Even with evaluation criteria, biased evaluators can twist the system.
Introduction
A company may have an evaluation system.
It may have clear-looking criteria.
Initiative.
Responsibility.
Teamwork.
Proactiveness.
Ownership.
Ability to involve others.
Business understanding.
At first glance, it looks like a proper system.
But if the evaluator is biased, the criteria can be twisted easily.
The process often goes like this:
Personal impression comes first.
The score is decided.
The evaluator finds a suitable reason afterward.
Then they present it as an objective evaluation.
This is not real evaluation.
It is personal impression wearing the clothes of a formal system.
And the situation becomes even worse when biased people are not only evaluating others, but also designing the rules.
When monkeys make the rules, the rules become monkey rules.
In other words, personal preferences, insecurities, status needs, and outdated values become formalized as rules, criteria, customs, and management policy.
What are “monkey rules”?
“Monkey rules” is a structural label.
It does not refer to animals.
It refers to a system where a manager’s personal biases, emotional needs, status concerns, or need for obedience are embedded into workplace rules.
On paper, the criteria may say:
- Initiative
- Responsibility
- Teamwork
- Proactiveness
- Ownership
- Ability to involve others
- Business understanding
But in monkey-rule translation, they may actually mean:
- Initiative = picking up the manager’s unfinished work
- Responsibility = filling organizational holes through overtime
- Teamwork = not disagreeing
- Proactiveness = moving in ways convenient for the manager
- Ability to involve others = making vague instructions work somehow
- Business understanding = reading the manager’s hidden expectations
- Ownership = taking responsibility without authority
That is the danger.
The words look clean.
The meaning is distorted.
Impression first, score second, reason last
A distorted evaluation often follows this order.
1. Personal impression comes first
“This person is difficult.”
“They do not work overtime.”
“They take paid leave.”
“They are too logical.”
“They do not appreciate my guidance.”
“They seem quiet but resistant.”
“They do not pick up my unfinished work.”
The impression is formed first.
2. The score comes next
The score is adjusted to fit the impression.
Initiative: low.
Teamwork: low.
Responsibility: low.
Ability to involve others: low.
3. The reason is created afterward
Then the evaluator attaches formal language.
“Lacks initiative.”
“Needs better collaboration.”
“Should take more responsibility.”
“Needs to be more proactive.”
“Should involve others more effectively.”
It looks like evaluation.
But in reality, the personal impression was converted into a score, then wrapped in evaluation language.
That is post-hoc evaluation.
Evaluation criteria can be hijacked by the evaluator’s values
Evaluation criteria alone do not create fairness.
In fact, vague criteria are easy to hijack.
Take “initiative.”
In a healthy system, initiative means identifying issues and taking appropriate action.
In monkey rules, it becomes:
Picking up the manager’s problems without being asked.
Reading vague expectations.
Taking responsibility without authority.
Filling organizational holes with personal effort.
Initiative becomes convenience.
Take “teamwork.”
In a healthy system, teamwork means cooperating with others to achieve the goal.
In monkey rules, it becomes:
Not disagreeing.
Not disturbing the atmosphere.
Following unreasonable instructions.
Avoiding problem-raising.
Teamwork becomes obedience.
Take “responsibility.”
In a healthy system, responsibility means understanding one’s role and completing necessary actions.
In monkey rules, it becomes:
Working overtime.
Not taking paid leave.
Picking up work outside one’s scope.
Compensating for poor organizational design with personal sacrifice.
Responsibility becomes suffering performance.
This is how monkey rules work.
Beautiful words are dangerous when undefined
The most dangerous evaluation words are often beautiful.
Initiative.
Responsibility.
Teamwork.
Ownership.
Proactiveness.
Ability to involve others.
They sound good.
But if they are not defined behaviorally, they become containers for personal preference.
“Lacks initiative” can justify a low rating.
“Has teamwork issues” can punish someone who disagrees.
“Lacks responsibility” can pressure someone who does not work overtime.
“Does not involve others” can shift responsibility onto someone without authority.
Abstract words become formal wrappers for subjective judgment.
Conditions that create monkey rules
Monkey rules do not appear randomly.
They grow under certain conditions.
1. Criteria are abstract
Words like initiative, responsibility, and teamwork are used without concrete behavioral definitions.
2. One evaluator has too much power
One manager’s personal impression becomes the score.
3. No behavioral logs exist
There is no record of date, context, behavior, expectation, and gap.
4. Reasons can be written after scores
The score is decided first, and the evaluator writes a plausible reason afterward.
5. The objection process is weak
The employee cannot safely correct factual errors or challenge the rating.
6. Overtime and paid leave become hidden criteria
Officially, they may not be evaluated.
In practice, they influence impressions.
7. Silence is treated as teamwork
People who do not disagree are seen as cooperative.
People who raise issues are seen as difficult.
8. Rule-makers are not audited
The bias of evaluators and rule designers is not checked.
When these conditions exist, an evaluation system easily turns into monkey rules.
When monkeys evaluate, the criteria are translated into monkey language
The core problem is not only the criteria.
It is how the evaluator translates the criteria.
The same behavior can be interpreted differently.
Raising an issue
- Good evaluator: identifies risks
- Monkey evaluator: too argumentative
Leaving on time
- Good evaluator: delivers within working hours
- Monkey evaluator: lacks responsibility
Taking paid leave
- Good evaluator: uses leave responsibly
- Monkey evaluator: lacks motivation
Clarifying role boundaries
- Good evaluator: manages responsibility clearly
- Monkey evaluator: avoids work
Proposing system improvements
- Good evaluator: thinks about recurrence prevention
- Monkey evaluator: too theoretical or difficult
Keeping records
- Good evaluator: manages work properly
- Monkey evaluator: cold or distrustful
If the evaluator is distorted, the same behavior receives a distorted meaning.
That is why “we have criteria” is not enough.
The organization must control how criteria are interpreted.
Rating bias is real
Rater bias is not something that disappears through good intentions.
Performance evaluation can include many rating errors.
Halo effect.
One positive impression improves ratings on unrelated dimensions.
Horn effect.
One negative impression lowers ratings across other dimensions.
Leniency.
The evaluator rates everyone too generously.
Severity.
The evaluator rates everyone too harshly.
Central tendency.
The evaluator pushes everyone toward the middle.
Research has also shown that idiosyncratic rater effects can account for a large share of rating variance.
In other words, ratings reflect not only the person being evaluated, but also the person doing the evaluation.
Therefore, evaluation systems should not rely on evaluator goodwill.
They should be designed on the assumption that evaluators are biased.
To prevent monkey rules, require facts before scores
The most important fix is to reverse the order.
Bad evaluation:
Impression
Score
Post-hoc reason
Better evaluation:
Concrete facts
Link to evaluation criteria
Calibration
Score
Explanation to employee
Before scoring, the evaluator should provide behavioral evidence.
If someone receives a low score on initiative, the evaluator should answer:
- What specific behavior supports the rating?
- What was the date?
- What was the context?
- What did the employee do?
- What behavior was expected?
- What was the gap?
- Were there examples of initiative?
- How were those examples reflected?
“Somehow lacks initiative” should not pass.
This makes subjective impression much harder to hide.
System design to prevent monkey rules
To prevent monkey rules, evaluation systems need safeguards.
1. Break criteria into concrete behaviors
Do not stop at “teamwork.”
Define behaviors such as:
- Shared necessary information
- Clarified discussion points before meetings
- Confirmed role boundaries
- Coordinated deadlines with other departments
- Explained facts respectfully to people with different views
2. Require facts before scores
Prevent impression-first scoring.
3. Require date, context, and expectation gap
Do not accept vague comments like “not proactive enough.”
4. Require both positive and negative evidence
If the rating is low, also ask whether there were positive examples.
If the rating is high, also ask whether there were issues.
5. Use multiple evaluators
Do not let one person’s impression determine everything.
6. Calibrate rating severity
Some evaluators are harsh.
Some are lenient.
Adjust for that.
7. Give employees a comment field
Let employees correct factual errors.
8. Create an objection process
Make it possible to challenge evaluations safely.
9. Do not use overtime or paid leave as hidden criteria
Evaluate output, role fulfillment, and process quality instead.
10. Evaluate evaluators
Use feedback from subordinates.
Do not allow people who lower others, lecture excessively, or manage through pressure to go unchecked.
Put rules on the rule-makers
To prevent monkey rules, the organization must apply rules to the people who evaluate and design rules.
That means:
- Evaluating evaluators
- Auditing the bias of rule-makers
- Keeping operation logs
- Recording exception decisions
- Rejecting evaluations based only on abstract words
- Allowing employees to review evaluation reasons
If rule-makers are unlimited, the system can be privatized.
Words like:
“For growth.”
“For responsibility.”
“For ownership.”
“For teamwork.”
can become wrappers for personal preferences.
Rules need rules.
The scariest part of monkey rules: they look righteous
Monkey rules are dangerous because they do not look evil.
They look correct.
“Have initiative.”
“Take responsibility.”
“Involve others.”
“Have ownership.”
“Value teamwork.”
“Grow.”
All of these sound good.
But if hidden inside them is:
Move conveniently for me.
Pick up my unfinished work.
Do not disagree.
Appreciate my lectures.
Follow my long-hours ideology.
Do not trigger my insecurity.
then the rules become monkey rules.
Do not trust beautiful words alone.
Trust audited operation.
Conclusion
When monkeys make the rules, the rules become monkey rules.
Even if evaluation criteria exist, biased evaluators can turn them into subjective impression.
Even if the system looks formal, distorted operation can make it personal.
Initiative, teamwork, and responsibility can all be translated into monkey language.
Therefore, the key to evaluation is not only writing good criteria.
The key is to prevent personal impressions from becoming scores.
That requires:
- Concrete behavioral definitions
- Facts before scores
- Date, context, and expectation gaps
- Multiple evaluators
- Employee comments
- Objection routes
- No hidden evaluation of overtime or paid leave
- Evaluating evaluators
- Rules on the rule-makers
Do not trust the monkey.
Do not trust unchecked discretion.
Put logs, facts, third parties, and objection routes between the evaluator and the score.
The essence of evaluation design is not beautiful wording.
It is reducing the space where evaluator impressions can secretly become scores.
参考情報・根拠メモ
※本文は職場違和感・人事評価の問題を一般化した意見記事。以下は背景理解のための参考情報。
評価者固有効果が評価点に大きく混ざる研究
- Scullen, Mount, & Goff (2000), “Understanding the latent structure of job performance ratings.”
- URL: https://pubmed.ncbi.nlm.nih.gov/11125659/
- ポイント：2つのデータセットで、評価者固有効果が評価点の分散の62％・53％を占めたと報告。評価は「評価される人」だけでなく「評価する人」も強く映す。
よくある評価者エラー
- Dartmouth College, “Common Rater Errors”
- URL: https://www.dartmouth.edu/hr/professional_development/for_managers/performance_management/common_rater_errors.php
- ポイント：ハロー効果、寛大化傾向、中心化傾向など、評価者が陥りやすいエラーが整理されている。
根拠に基づく意思決定
- CIPD, “Evidence-based practice for effective decision-making”
- URL: https://www.cipd.org/en/knowledge/factsheets/evidence-based-practice-factsheet/
- ポイント：意思決定では、科学的研究、組織内データ、専門家の経験、利害関係者の情報など、複数の証拠を批判的に評価することが重要とされる。
公正で透明性の高い評価基準
- 厚生労働省「職業能力評価基準」関連資料
- URL: https://www.mhlw.go.jp/content/11800000/000681928.pdf
- ポイント：従業員の職業能力を適正に評価するには、公正で透明性の高い評価基準が必要と説明されている。
評価制度の公正性・納得性
- 厚生労働省関連資料「評価制度により職員の意欲を引き出すためには、評価の透明性...」
- URL: https://www.mhlw.go.jp/shingi/2005/09/dl/s0908-4i6.pdf
- ポイント：評価制度の公正性・納得性を担保するため、評価に関する苦情や不満等に適切に対応する仕組みの重要性が示されている。
