Interactive · Based on the Millennium Ecosystem Assessment (2005)
Four Scenarios
Four plausible worlds for 2050. Two axes — how we organise ourselves, and how we treat ecosystems. None of them are predictions. All of them are happening somewhere right now.
Explore the quadrantIn 2005, an international team of scientists published four internally consistent stories about the global future — stress tests for how ecosystems, societies and human well-being might co-evolve to 2050.
Two decades later, each of the four is happening somewhere. The value isn't in picking one. It's in seeing what each world costs, what it buys, and what it breaks.
Ecosystems and Human Well-being: Scenarios, Chapter 8.
Coordinating lead authors: Steven Cork, Garry Peterson, Gerhard Petschel-Held.
Island Press, 2005 — an output of the Millennium Ecosystem Assessment commissioned by Kofi Annan.
01 · The Framework
Two axes, four worlds
The framework turns on two critical uncertainties. Hover a quadrant to preview; click to dive in.
Every scenario sits in one quadrant. Every quadrant produces a distinctive world.
02 · The Four Worlds
In depth
Select a scenario. Each card below expands with the full story.
Global · Reactive
Global Orchestration
A globally connected world that bets on growth, free trade and global public goods — and fixes ecosystem problems after they become visible.
Ethos "The rising tide will lift the boats and fix the mess."
Defining features
- Dominant approach
- Sustainable development via economic growth; global public goods
- Economic approach
- Deep reduction of tariff barriers; externalities internalized through markets
- Social policy
- Global public health, global education, poverty reduction
- Dominant organisations
- Transnational companies; global NGOs and multilateral institutions
Globalisation / Regionalisation dynamics
Dense lattice of supranational institutions. Borders fade for goods, capital, information. Cultural convergence toward a global middle-class aesthetic.
Reactive / Proactive dynamics
Monitoring exists, but pre-emptive action is modest. Problems are solved after they become visible — well-funded but late. Works for fast, legible feedbacks. Fails for slow, non-linear ecosystem change.
Developed world
Prosperity spreads; OECD GDP/capita ~$70k by 2050. Environmental problems partly exported. New health burdens (metabolic, chemical, mental). Homogenized culture.
Majority world
~1 billion escape poverty. Access to health and infrastructure expands. But ecosystem-service breakdowns hit the poor hardest. Cultural and ecological costs accompany success stories.
Potential benefits
- Largest aggregate economy in 2050 (~$180 T)
- Strongest poverty reduction
- Global coordination on markets, transport, health, pests, disease
- Wealth creates capacity to act once priorities shift
Risks
- Progress on environment may be too slow to protect local ecosystems
- Ecosystem breakdowns amplify inequality
- Reactive management is more expensive than preventive
- Cultural homogenization erodes local adaptive knowledge
- Brittle under cascading shocks
Signature imagery
Container ships. Mumbai teenager in Kansas colors. Davos livestreams. A coral reef quietly bleaching while the tourism industry retrofits the hotels.
Regional · Reactive
Order from Strength
A fragmented fortress world of regional blocs that fortify borders and exploit ecosystems reactively — the scenario the MA authors agreed was the most dangerous.
Ethos "Protect our own. Outside is threat."
Defining features
- Dominant approach
- Reserves, parks and national-level conservation
- Economic approach
- Regional trading blocs; mercantilism; tariffs as revenue
- Social policy
- Security and protection (of borders, of nations)
- Dominant organisations
- Multinational companies operating within blocs
Globalisation / Regionalisation dynamics
Economic globalization retreats. Borders close to people, capital, goods, information. Multilateral institutions collapse. Identity politics dominates.
Reactive / Proactive dynamics
Ecosystems assumed to be substitutable by technology — but the tech ecosystem that could provide substitutes never develops in a fragmented world. Monitoring and research collapse along with cooperation.
Developed world
High domestic well-being via aggressive trade and protected reserves — but inequality inside rich blocs rises too. Civil liberties erode. Innovation slows without global flows.
Majority world
The worst scenario for the majority world by a wide margin. Commodity exports continue; rehabilitation funding vanishes. Failed states, pirate zones, mass migration pressure.
Potential benefits
- Security from separation of aggressors (for those inside blocs)
- Reduced invasive-species spread because trade is suppressed
- Local industry protected from global competition
- Locally appropriate policies without international override
Risks
- Extreme inequality within and between blocs
- Malnutrition, loss of liberty, declining well-being
- Severe global environmental degradation
- Security built on exclusion is fragile
- Offloaded harms eventually return (climate, disease, pollution)
Signature imagery
Walls and drones. Abandoned NGO offices. Piracy. Drought-emptied villages. Unexpected thickets of wildlife reclaiming the DMZs between hostile nations.
Regional · Proactive
Adapting Mosaic
A patchwork of regional experiments in ecosystem stewardship — messier and slower than the globalized worlds, but structurally more resilient, with networks that weave bottom-up cooperation over decades.
Ethos "Act locally. Learn everywhere. Share what works."
Defining features
- Dominant approach
- Local–regional co-management; common-property regimes
- Economic approach
- Integration of local rules into regional scale; locally-tuned markets
- Social policy
- Local communities, local equity, social capital
- Dominant organisations
- Cooperatives; global organizations with local chapters
Globalisation / Regionalisation dynamics
Frustration with top-down reform pushes governance down to watersheds, catchments, cities. New international cooperation eventually re-forms around shared practices, not shared rules.
Reactive / Proactive dynamics
Communities invest in understanding their own ecosystems before services break. Adaptive management — plan, act, monitor, learn — becomes the dominant paradigm. Failures are visible and learned from.
Developed world
Slower growth than in global scenarios, but richer local cultural and ecological life. Diverse regional forms — Nordic, satoyama, Midwestern organic, Mediterranean terraces.
Majority world
Two-stage story: early neglect of global inequality, later correction as networks mature. Where social capital exists, remarkable ecological and livelihood recovery. Where it doesn't, stagnation.
Potential benefits
- High coping capacity for local change (proactive)
- Win-win management where knowledge and stewardship align
- Strong bottom-up international networks eventually
- Cultural and ecological diversity preserved
- Structurally the most resilient to surprise
Risks
- Early neglect of global commons (climate, oceans, pandemics)
- Inattention to inequality in early decades
- Less aggregate economic growth
- Local experiments sometimes fail catastrophically
- Slow and messy cross-boundary coordination
Signature imagery
Watershed councils. Farmer cooperatives with solar rigs. Restored wetlands. A UN that becomes a sharing platform. Civic festivals. The uneven terrain of an actively composed world.
Global · Proactive
TechnoGarden
A globally connected, technologically proactive world that prices, tracks and engineers ecosystem services. The highest ceiling for well-being — and the least-visible floor.
Ethos "Engineer nature into a managed garden."
Defining features
- Dominant approach
- Green tech, eco-efficiency, tradable ecological property rights
- Economic approach
- Global markets + ecological property rights + payments for ecosystem services
- Social policy
- Technical expertise; openness; competitive
- Dominant organisations
- International professional associations; technocratic NGOs
Globalisation / Regionalisation dynamics
Deep markets with priced ecosystem services baked in. Research, IP and innovation flow freely. Strong global cooperation on climate, pollution, conservation — because those problems are now priced.
Reactive / Proactive dynamics
Heavy investment in biotechnology, ecological engineering, real-time monitoring. Payments for ecosystem services become normal. Artificial ecosystems designed for specific services.
Developed world
High material well-being, high sophistication. Most people live in highly managed environments. Direct experience of non-human nature becomes rare and ritualized.
Majority world
Significant benefit from technology transfer. Rapid urbanization rebuilt around green tech. But poorer regions are downstream consumers, not producers — dependency in new forms.
Potential benefits
- Market-based win-wins between economy and environment
- Best climate outcome by 2100 (0.06 °C/decade)
- Optimized ecosystem services at unprecedented efficiency
- Broadly distributed material well-being
- High innovation and deployment capacity
Risks
- Technological failures cascade across continents
- Wilderness eliminated as 'gardening' of nature intensifies
- Little experience of non-human nature → simplified views
- Engineered systems optimized for services, not resilience
- Concentration of ethical decisions in a technocratic class
Signature imagery
Vertical farms. Bioengineered coral. Drones pollinating orchards. A forest audited for carbon flux in real time. A hum of optimization that stops only when something snaps.
03 · Comparison
Across eight dimensions
Choose a dimension. The four scenarios diverge sharply on each. Counter-intuitive findings emerge when you flip between them.
Size of the global economy at mid-century. Highest in Global Orchestration.
trillion 1995 US$
Long-run warming rate — the lower the better. TechnoGarden decarbonises fastest; Order from Strength cannot cooperate.
°C per decade
Rate of biodiversity decline at mid-century. Order from Strength is worst; TechnoGarden's decline stabilises later but at a simplified baseline.
rate of loss
Risk of emerging-disease outbreak. Good outlook in globally cooperative worlds; poor in fragmented ones.
outlook
Gap between rich and poor nations. GO narrows it through growth; OS widens it through fragmentation.
level
Inequality inside nations. Adapting Mosaic's local stewardship levels it most; reactive growth in GO creates new gaps.
level
Ability to absorb shocks — pandemics, failures, tipping cascades. AM buys it through many small failures. TG optimises it away.
structural capacity
Higher in fragmented, poverty-persistent worlds; lower where development lowers fertility.
billion
04 · The Shape of Catastrophe
Not the same kind of broken
The MA's most important single finding: the shape of the risk distribution differs between scenarios, not just its average. Each world breaks in a different way.
- Global Orchestration Occasional medium-to-large events before recognition.
- Order from Strength Fat-tailed. Frequent local outbreaks escalate because no controls exist.
- Adapting Mosaic Many small failures; very few global ones — networks catch them.
- TechnoGarden Rare events — but when one hits, it hits very large, tightly optimised.
05 · Who Bears What
The two worlds inside each scenario
Every scenario looks different from OECD and from Sub-Saharan Africa. Flip the view to see both.
Global Orchestration
Prosperity spreads; OECD GDP/capita ~$70k by 2050. Environmental problems partly exported. New health burdens (metabolic, chemical, mental). Homogenized culture.
~1 billion escape poverty. Access to health and infrastructure expands. But ecosystem-service breakdowns hit the poor hardest. Cultural and ecological costs accompany success stories.
Order from Strength
High domestic well-being via aggressive trade and protected reserves — but inequality inside rich blocs rises too. Civil liberties erode. Innovation slows without global flows.
The worst scenario for the majority world by a wide margin. Commodity exports continue; rehabilitation funding vanishes. Failed states, pirate zones, mass migration pressure.
Adapting Mosaic
Slower growth than in global scenarios, but richer local cultural and ecological life. Diverse regional forms — Nordic, satoyama, Midwestern organic, Mediterranean terraces.
Two-stage story: early neglect of global inequality, later correction as networks mature. Where social capital exists, remarkable ecological and livelihood recovery. Where it doesn't, stagnation.
TechnoGarden
High material well-being, high sophistication. Most people live in highly managed environments. Direct experience of non-human nature becomes rare and ritualized.
Significant benefit from technology transfer. Rapid urbanization rebuilt around green tech. But poorer regions are downstream consumers, not producers — dependency in new forms.
06 · Two Decades Later
How the scenarios have aged
The MA was published in 2005. Some of its warnings arrived early. Some things it didn't see coming.
Got right
- COVID-19 (2020–2023) matched the MA's warnings about emerging disease in fragmented-cooperation worlds.
- Authoritarianism and border hardening — Brexit, trade wars, anti-immigrant politics — map onto Order from Strength.
- Biodiversity collapse accelerating — the 2019 IPBES report, the Living Planet Index.
- Climate change tracking upper-GO trajectories. 1.5 °C crossed earlier than most 2005-era models projected.
- Rise of payments for ecosystem services — REDD+, carbon markets, biodiversity credits — is TechnoGarden logic.
- Civic and regional networks — Adapting-Mosaic logic at work in food systems, watershed governance, urban climate action.
- Economic multipolarity — Global Orchestration integration plus Order-from-Strength blocs, simultaneously.
Missed or wrong
- AI and the digital revolution — the shape of information, labour and knowledge in the 2020s fits none of the scenarios cleanly.
- Social-media-mediated polarisation — the MA assumed information flows would help cooperation. Often they undermine it.
- The 2008 financial crisis and its long tail of distrust in global institutions.
- Climate extremes already visible in the 2020s — the MA projected impacts of this scale for 2050.
- Solar and EV cost collapses — faster than any TechnoGarden projection.
- Persistence of extreme global inequality despite deep integration.
- The Anthropocene framing — reframing the whole question in ways the MA doesn't fully engage.
The world we are actually in
A hybrid. Fragments of all four scenarios, operating in different regions and at different scales at the same time. Global cooperation on some issues. Fortress dynamics in others. Regional experimentation in food and climate. Mass deployment of green tech. Rapid erosion of some tipping systems.
The question the scenarios still pose most powerfully is not which one we live in, but which way each decision nudges us.
07 · Robust Findings
What holds across all four
Ecosystems will continue to change. No scenario returns to a pre-industrial baseline. The question is which services survive, and for whom.
Some ecosystem services will decline in every scenario — but which ones, and where, varies dramatically.
Adverse consequences are likely in all scenarios. The difference is severity, distribution and reversibility.
Critical ecological feedbacks determine outcomes — reefs, Arctic ice, Amazon, permafrost, ocean oxygen, pollinator networks — and they are only partly under human control in any scenario.
Globally connected + reactive worlds (GO): global problems are eventually addressed; local services get under-served.
Compartmentalised + reactive worlds (OS): well-being declines. Unambiguous.
Globally connected + proactive (TG): wealthy benefit most; local processes ignored; efficiency without resilience.
Compartmentalised + proactive (AM): strong focus on local ecosystems; global commons neglected until networks mature.
You do not get to pick one of these worlds. Nobody does. What humanity can do is recognise which way each decision nudges us — and notice the cost, the gain, and the break.
Source: Cork, S., Peterson, G., Petschel-Held, G. et al. (2005). "Four Scenarios." In Ecosystems and Human Well-being: Scenarios, Volume 2 of the Millennium Ecosystem Assessment (Island Press). Chapter 8.