Systems Change 101

By Swan 🔸, Lin BL, Karen Singleton, Ulf Graf 🔹, Ben R Smith @ 2025-06-23T08:15 (+45)

In alphabetical order: Ben Smith, @Karen Singleton, @Lin BL, Rebecca Zanini, @Swan 🔸 , @Ulf Graf 🔹 

Acknowledgements: Rosanna Zimdahl, Ruben Dieleman, Samuel Hilton and Simon Holm. 

Preface/TL;DR 

We believe systems thinking offers valuable tools for identifying more impactful interventions.

This post aims to help you:

• Understand systems thinking and systems change and advocate for their value to EA.
• Apply systems thinking to your cause area to identify more impactful interventions.
• Find relevant resources to deepen your knowledge in areas that interest you.

The most up-to-date version of this post can be found on this website. 

Introduction 

Advocates push for improved conditions in factory farms, yet industrial animal agriculture persists. Direct cash transfers are distributed to alleviate poverty, yet billions remain ensnared in its grip. Life-saving malaria nets are funded to protect children from deadly disease, yet preventable illnesses still claim millions of lives annually.

Why do these problems persist so stubbornly? Because the choices we make—what we eat, how we use technology, how we respond to global crises—are all shaped by underlying systems that are often invisible or non-obvious. These invisible systems lock in behavioural patterns that resist change, creating 'wicked’ problems that defy simple solutions. Uncovering these systems reveals leverage points for lasting impact. Failing to consider systemic effects can lead to ineffective interventions, such as in the classic example of the PlayPump.

What if we could change the underlying systems that drive suffering, catastrophic and existential risks? As Effective Altruists are committed to doing good better, skilful identification and application of systemic interventions can offer dramatic leverage, in EA’s terms.

But what is a system?

A system is a set of interconnected elements—such as people, institutions, resources, rules, and beliefs—that interact in patterned ways to produce behaviour and outcomes over time. What defines a system is not just its parts, but how those parts influence one another through relationships, feedback loops, delays, flows of information, power, and resources. These dynamics can give rise to behaviours and outcomes that are emergent, nonlinear, and sometimes resistant to change, yet also capable of sudden transformation when key thresholds or leverage points are activated.

Thinking about systems helps us both zoom in and zoom out, revealing the macro & micro hidden structures, incentives and feedback loops that drive these issues. Instead of just mitigating harm, we can influence the underlying systems to prevent harm in the first place.

Consider:

• In AI safety, technical solutions alone won't prevent misalignment if economic and geopolitical structures reward the rapid development of capabilities over safety measures.
• In biosecurity, stockpiling vaccines won't protect us if we can't coordinate global responses, and detection networks remain fragmented.
• In animal welfare, cage-free campaigns help millions of animals but don't necessarily address the economic incentives and cultural narratives that will lead to future animal suffering.
• In global health & development, income support doesn’t fix the problem if people remain trapped in polluted, unsafe and economic systems vulnerable to corruption, exploitation & extraction.

These areas also cannot necessarily be treated as neatly separate issues - they are all to a greater or lesser extent interconnected through our global civilizational ‘operating system’, which is perhaps most clearly highlighted through the potentially transformative impact of the diffusion of advanced AI.

What is systems change?

Understanding systems 

Understanding a system can give us a better ‘map’ for the ‘territory’, and thereby have a better sense of if, where and how to intervene.

As previously mentioned, a system is a set of interconnected elements—such as people, institutions, resources, rules, and beliefs—that interact in patterned ways to produce behaviour and outcomes over time. What defines a system is not just its parts, but how those parts influence one another through relationships, feedback loops, delays, flows of information, power, and resources. These dynamics can give rise to behaviours and outcomes that are emergent, nonlinear, and sometimes resistant to change, yet also capable of sudden transformation when key thresholds or leverage points are activated.

While everything is ultimately interconnected, effective systems thinking requires drawing practical boundaries and abstracting a simplified representation of the system. The skill lies in capturing all important features in a tractable and resource-efficient manner without making the analysis too complex to be useful. 

After identifying a system of interest, we can classify it based on the level of order or constraint, which can help us to decide what type of response is appropriate. One prominent sense-making framework is the Cynefin Framework, which has 5 domains:

• Confusion & aporia: Not knowing what the type of system is.
• Clear/simple/obvious systems: Self-evident linear cause-and-effect relationships
• Complicated systems: Linear, but require deep expertise or analysis to understand
• Complex systems: Entangled - non-linear cause and effect, only understandable in retrospect, featuring emergent, unpredictable behaviours
• Chaotic systems: No substantial behaviour patterns - highly sensitive and unpredictable, appearing to be almost random.

Systems change is a process that fundamentally alters how a system functions by modifying its components and relationships to address underlying dynamics rather than symptoms. It creates enduring improvements by shifting a system's behavioural patterns, preventing issues from recurring. The “inverted triangle” framework provides a model of systems change that recognises three interconnected levels: 

• Structural changes: Tangible, visible elements like policies (e.g. agricultural subsidies), practices (e.g. corporate reporting requirements), and resource flows (e.g. how funding is allocated)
• Relational changes: How people and institutions interact, including power dynamics (who influences decisions), relationships between stakeholders (government-business-community), and how information flows
• Transformative changes: The deepest level involves shifts in collective beliefs, cultural narratives, assumptions, and mental models that shape how we understand problems and solutions

Effective systems change requires collaboration across sectors and disciplines, addressing political, environmental, social, technological, legal and economic dimensions simultaneously.

Why Systems Change Matters 

Many of today’s most pressing challenges—climate change, poverty, animal welfare, AI safety—are "wicked problems": complex, deeply entangled, and resistant to simple solutions. Solving these problems in isolation is akin to playing Whack-a-Mole: fixing one issue can cause others to emerge elsewhere.

In a nutshell, systems change matters because it:

• Helps avoid unintended consequences from isolated solutions
• Helps ensure long-term change is sustainable

In more detail, systems change:

• Helps avoid unintended consequences from isolated solutions
  • E.g. Anti-poverty programs focused solely on financial aid may inadvertently create unintended repercussions
  • E.g. Drug enforcement policies targeting supply routes without addressing demand can drive criminal networks to shift and adapt.
  • E.g. Large-scale (factory) farming provides food for a significant part of the population, yet drives antibiotic resistance and climate change and is the cause of huge animal suffering.
• Helps ensure long-term change is sustainable
  • E.g. Fixing governance and infrastructure issues that prevent clean water access instead of repeatedly treating waterborne diseases.
  • E.g. Focusing only on short-term symptom treatment in addiction services (e.g. funding rehab but not addressing housing/employment) leads to high relapse rates.

Without a systemic approach, well-intentioned interventions can backfire, create new problems elsewhere, or require continuous effort without creating sustainable solutions. Designing solutions that work with the complexity of the systems we are trying to change is essential to avoid these risks.

Common System Behaviours

Understanding how systems behave helps us identify effective intervention points. The behaviour of systems is usually characterised by the following features: 

Feedback Loops: 

• Reinforcing (positive) feedback loops amplify changes, potentially creating runaway effects. For example, Arctic ice melt reduces sunlight reflection, accelerating global warming and causing more ice to melt.
• Balancing (negative) feedback loops stabilise systems by counteracting changes. Predator-prey populations naturally balance each other—as predators increase, prey decrease, eventually causing predator populations to decline.

Nonlinearity: Small changes can sometimes trigger disproportionately large responses in systems, especially near tipping points. Multiple indirect interactions can trigger these responses instead of linear direct interactions.

Emergence: Different elements working together can create impacts far greater than the sum of their individual effects, and with features none of the individual elements had.

Tipping Points: Systems often resist change until reaching critical thresholds where they rapidly shift to new states. Climate scientists worry about tipping points like permafrost melt, where gradual warming suddenly releases massive methane amounts, dramatically accelerating climate change.

Cascading Effects: Changes in one aspect of a system can cascade through interconnected systems, creating chains of impacts that amplify or spread across domains.

Trade-offs: When positive changes in one aspect of a system may lead to negative changes in another aspect of the system.

Common System Archetypes 

In Thinking in Systems, Donella Meadows identified eight different system archetypes describing common and usually recurring patterns of system behaviour. Three particularly relevant for systems change work include:

Tragedy of the Commons: When actors have open access to a shared resource, each maximises their benefit, potentially depleting the resource for everyone (overfishing). 

Success to the Successful: Resources flow disproportionately to those already succeeding, creating widening gaps (economic inequality, market concentration).

Shifting the Burden: Quick fixes address symptoms while neglecting underlying causes, creating dependency on short-term solutions (relying on food aid without building agricultural capacity).

Finding Leverage Points 

Not all interventions have equal impact. Donella Meadows identified a hierarchy of 12 leverage points—places to intervene in a system, ranking them from least to most leverage.

To facilitate strategic thinking about interventions, Meadows’ 12 leverage points can be clustered into four levels, which maintain the hierarchy while simplifying strategic analysis: 

• Parameters: concrete, measurable elements that adjust system behaviour (e.g., subsidies, taxes, quotas)
• Feedback: loops that either amplify or balance system behaviour (e.g., reinforcing loops, balancing loops)
• Design: rules, structures, and information flows that shape how the system operates (e.g., laws and policies)
• Intent: fundamental purpose and worldview that drives the system (e.g., goals (what the system optimises for), paradigms (core beliefs out of which the system arises)) 

Meadows points out that higher leverage points shape everything below them in the hierarchy. When the intent of the system shifts, there is a cascading effect — new worldviews generate new goals, which demand new rules, which require new information flows, which ultimately change parameters. This cascading effect means that paradigm shifts, though typically more challenging to achieve, create more comprehensive and lasting change.

For instance, addressing climate change through carbon pricing (parameters) works within existing economic assumptions, while shifting from endless growth paradigms to regenerative sustainability paradigms (intent) would automatically reshape climate goals, policies, information systems, and economic instruments.

How do you apply systems change? 

This section explores the principles for effective systems change and then presents how they play out across different cause areas and a few successful examples.

Principles for Effective Systems Change 

Effective systems change often begins with a shift in how we understand problems and where we intervene. To start with:

• Model the System: Understanding the system, who the actors within it are, and what flows between them (e.g. money, information). This can help with working out where and how to intervene with these flows and avoid unintended consequences.

After this, some specific things to consider are: 

• Start Small and Build: Complex systems that work tend to evolve from simpler ones that worked (Gall’s Law). Begin with a contained intervention and iterate.
  • Example: the World Wide Web was first developed in 1989, expanded globally when made royalty-free in 1993, and is now ubiquitous.
• Look for Tipping Points: Systems often resist change until a critical threshold is reached. Identify leverage points where small inputs can trigger large shifts.
  • Example: the Montgomery bus boycott in the US civil rights movement.
• Address Power Dynamics: Understand who holds power and how structures maintain the status quo. Lasting change often requires shifting these relationships.
  • Example: antitrust actions to reduce the influence of large companies, e.g. Standard Oil in 1911.

By applying these principles, we can design interventions that address deeper causes, reduce the risk of unintended consequences, and create a more durable impact.

Applying Systems Change to Different Cause Areas 

In this section, we present concrete examples from typical EA cause areas, showing how solutions and lessons can transfer across domains, and how failures in one area do not imply failure in another.

Concrete examples in a typical EA cause area (alphabetically)

AI safety & governance

In AI safety, technical solutions alone won't prevent harm if economic and geopolitical structures and incentives reward the rapid development of capabilities over safety measures in an ‘arms race’ between developers and nation states. High leverage interventions include:

• Efforts to promote the paradigm of seeing AI as a global commons risk, not just as a technology, such as the EA movement’s campaigns for this as a cause area, including 80,000hrs publications, public warnings from the ‘godfathers’ of AI, and campaigns such as Pause AI and Control AI.
• Efforts towards global governance regimes, including OECD Principles on AI, the Bletchley Declaration on Frontier AI, calls for international AI agencies, and setting up a global network of AI safety/security institutes.
• Efforts towards controlling model release and use include the EU AI Act, proposals for training run disclosures, and AISIs' pre-deployment testing of frontier systems.
• Introduction of restrictions on high-end chips by the US as a technical constraint aimed at slowing down or redirecting the development of AI to targeted regions.

Animal Welfare: Factory farming 

• A systems approach means working strategically across different levels, including short-term improvements and long-term paradigm shifts. If we only pursue what’s currently feasible, we risk entrenching the existing system. We want to shift away from treating animals as commodities and challenge the norms that permit their exploitation. This could include:
  • Recognising animal sentience within legal frameworks. Examples include the UK Animal Welfare (Sentience) Act 2022, which formally recognises all vertebrates, as well as cephalopod molluscs (like octopuses and squid) and decapod crustaceans (like crabs and lobsters) as sentient beings and EU Article 13 Treaty on the Functioning of the European Union (Consolidated Version).
  • Implementing true cost accounting to reflect the full environmental, 

health and ethical impacts of animal agriculture. The Global Alliance for the Future of Food champions True Cost Accounting (TCA) document provides a holistic framework for valuing externalities.

• Embedding sentience impact assessments across all areas of policymaking. The establishment of the Animal Sentience Committee (ASC) as part of the UK Animal Welfare (Sentience) Act 2022 means the ASC has a statutory duty to scrutinise any government policy (from any department – e.g., trade, transport, housing, not just environment/agriculture) to assess whether due regard has been given to animal sentience.
• Scaling alternatives to animal products to displace demand. Examples include alternative proteins, pioneers in developing and commercialising plant-based meat products (Impossible Foods and Beyond Meat), precision fermentation & cultivated meat companies (e.g., Upside Foods and Motif FoodWorks).
• Corporate campaigns for cage-free eggs successfully applied systems change within the market and policy spheres, key leverage points within the food system. By pressuring major corporations (McDonald's, Walmart, Nestlé), these campaigns shifted the standard practice of large-scale egg production. This created a ripple effect through supply chains, influencing suppliers, competitors and setting new expectations for consumers and producers. This momentum helped influence the European Commission’s 2021 commitment to phase out cages for all farmed animals by 2027 (though implementation has since stalled). While undeniably benefiting millions of animals, these campaigns also highlight the need for complementary, broader systems-level interventions. To secure long-term gains and fully mitigate future animal suffering, a systems approach must move beyond solely welfare improvements within factory farming. A systems approach aims to transform the entire system toward one where animal use is no longer necessary or socially acceptable. While a tipping point may seem distant, systems change is not about waiting for one; it’s about deliberately shaping the social, political and economic conditions that make transformative change possible over time.

Biosecurity

Some past examples of systems change in biosecurity: 

• Smallpox killed an estimated 300 million people in the 20th Century alone before its eradication in 1980. The World Health Organisation started planning to eradicate smallpox in 1959, which initially lacked resources, and efforts were scaled with the Intensified Eradication Program in 1967. This involved mass vaccination campaigns, enhanced surveillance and containment strategies, and required international collaboration and global coordination. When smallpox eradication was attempted, no human disease had ever been fully eradicated, and smallpox remains the only fully eradicated human disease.
• The Biological Weapons Convention bans the development, production, acquisition, transfer, stockpiling and use of biological weapons. Negotiations started in 1969, and it entered into force in 1975. While imperfect and lacking a formal verification mechanism, with negotiations limited by tensions due to the Cold War, it was the first multilateral agreement that banned a whole category of weapons. It led to the dismantling of over 20 bioweapons programs, and now has a membership of 189 States Parties and four Signatory States. However, the Biological Weapons Convention has notable areas that require strengthening, including in verification and enforcement mechanisms. 

Some examples where systems change would likely be high leverage: 

• Developing new interventions: Broad-spectrum technologies are able to tackle risks from a range of different pathogens, including novel, currently unknown pathogens. Development of new broad-spectrum interventions can help prepare for future pandemics, for example: the development of new broad-spectrum antivirals, novel air disinfection methods such as far-UVC, and improvements to PPE. However, there are fewer market incentives to develop such technologies, for example, because they are less likely to be profitable and due to current regulatory systems. Attempts to incentivise their development include advanced market commitments, competitions, and funding focused research organisations (FROs) to do foundational research.
• Implementation of existing interventions: For example, elastomeric respirators provide better protection than standard N95/FFP2 respirators, and are reusable, so that one respirator may be able to last someone for an entire pandemic. Stockpiling such respirators would be one step toward effectively preparing for the next pandemic. Implementing and combining existing indoor air quality measures (e.g. ventilation, filtration, ultraviolet devices) in buildings can reduce the spread of both potential future pandemics and more frequent illnesses such as seasonal influenza. 

Climate Change

Some examples of high-leverage interventions at the international governance level include:

• The Montreal Protocol (1987) was a highly effective binding international treaty, reducing the use of CFCs and helping repair the ozone layer.
• UNFCCC (1992) United Nations Framework Convention on Climate Change treaty created the framework under which later agreements were made.
• Paris Agreement (2015) and Net Zero Commitments (2020s), although effectiveness is debatable due to soft enforcement and the non-binding nature of the latter.

Some lower-level and localised examples include:

• The Swiss CO2 Levy and Canada Carbon Rebate are examples of green basic income programmes that charge the use of fossil fuels and finance basic income, creating an incentive to reduce use.
• Equal Right has outlined a ‘Cap and Share’ policy, which limits the extraction of fossil fuels and finances basic income (Equal Right, 2023). It is being tested in Tuvalu and Palau.
• Initiatives that fund local protection and reforestation, and forest management include Basic income for nature and climate from Cool Earth, which gives unconditional cash transfers to individuals in important conservation areas to disincentive extraction and exploitation. Costa Rica has a similar programme: “Payments for Environmental Services Programme” (United Nations Climate Change, 2023; World Bank Group, 2022). 

Global Health and Development (GHD) 

• The scale-up of antiretroviral therapy (ART), supported by global partnerships such as PEPFAR and the Global Fund, turned HIV from a death sentence to a manageable condition for millions. Activism by groups like Médecins Sans Frontières and support from countries like Brazil and South Africa led to global policy shifts enabling the generic production of ARTs, dramatically lowering drug costs from $10,000 per patient to less than $100.
• Oral rehydration therapy (ORT), a simple and low-cost intervention, has saved tens of millions of lives from diarrhoeal diseases since its widespread adoption in the 1970s. There are still huge supply chain and logistics issues that prevent it from being accessible. Clear Solutions is an example of a non-profit working in the space.
• Global extreme poverty fell from about 40% in 1990 to roughly 8.5% in 2024, representing immense progress. However, poverty reduction has now stalled, with 3.5 billion people still living on less than $6.85 per day and extreme poverty projected to decline only slowly by 2030. Without more inclusive growth and investment in jobs, education, and services, progress will remain slow for decades. This World Bank report lists climate change as a fundamental risk to further poverty reduction.
• Fixing governance and infrastructure issues that prevent clean water access addresses the deeper causes of disease. For example, investments in sustainable water and sanitation systems through WASH (Water, Sanitation, and Hygiene) and Global Water initiatives help prevent repeated outbreaks of waterborne diseases, rather than relying solely on medical treatment. Sanitation is the linchpin of a lot of GHD efforts. 

Great power conflict 

Examples here focus on nuclear warfare. 

• The recognition of the logic of Mutually Assured Destruction (MAD) shaped Cold War-era policy with the development of second-strike capability. Powerful incentives for deterrence were created, but they also produced long-term strategic instability.
• The Nuclear Non-Proliferation Treaty (NPT) (1968) created a global nuclear order with near-universal nation-state membership, combining disarmament goals with verification and safeguards aimed at limiting the spread of weapons.
• The Intermediate-Range Nuclear Forces (INF) Treaty (1987) successfully eliminated an entire class of nuclear weapons in Europe, though its collapse highlights the fragility of arms control agreements.

Other

We’ve focused on the examples above on core EA cause areas. Here we note some other broader social change movement examples that have occurred (or are still in motion) through the combination of interventions at multiple layers of leverage in the system over time.

• Abolition of the Transatlantic Slave Trade
• Indian Independence Movement
• Women’s Suffrage/Women’s Rights
• US Civil Rights Movement
• Abolition of Apartheid
• Eastern Bloc democratic transitions
• Arab Spring

Causes not in scope 

For the scope of this article, we have chosen not to provide examples of the following: global priorities research, improving institutional decision-making and community building. 

Solutions and learnings from one area can maybe be used in another 

• We can learn from how policymakers approached genetically modified organisms (GMOs) and CRISPR regulation, navigating public concern, transparency, and international coordination, and apply similar approaches to AI governance.
• Focused Research Organisations (FROs), pioneered in the biotech sector and other tech areas, provide a model for mission-driven institutions that sit between academia and industry. This model could be applied in other fields such as climate tech or AI safety.
• Gavi, the Vaccine Alliance, shows how global financing partnerships can accelerate equitable access to critical technologies. Similar models could further support global biosecurity or climate adaptation tools.
• The Global Fund to Fight AIDS, Tuberculosis and Malaria demonstrates how pooled international funding with performance-based financing can drive large-scale health outcomes. Creating more pooled funding initiatives in other cause areas can be a step forward.
• Lessons from animal welfare advocacy show the importance of combining inside-game (corporate and legal reform) and outside-game (public mobilisation) strategies. The global cage-free movement leveraged coordinated corporate campaigns and institutional procurement to improve conditions for hundreds of millions of animals, a pragmatic systems intervention, albeit one that operates within the existing industrial model.
• Advocacy for alternative proteins has supported deeper shifts by shaping regulatory frameworks, catalysing public investment and transforming consumer narratives. This is helping to build the foundations for a post-animal food system. These examples show how coordinated multi-lever strategies can push toward deeper system-level change. 

Failures in one area don’t necessarily mean failure in another

Setbacks or failures in one area or context do not mean a similar approach will not work in another, particularly if lessons are learned from why the original approach failed: 

• Bans on specific practices (e.g., fur farming, foie gras) sometimes shift production to other countries with weaker standards, rather than reducing overall supply. While this shows that such bans are possible, future interventions in this or other cause areas would need to consider the international and global nature of the system to increase the chances of the target outcome being reached.
• Some consumer-facing labelling schemes (e.g. "humane certified") have led to consumer confusion and false reassurance, slowing demand for more fundamental shifts. This failure illustrates risks of poorly designed public communication, but does not invalidate public engagement efforts in other cases in this or other cause areas.
• The Biological Weapons Convention, while a massive international accomplishment, lacks formal verification mechanisms, which have reduced its effectiveness and limited its enforcement. After signing the BWC, the Soviet Union continued to develop biological weapons in one of the largest biological weapons programmes in history. However, it was still the first treaty that bans an entire class of weapons; the subsequent Chemical Weapons Convention had inbuilt verification mechanisms, and future treaties in multiple areas can learn from its successes and failures.
• Over-reliance on voluntary corporate pledges has led to limited progress in emissions reduction. This shows the limits of voluntary approaches but does not mean that public-private partnerships cannot be effective when combined with strong regulation and accountability, as seen in other sectors like public health.

Early attempts to develop renewable energy focused primarily on developing and improving individual technologies (e.g. solar panels, wind turbines) without addressing grid integration challenges. However, there were challenges integrating renewables into existing grids, which highlighted the need for systems thinking about how to implement these new technologies. This is an important consideration for any cause area that looks at emerging technologies, including considerations around synthetic biology and artificial intelligence. 

Next Steps & Call To Action 

So now you have a better understanding of systems change and are wondering what to do next? 

Our goals for the post were for you to:

• Understand systems thinking and systems change and advocate for their value to EA.
• Apply systems thinking to your cause area to identify more impactful interventions.
• Find relevant resources to deepen your knowledge in areas that interest you.

Below are some resources for you to take action. We are also hoping that this post can act as a conversation re-starter on this topic. 

 Resources

Now that you have an introduction to systems change and its role in effective change, here are some resources we recommend to build your understanding from general concepts to practical strategies.

If You Only Have 30 Minutes

Accessible resources to orient systems change through an EA lens.

• Systems Change: Context (Video)
  • A short video outlining what systems change means and why it’s critical for tackling complex social problems. Link.
• Medium: Tools for Systems Thinkers – Leyla Acaroglu
  • A visual introduction to six fundamental systems concepts. Link.
• Complexity and the Search for Leverage (EA forum post)
  • Why EAs should care about systems complexity and how to think about impact within it. Link.
• TEDx: Systems Thinking 101
  • 15-minute talk covering systems thinking and causal loops using real-world examples of fast fashion and industrial agriculture. Link.

If You Have a Few Hours

Systems Thinking for EA Foundations

Establish what systems thinking is and why it matters for systemic impact.

• Thinking in Systems – Donella Meadows (Top Pick)
  • The classic, essential book on systems thinking with timeless tools like stock-and-flow and leverage points. Link.
• Intro to Systems Thinking - Daniel Kim.
  • Intro article with diagrams and mental models. Link.
• Systems Thinking for Social Change. Link.
  • A suite of practical videos on applying systems thinking to social challenges
• Ashoka Systems Change Masterclass
  • Action-focused course on how to shift systems through changemaking. Link.
• TEDx: Systems Thinking is Not Optional: Lessons From a Pandemic (Steve Woodsmall).
  • Unpacks how COVID exposed systemic weaknesses and what systems thinking offers for future preparedness. Link.
• How to Create Systems Change Together (Adrian Röbke)
  • Explores the mindsets and relationships needed to work across silos and shift complex systems collaboratively. Link.

Thinking Systemically: Books on Complexity, Change and Resilience

Find insights into how systems behave, change and sometimes break. Useful for those seeking to understand dynamic systems, design better interventions, or build resilience into their thinking.

• Systems Thinking For Social Change: A Practical Guide to Solving Complex Problems, Avoiding Unintended Consequences, and Achieving Lasting Results – David Peter Stroh
  • A hands-on guide to solving complex problems using systems tools in real-world contexts. Link.
• Getting to Maybe: How the world is changed – Frances Westley, Brenda Zimmerman and Michael Patton
  • Shows how social innovators navigate complexity and unpredictability to create change.
• Antifragile: Things that Gain from Disorder – Nassim Nicholas Taleb
  • Explores how systems benefit from shocks and disorder, and how to design for resilience. Link.
• Skin in the Game: The Hidden Asymmetries in Daily Life – Nassim Nicholas Taleb.
  • Argues that accountability and asymmetry are crucial for functioning systems.
• Complexity: A Guided Tour – Melanie Mitchell
  • A lucid overview of complex adaptive systems, including evolution, AI, and networks. Link.
• Introductory Systems Thinking Toolkit for Civil Servants (PDF toolkit)
  • A concise, practical intro to systems thinking for decision-makers, with diagrams and real-world framing. Link.
• Polycrisis patterns: applying system archetypes to crisis interactions. Link.
  • The article uses system archetypes to identify and illustrate three patterns of polycrisis, and shows how to analyze how multiple crises interact and emerge.
• OECD: Systems practice workbook. Link.
  • A workbook about how to use a systems approach for making social change in practice.

Systems Change for Animal, Health and Ecological Welfare

Zoom in on systemic harm and interlinked interventions.

• One Health – World Organisation for Animal Health
  • An integrated, unifying approach that aims to sustainably balance and optimise the health of people, animals and ecosystems recognising they are closely linked and interdependent. Link.
• Industrial Food Animal Production: How It Works And How We Resist – Dr Rachel Mason.
  • A detailed overview of the industrial meat system and how activists are working to change it. Link.
  • Podcast version. Link.
• Summary for policymakers of the thematic assessment of the interlinkages among biodiversity, water, food and health.
  • A high-level summary showing how ecological systems interlink, with consequences for human and planetary health. Link.
• Thematic assessment report of biodiversity loss and the determinants of transformative change and options for achieving the 2050 vision for biodiversity.
  • Summary for policymakers, by Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services. Link.
  • The paper discusses how animal agriculture is not just a site of production but a complex system with cascading effects on human health, ecosystems, microbial resistance and animal welfare. Link.
• Climate Policy Explorer.
  • Shows the best policies for the reduction of CO2 emissions in different countries and sectors. Link

Changing Complex Socioeconomic Systems

Resources that analyse or guide change in industrial, economic and governance systems.

• Influence for Good – How funders can shift systems (Ashoka)
  • A guide for funders and changemakers on how to enable systems-level transformations with resources and leverage. Link.
• Post-growth: The Science of Wellbeing within Planetary Boundaries (Kallis et al.)          
  • A visionary framework for moving beyond GDP toward sustainable, systems-aligned wellbeing. Link.
• Prioritising Profits over People and Planet: The Devastating Impacts of Large Businesses on the Right to a Clean, Healthy and Sustainable Environment (UN Policy Brief)
  • Highlights how unchecked corporate systems endanger human and planetary health. Link.
• Poverty, Prosperity & Planet Report 2024 (World Bank)
  • Lays out pathways to escape interlocking global crises through systems-aware development strategies. Link.
• Public Service and Patterns of Change
  • Looks at institutional behaviour and systems change from within (pgs 2–7). Link.

Complexity, Risk, and Uncertainty

For EAs grappling with AI, biosecurity and long-term risk.

• Santa Fe Institute – Complexity Explorer
  • Research centre for complex systems science, free complexity courses. Link.
• Systemic Cascading Risks (EA forum post)
  • Explores how systems thinking can help address existential and institutional fragility. Link.
• Model-Based Policy Under Deep Uncertainty (EA forum post)
  • Explains how to model systemic interventions under high uncertainty and limited predictability. Link.
• Joe Norman's Complexity Syllabus
  • Curated resources by a systems complexity expert. Link.
• From Intelligence to Wisdom – Simon Haberfellner
  • AI safety through a systems lens. Link.

EA-Specific Reflections on Systems Change

Direct engagement with systems thinking from within EA.

• Effective Altruism and Systemic Change (EA forum post)
  • Overview of common critiques and pathways for integrating systemic approaches into EA thinking. Link.
• Interrelatedness of x-risks and systemic fragilities (EA forum post)
  • For those bridging global catastrophic risks and systems theory. Link.
• Introducing SyDFAIS (EA forum post)
  • A systems design framework for AI Safety field-building. Link.
• Pragmatic AI Safety #3: Complex Systems for AI Safety (EA forum post)
  • Breaks down how complex systems theory applies to understanding and intervening in AI dynamics. Link.

Ongoing Learning, Blogs, Podcasts & Communities: Deepen Your Practice

For those seeking to embed systems thinking in their work.

• Systems Innovation Network eBooks. Link.
• Systems Innovation Playlists
  • Curated video series covering systems mapping, leverage points and change-making techniques. Link.
• School of System Change Blog
  • Features practical insights, reflections, and case studies from practitioners applying systems thinking globally. Link
• What is ‘Good’ Systems Change? (UNDP)
  • A deep dive into evaluating impact in complex systems without relying on simplistic metrics. Link
• Collective Impact Forum
  • Focuses on multi-actor systemic change processes. Link

·       Systems Innovation Network: Learning series

• A learning and collaboration community built around systems tools and transformation frameworks. Link.
• Ashoka Systems Change Podcast
  • Interviews with systems entrepreneurs and funders working to reshape systems in health, education, and beyond. Link
• Ashoka Fellows, Young Changemakers, and Programs
  • Peer networks for people driving systemic change across fields. Link.
• Systems Thinking. Link.
• Transition Network Groups
  • A grassroots global network experimenting with post-carbon, resilient local systems. Link.
• Institute for Healthcare Improvement (IHI)
  • An established leader in using systems change to improve public health and healthcare delivery. Link.
• CSER (Cambridge) – Global Systemic Risk. Link.
• Systemiq / Metabolic / Vinnova
• Applied systems change consultancies; learn from case studies. Link. Link. Link.

Bonus: Depth / Fringe / Niche

Not essential, but great for advanced users.

• The Grammar of Systems
  • A more abstract framing of systems and order/chaos. Link.
• Limits to Growth (Classic) & Update to Limits
  • Classic systems modelling resource depletion and sustainability trade-offs. Link. Link.
• The Donella Meadows Project. Link.
• Cynefin Framework
  • If familiar with complexity and decision science. Link.
• Systems and Complexity in Organisation. – Hosts more advanced theory like the Grammar of Systems. Link.