A student can know that the climate is warming, explain greenhouse gases accurately, describe biodiversity loss and still do nothing differently.
Another student can care deeply about environmental problems and feel paralysed because the scale is enormous.
A third can recycle carefully while misunderstanding which actions matter most.
A fourth can join a clean-up project, feel useful for one afternoon and then conclude that serious environmental decisions belong to governments, scientists and companies rather than ordinary people.
These examples reveal a gap that environmental education often underestimates.
Knowledge is necessary for informed action, but knowledge does not automatically become agency.
Environmental agency is the capacity to understand environmental problems, judge possible responses, believe that action can matter, participate individually and collectively, and revise action as evidence, constraints and responsibilities become clearer.
Current OECD work makes this mechanism especially visible. The PISA 2025 Assessment and Analytical Framework places environmental science inside a wider picture of science identity and agency. The newly published PISA 2025 Results also report students’ environmental awareness, attitudes and reported engagement. OECD’s framework is careful not to imply that environmental education should prescribe one ideological set of behaviours. Instead, it emphasises informed agency: the ability to use scientific knowledge, evaluate evidence and make reasoned decisions about socio-scientific issues.
That distinction is important.
Schools should not measure environmental education by whether every learner adopts the same lifestyle.
They should ask whether learners can move through a more demanding sequence:
understand the system → judge evidence → identify leverage and constraints → select proportionate action → coordinate where needed → observe consequences → revise.
The educational question is therefore not:
“How do we make students care about the environment?”
It is:
“How do we help learners turn environmental knowledge and concern into informed, realistic and revisable participation without pretending that individual behaviour alone solves system-scale problems?”
The 50-second answer
Environmental agency works when learners can connect five layers.
- System knowledge: What is happening, through what mechanism, at what scale?
- Judgement: Which claims and proposed solutions are supported by evidence?
- Leverage: Where can action change outcomes — individual behaviour, household practice, school systems, community decisions, markets, infrastructure or public policy?
- Efficacy: Do I understand how my action, especially coordinated action, could matter?
- Feedback: Did the action work, and what should change next?
A learner with knowledge but no efficacy may become fatalistic.
A learner with concern but weak system knowledge may spend effort on low-impact action.
A learner with agency language but weak evidence standards may become vulnerable to misinformation.
A learner focused only on personal habits may miss institutional leverage.
A learner focused only on institutions may conclude personal behaviour never matters.
Environmental agency therefore depends on linking scientific literacy with decision literacy.
First distinction: agency is not the same as awareness
Awareness answers:
“Do I know this problem exists?”
Agency asks:
“Can I reason about what should happen next and participate appropriately?”
A learner may be highly aware of climate change and know little about energy systems.
They may know plastic pollution is harmful and have no idea which plastic pathways dominate locally.
They may know biodiversity is declining but not understand habitat fragmentation, invasive species, land-use change or ecological trade-offs.
Awareness is the beginning.
Agency needs mechanism.
This is why emotionally powerful environmental messaging can produce shallow education if learners leave knowing that a problem is serious but not how systems work.
Fear without mechanism can become anxiety.
Hope without mechanism can become slogan.
Agency needs both reality and leverage.
Stage 1: build a causal model of the environmental problem
Before discussing action, learners need a model.
Consider urban flooding.
A weak model says:
heavy rain → flooding
A stronger model includes:
- rainfall intensity and duration;
- impervious surfaces;
- drainage capacity;
- soil infiltration;
- topography;
- river and coastal conditions;
- development patterns;
- maintenance;
- climate trends.
Now interventions can be analysed.
Drain enlargement addresses one part.
Green infrastructure changes infiltration and storage.
Land-use planning changes exposure.
Early warning changes consequences rather than rainfall.
Insurance changes financial risk.
Individual behaviour may have limited direct leverage but can still interact with community preparedness and political choice.
A causal model prevents environmental education from becoming “problem → moral action” without the middle.
eduKateSG’s broader owner How Science Works explains how evidence and models build knowledge. Environmental agency takes that knowledge into action selection.
Stage 2: distinguish scale
Environmental problems operate at different scales.
A leaking tap can be fixed at household scale.
Urban air quality involves transport, industry, weather and regulation.
Climate change is global but mediated through national energy systems, cities, firms and households.
A school biodiversity project may meaningfully improve one habitat while barely changing regional species decline.
Scale mismatch creates bad action.
Suppose students are taught about global carbon emissions and the only response offered is “turn off classroom lights.”
Turning off unnecessary lights is sensible.
But if the lesson implies that the climate problem is solved mainly through student light-switch behaviour, the causal model is distorted.
A more accurate lesson asks:
At which level does this action operate, what outcome could it realistically influence, and what larger systems also matter?
This protects students from two errors:
“My bottle choice will save the planet.”
and:
“My behaviour never matters because only governments matter.”
Both are scale failures.
Stage 3: identify leverage rather than listing actions
Environmental education often produces action lists.
Recycle.
Use less water.
Walk more.
Plant trees.
Reduce waste.
These actions vary dramatically in mechanism and impact.
Agency requires leverage analysis.
Ask:
- What variable does this action change?
- How large is the change likely to be?
- Who controls the relevant system?
- What barriers limit the action?
- Could the action create trade-offs?
- Is coordination required?
- How would we know if it worked?
Imagine a school wants to reduce food waste.
Students could make posters.
Or they could measure where waste occurs, identify whether portions, menu design, queue timing or student preferences drive it, test a change, and measure waste again.
The second route teaches agency because it links action to system evidence.
Stage 4: distinguish private action from collective action
Environmental problems often involve public goods.
Clean air.
Stable climate.
Biodiversity.
Safe water.
One person can contribute, but outcomes depend on many actors.
This means individual behaviour is only one form of agency.
Collective agency can include:
- school decision-making;
- community projects;
- citizen science;
- public consultation;
- consumer coordination;
- professional practice;
- institutional purchasing;
- policy participation;
- technical innovation.
Education should help learners see these channels without turning the classroom into partisan advocacy.
The teacher’s job is not to tell students which political position to adopt.
It is to make the decision architecture visible.
Who has authority?
Who bears costs?
Who receives benefits?
Which evidence is relevant?
What trade-offs exist?
What forms of participation are legitimate?
That is civic scientific literacy.
Stage 5: efficacy requires plausible pathways
“You can make a difference” is motivational language.
It is not enough.
Learners need to know how.
Suppose students want to reduce heat exposure around the school.
A plausible pathway might be:
- measure temperatures across school surfaces;
- map shade and heat hotspots;
- compare materials;
- identify areas under school control;
- propose shading or planting interventions;
- estimate costs;
- present evidence to school leadership;
- implement one feasible change;
- measure temperature again.
Now efficacy is evidence-based.
The student learns:
data can enter a decision process.
Even if the proposal is rejected, the learner gains a more realistic model of agency than “care more”.
Stage 6: action should generate feedback
Agency without feedback becomes performance.
A school holds a clean-up day.
Photographs are taken.
Everyone feels positive.
What changed?
How much waste was removed?
Did litter return?
Where did the waste originate?
Was prevention addressed?
Did student behaviour change?
Environmental action should be treated like inquiry.
baseline → intervention → measurement → interpretation → revision
This is one of the strongest educational moves because it connects environmental participation with scientific method.
The action becomes something students can learn from rather than merely something they perform.
Stage 7: agency includes choosing not to act on weak evidence
Environmental urgency can create pressure to act quickly.
Sometimes quick action is necessary.
Sometimes restraint is responsible.
A proposed intervention may have:
- weak evidence;
- hidden ecological trade-offs;
- high cost;
- low scalability;
- unintended consequences.
Agency is not measured by activism volume.
It includes the capacity to say:
“The goal is valuable, but this proposed method is poorly supported.”
That is why epistemic judgement and environmental agency must travel together.
Environmental agency is not environmental anxiety
Concern can motivate action.
Persistent helplessness can reduce it.
Environmental education sometimes presents a long sequence of crises:
warming, extinction, pollution, water scarcity, deforestation, ocean change.
If every unit ends with catastrophe and vague encouragement, learners may conclude:
“Everything is getting worse and nothing I do matters.”
A more responsible curriculum pairs seriousness with response architecture.
For each major problem:
- what mechanisms created it?
- what interventions have worked?
- where have interventions failed?
- which actors control leverage?
- what uncertainties remain?
- what can different people realistically do?
Hope should be built from evidence of leverage, not denial of difficulty.
Environmental agency is not consumer purity
A learner can begin to think environmental responsibility means making every purchase morally perfect.
This is impossible.
Products have supply chains.
Information is incomplete.
Trade-offs exist.
Individuals have unequal budgets and constraints.
Education should not turn environmental agency into personal purity.
Better:
make informed choices where you have meaningful control, understand system constraints, and recognise where collective or institutional action has greater leverage.
This is more realistic and less likely to convert environmental learning into guilt.
Environmental agency is not “small actions are useless”
The opposite error also appears.
“Individual action does not matter. Only system change matters.”
But systems are composed partly of individual choices, social norms, professional decisions, votes, purchases, technologies and institutions.
Some individual actions:
- aggregate;
- signal norms;
- reduce personal impact;
- build political support;
- create markets;
- develop skill;
- model feasibility.
The question is not individual versus system.
It is how levels interact.
Environmental agency is not one correct lifestyle
Families live in different contexts.
Urban transport differs from rural transport.
Energy systems differ.
Housing differs.
Income differs.
Dietary options differ.
Climate differs.
A universal action list can therefore become misleading or unfair.
Teach principles and evidence.
Let learners analyse local constraints.
Agency becomes contextual rather than performative.
A climate example: from carbon footprint to systems map
Suppose a class studies household carbon emissions.
Students calculate simplified footprints.
This can be useful.
Then extend the model.
Which emissions are directly controlled?
Which depend on national electricity generation?
Which depend on transport infrastructure?
Which depend on housing?
Which depend on product supply chains?
Now the learner sees nested control.
A household can choose efficiency measures.
A city controls transit and land use.
A utility controls generation mix within regulation.
Firms design products.
Governments shape standards and incentives.
The learner’s agency becomes more realistic because responsibility is mapped rather than moralised.
A biodiversity example: action requires diagnosis
A school notices fewer butterflies.
Students propose planting flowers.
Good instinct.
But what is limiting the population?
Nectar?
Larval host plants?
Pesticides?
Habitat connectivity?
Seasonal timing?
Local climate?
A stronger project investigates species requirements first.
Perhaps the site already has abundant flowers but lacks host plants.
Then “plant more flowers” would be attractive but weak.
Agency improves when action follows ecological diagnosis.
A waste example: measure before moralising
Students complain that classmates “do not care” because recycling bins contain contamination.
The school could run an awareness campaign.
First inspect the system.
Are labels clear?
Are bin colours consistent?
Are recycling and general-waste bins co-located?
Does the local recycler accept the materials students think it accepts?
Are food-contaminated items common?
Is the collection actually separated downstream?
Behaviour may reflect system design more than attitude.
This is an important agency lesson:
before blaming people, inspect the choice architecture.
Common failure mode: the poster project
Environmental topic.
Students make posters.
“Save Earth!”
“Reduce, reuse, recycle!”
The project looks active.
What did students reason about?
Perhaps very little.
Posters can be useful communication products.
They should follow analysis.
What audience?
What behaviour?
What barrier?
What evidence?
What message?
What measure of success?
Then communication becomes part of an intervention rather than environmental decoration.
Common failure mode: catastrophism without action architecture
A curriculum can unintentionally train despair.
Every unit demonstrates severe damage.
No unit analyses successful intervention.
Students learn the problem space but not the solution space.
Include cases where policy, technology, conservation or behaviour change produced measurable improvement.
Also include failures.
Agency needs evidence that systems can change and humility about how difficult change can be.
Common failure mode: tokenism
A school runs one Earth Day.
Students plant one tree.
Environmental agency is declared achieved.
Single events can be meaningful.
Agency needs repeated opportunities to reason, decide, act and reflect across the curriculum.
The event should connect to learning architecture.
Common failure mode: individualising structural problems
A student is told:
“Climate change is your responsibility. Take shorter showers.”
This is too simple.
Household conservation matters.
Climate mitigation involves energy, industry, transport, land use, buildings, technology and policy.
Teach distributed responsibility.
Otherwise students may carry guilt without understanding leverage.
Common failure mode: politicising without teaching evidence
Environmental questions often intersect with politics.
That is unavoidable.
Schools should not resolve contested policy choices by instructing students which party, candidate or ideology to support.
They can teach learners to analyse:
- scientific evidence;
- costs and benefits;
- distributional effects;
- uncertainty;
- time horizons;
- implementation constraints;
- ethical trade-offs.
A learner may then reach a reasoned position.
The education lies in the quality of reasoning.
Common failure mode: action without feedback
Students start a recycling programme.
No one measures anything.
Months later, nobody knows whether waste changed.
Agency has become ritual.
Build feedback into the plan.
Measure before.
Measure after.
Compare.
Revise.
If you are a learner
When you encounter an environmental problem, do not jump immediately from emotion to action.
Build a simple model.
What is causing the problem?
At what scale?
Who controls the important variables?
What actions are proposed?
What evidence supports them?
What can I do individually?
What requires coordination?
How would we know if it worked?
This protects you from both helplessness and performative action.
Then choose one action proportionate to your role.
You do not need to solve the entire system to have agency.
If you are a parent
Environmental concern can become emotionally heavy.
If a child says:
“The planet is doomed,”
do not answer with false reassurance.
Ask:
“Which problem are you thinking about?”
Make it specific.
Then investigate mechanism and response.
What is changing?
What are scientists confident about?
What interventions exist?
Who controls them?
What can a family reasonably do?
What is outside the family’s control?
This moves conversation from diffuse fear toward informed agency.
Also avoid moralising every household choice.
Environmental learning should deepen judgement, not make family life a constant purity test.
If you are a teacher
For each environmental unit, add an agency layer.
After teaching the mechanism, ask:
- What interventions follow logically?
- What evidence supports each?
- What scale does each operate at?
- Who has authority?
- What trade-offs appear?
- How would success be measured?
Do not require students to adopt one approved behaviour.
Require them to reason accurately.
Use local problems where possible.
A school-energy audit.
Heat mapping.
Biodiversity survey.
Waste measurement.
Water use.
Then connect local action to larger systems so students understand both reach and limitation.
A classroom mini-sequence: from concern to tested action
Topic: cafeteria food waste.
Lesson 1: Measure waste by category for five days.
Lesson 2: Interview kitchen and students.
Lesson 3: Build causal hypotheses: portion size? menu preference? queue timing? packaging?
Lesson 4: Select one low-cost intervention.
Lesson 5: Implement for one week.
Lesson 6: Measure again.
Lesson 7: Decide whether the intervention should continue, change or stop.
This sequence teaches science, data, systems thinking and agency.
It also teaches an important civic habit:
good intentions are hypotheses until outcomes are measured.
A Clementi-floor test: can the learner locate leverage?
Give students one environmental claim:
“We should reduce plastic pollution.”
Ask them to map five levels.
Individual.
Household.
School or organisation.
Industry.
Government.
For each level:
What action is possible?
What mechanism does it change?
What constraint appears?
What evidence would show impact?
A learner who can do this has moved beyond awareness.
They are beginning to reason about agency.
Frequently asked questions
What is environmental agency?
The capacity to understand environmental problems, evaluate evidence and possible responses, recognise where action can matter, participate appropriately and revise action based on outcomes.
Is environmental agency the same as environmental behaviour?
No. Behaviour is one output. Agency includes judgement, efficacy, understanding of scale, participation and the ability to choose among actions.
Should schools tell students what environmental actions to take?
Schools can teach evidence-supported practices and create environmental routines. For contested policy choices, the educational role is to teach evidence, mechanisms, trade-offs and legitimate participation rather than instructing students toward a political identity.
Do individual actions matter?
Some do strongly, some weakly, and many interact with system-level conditions. The important skill is matching action to scale and leverage.
Can environmental education increase anxiety?
It can if students receive repeated catastrophic information without credible pathways for understanding and response. Seriousness should be paired with evidence about intervention and agency.
What is environmental self-efficacy?
A learner’s belief that they can perform environmental actions or contribute effectively to relevant outcomes. It is one component of broader agency.
Is activism required for environmental agency?
No. Agency can include personal behaviour, professional practice, community projects, informed consumption, citizen science, institutional change, public participation and many other forms.
How can schools avoid token environmental projects?
Connect action to curriculum, diagnose the system, establish a baseline, measure outcomes and reflect on what changed.
What if students disagree about environmental policy?
That can be educationally productive if evidence standards, respectful discussion and distinction between scientific findings and policy value judgments remain clear.
What is the simplest agency question?
“Where in this system could an action realistically change an outcome, and how would we know?”
Evidence and caveats
The OECD PISA 2025 Assessment and Analytical Framework places environmental knowledge and agency within the current science-literacy architecture. The framework emphasises using scientific knowledge and evidence in relation to real-world and socio-scientific issues while also examining attitudes, identity and agency.
The PISA 2025 Results, published on 8 September 2026, provide contemporary international data on science learning, environmental awareness and student attitudes. These large-scale data are valuable for describing patterns but remain primarily observational.
Several caveats matter.
Reported attitudes are not identical to behaviour. Students may say they care about environmental issues without acting, or act without endorsing survey language.
Behaviour is constrained by infrastructure. A learner cannot choose public transport where none exists.
Agency is culturally and politically contextual. Legitimate participation routes differ across societies.
Environmental outcomes are multi-causal. A school project may influence one variable without changing the larger system.
Education should not overclaim individual responsibility. Large environmental problems involve institutions, markets, technology and governance as well as personal behaviour.
The responsible conclusion is:
environmental education becomes more powerful when students learn not only what environmental systems are doing, but how evidence can inform proportionate action at the right scale.
The deeper idea: agency is knowledge that can enter the world
Science education can stop at explanation.
Why is the climate warming?
Why are species declining?
How does pollution move?
Those questions matter.
Agency adds:
What follows from the explanation?
Which actions fit the mechanism?
Which actor has leverage?
What trade-off appears?
What evidence would change our mind?
What happened after we acted?
That is not activism disguised as science.
It is science extended into decision.
The learner is no longer only someone who knows what the system does.
They are becoming someone who can reason about how humans enter the system — carefully, collectively and answerably to evidence.
Sources and further reading
- OECD. PISA 2025 Assessment and Analytical Framework. Published 29 May 2026. https://www.oecd.org/en/publications/pisa-2025-assessment-and-analytical-framework_01c7e80b-en.html
- OECD. Science Framework for PISA 2025. https://pisa-framework.oecd.org/science-2025/
- OECD. PISA 2025 Results (Volume I): Learning and Equity in a Changing World. Published 8 September 2026. https://www.oecd.org/en/publications/pisa-2025-results-volume-i_73451bc5-en.html
- UNESCO. Education for Sustainable Development. https://www.unesco.org/en/sustainable-development/education
- UNESCO. Greening Education Partnership. https://www.unesco.org/en/sustainable-development/education/greening-future