes should be influenced by the role but should not encourage manipulation or narrow optimization.
This is not a long KPI catalog. It is a decision-focused view of performance.
3. Decision rights
Specify what the role may decide, what requires consultation, and what must be escalated. Clarify budget, policy, data, people, and operational authority.
As [leaders need decision protection](/blog/leaders-need-decision-protection) argues, people need protection when they make a reasonable decision within mandate, evidence, and process. They also need accountability when they conceal risk or ignore required controls.
4. Guardrails and dependencies
List legal, ethical, financial, safety, and service constraints. Identify the teams, data, approvals, and infrastructure the role depends on.
Accountability is unfair when leaders demand an outcome but withhold the authority or dependency needed to deliver it.
5. Evidence and review cadence
Agree on the evidence that will inform review: service outcomes, quality measures, stakeholder feedback, case records, decision logs, project results, and system improvements.
Use monthly or quarterly conversations to inspect evidence, understand constraints, and adjust priorities. Annual classification can remain, but it should summarize a year of documented learning rather than surprise the employee.
6. Capability and remedy plan
Compare role requirements with demonstrated capability. Select development through real assignments, coaching, peer review, simulation, or targeted learning. Define what improvement will look like.
If performance remains weak, the remedy may involve clearer expectations, better support, role redesign, reassignment, or formal action. The contract makes that sequence explicit.
Leaders also need contracts between roles
Many failures occur at interfaces rather than inside a position. Policy and execution, sales and operations, central and local teams, or product and risk may each perform their tasks while the handoff fails.
Role contracts should therefore include service expectations between roles: what is handed over, in what form, by when, with which evidence, and how exceptions are resolved.
This supports [closed-loop execution](/blog/leaders-need-closed-loop-execution), where intent, frontline signals, correction, and verified outcomes form a continuous operating loop.
Avoid turning role contracts into bureaucracy
The document should be short enough to use. Leaders should not create dozens of measures or rewrite every routine task. They should focus on high-value outcomes, consequential decisions, essential guardrails, and the evidence needed for learning.
Automation can support evidence collection, but it should not turn performance into surveillance. Quantitative indicators require context. Sensitive data needs clear access, retention, and remedy rules.
The role contract is a leadership conversation supported by evidence, not a score generated without explanation.
Calibrate roles before calibrating people
Organizations often compare employee ratings without first checking whether roles have comparable scope, authority, resources, and risk. This produces false precision. Two people may hold similar titles while one manages a stable process and the other inherits a fragmented system with unresolved dependencies.
Leaders should first calibrate the architecture of the roles. Review whether outcomes are similarly demanding, whether decision rights match accountability, and whether evidence is equally available. Then compare how managers interpret quality and performance.
Calibration sessions should use real cases rather than abstract labels. Participants can examine an effective decision, a missed outcome, an ethical escalation, or a cross-team failure. The purpose is to improve shared judgment, not force every situation into an identical score.
This process also reveals design problems. If several people fail at the same handoff, the remedy may be a changed workflow or dependency rather than individual training. Good performance leadership improves the system around the role as well as the person inside it.
What leaders should measure
At system level, leaders should track:
- percentage of critical roles with current outcome and decision-right definitions;
- clarity of expectations reported by employees and managers;
- frequency and quality of evidence-based reviews;
- time needed to resolve cross-role handoff failures;
- capability gaps connected to actual work;
- improvement after development interventions;
- decisions escalated because authority was unclear;
- repeated failures caused by missing dependencies;
- fairness and consistency across comparable roles;
- service or business outcomes associated with role redesign.
These measures show whether the organization is improving performance architecture, not simply completing appraisal forms.
Role contracts strengthen corrective accountability
Accountability should distinguish error, negligence, uncertainty, and responsible experimentation. A clear role contract makes that distinction easier because expectations, authority, guardrails, and evidence are visible.
It also complements [corrective accountability in innovation leadership](/blog/innovation-leadership-corrective-accountability). When a decision produces harm or failure, the response should contain the problem, establish facts, repair consequences, improve the system, and assign responsibility proportionately.
The objective is neither blame avoidance nor tolerance of poor performance. It is a credible system in which people know what good work means and how gaps will be addressed.
Conclusion
Annual ratings are outputs of a performance system. They should not be mistaken for the system itself.
The real work happens earlier: defining value, clarifying decisions, securing dependencies, inspecting evidence, developing capability, and correcting gaps. Role contracts connect those activities in a form leaders and employees can actually use.
When designed well, they make accountability more demanding and more fair. People receive clearer authority and support, while organizations gain better evidence about whether roles, capabilities, and structures fit the work.
Key Takeaways
- Annual ratings summarize performance but do not create it.
- Role contracts connect purpose, outcomes, authority, guardrails, evidence, and capability.
- Frequent evidence-based reviews reduce surprise and improve course correction.
- Accountability requires access to the dependencies needed for delivery.
- Cross-role service expectations are essential because many failures occur at handoffs.
FAQ
What is a role contract?
It is a living agreement connecting a position's purpose, outcomes, decision rights, guardrails, evidence, capability needs, and review-and-remedy cycle.
Is a role contract the same as a job description?
No. A job description usually lists responsibilities. A role contract clarifies value, current outcomes, authority, evidence, dependencies, and how improvement will be managed.
How often should role contracts be reviewed?
Evidence should be discussed monthly or quarterly, while the contract itself should be refreshed when priorities, authority, risks, or organizational design materially change.
Can role contracts work in the public sector?
Yes. They are especially useful where mandates, legal guardrails, service outcomes, and inter-agency dependencies must be made explicit without reducing public value to a single metric.