You might be surprised at how many people are asking this same question. But you probably won’t be surprised at how many people also think they already know the answer – and believe they’re “right.”
Talking about government is tricky because it almost always drags us straight into politics. And once politics is involved, most of us start thinking about whichever party or group we usually support. We judge how government is performing through the lens of who we voted for, who we like, or who we can’t stand.
I’m sure you can relate to feeling more forgiving when the people in charge are the ones you support – and much less forgiving when they’re not.
But here’s the part that’s hard for many of us to get our heads around:
It no longer matters.
The reason government isn’t working today has very little to do with politics in the way we usually think about it. That’s why you’ll hear people say things like “they’re all the same” or talk about a “uniparty.” It’s not that all politicians secretly belong to one big club. It’s that every political party we can choose from today is still part of the same way of thinking.
And that way of thinking – not a group of elites, not a secret organisation, not the banks or big business – is what has shaped every decision that has led us to where we are now.
This is what “the establishment” really means. Not a class. Not a conspiracy. A worldview.
A worldview that every major political party shares, even the ones that claim to be “anti‑establishment.” And once you understand that, you start to see why government feels stuck, confused, or unable to fix anything – no matter who gets elected.
How government works today – how politics works, how elections work, how decisions are made, and how those decisions affect you and me – is far too complex to cram into one message here. And trying to do that would only confuse things further.
Yes, we need to understand what has happened, why it has happened, and where politics is heading now. I’ll give you links that will help you explore those questions and ask new ones too.
But for now, what matters just as much – if not more – is finding others who are looking for direction, who want to put reason before fear, and who are open to working together to take positive steps in the same way.
Whatever difficulties lie ahead, we will be far better able to face them together than alone.
Artificial intelligence has become the centre of one of the most polarising debates of our time. My earlier work on AI – explored in pieces such as Actions Speak Louder Than Digital Words – argued that digital tools amplify human behaviour, incentives, and blind spots rather than replacing them. That principle has only grown more relevant as AI has advanced.
This essay builds on that foundation, examining why the real threat is not AI wiping out humanity, but the risk of human redundancy created by the systems we have built around it. For readers who want to explore the wider context of this argument, links to my previous work are included in the Further Reading section at the end.
PART 1 – The False Choice Between AI Utopia and AI Armageddon
The debate about artificial intelligence has become trapped in a false choice. On one side, we are promised a golden age where machines liberate humanity from work, scarcity, and limitation. On the other, we are warned of an approaching Armageddon in which superintelligent systems escape human control and bring civilisation to an abrupt end.
These extremes dominate headlines, policy conversations, and public imagination. They are compelling, dramatic, and emotionally charged – but they are not the reality we face.
AI could end humanity. It is technically possible. But the probability is dwarfed by dangers that lie much closer to home, dangers created not by machines but by the systems, incentives, and worldviews that shape how we build and deploy them. The real risks are not futuristic; they are present, familiar, and deeply human.
My earliest work on AI, long before today’s existential‑risk debate took hold, focused on human behaviour rather than machine capability. I argued that digital tools amplify our choices, our incentives, and our blind spots – that technology reveals who we are more than it changes who we are. That principle has only grown more relevant as AI has advanced. The stories we tell about AI today say far more about us than they do about the technology itself.
The truth is simple: AI is not on a collision course with humanity. Humanity is on a collision course with the consequences of its own systems – and AI is merely accelerating the impact.
We are not facing a future determined by machines. We are facing a future determined by the choices we make now, the structures we maintain, and the hard truths we are willing – or unwilling – to confront.
The danger is not that AI will suddenly become uncontrollable. The danger is that we continue to behave as if the systems we built are sustainable when they are not.
This piece is not about dismissing risk or denying possibility. It is about grounding the conversation in reality, recognising the fragility of the world we already inhabit, and understanding that the most important decisions about AI are not technical – they are human.
PART 2 – The AI Doom Narrative Is a Story, Not a Forecast
The idea that artificial intelligence is on the verge of wiping out humanity has become one of the most powerful stories of our time. It is repeated by influential technologists, amplified by media hungry for drama, and absorbed by a public already anxious about rapid change.
But despite its emotional force, the “rogue superintelligence” narrative is not a forecast. It is a story – shaped by worldview, psychology, incentives, and imagination far more than by the realities of how AI actually develops.
Stories matter. They shape policy, investment, public fear, and political attention. But stories can also mislead, especially when they are told by people who are brilliant in one domain and inexperienced in others.
The individuals driving the AI‑doom narrative are often exceptional engineers and theorists. They understand algorithms, scaling laws, and computational capability. What they do not understand – and what they rarely acknowledge – are the economic, political, social, and infrastructural constraints that define the real world.
The doom narrative assumes a world of perfect conditions: perfectly connected systems, perfectly maintained infrastructure, perfectly aligned incentives, perfectly centralised control, and perfectly predictable human behaviour. It imagines AI as a single, unified entity capable of seamlessly accessing and manipulating everything. But the world we inhabit is fragmented, fragile, and full of competing interests. Systems fail. Infrastructure breaks. Governments disagree. Markets fluctuate. Human behaviour is messy and unpredictable.
The idea that AI could effortlessly take over such a world is not grounded in reality; it is grounded in abstraction.
It is also grounded in a particular worldview – one that sees intelligence as the ultimate force in the universe, optimisation as the highest good, and complexity as something that can be mastered from above.
This worldview is not malicious. It is simply narrow. It comes from people who spend their lives in digital environments where problems are clean, variables are controllable, and outcomes can be simulated.
But the real world is not a simulation. It is a living, evolving, politically contested, economically constrained, socially complex environment where intelligence alone is not enough to shape outcomes.
The doom narrative also serves psychological and economic purposes. It positions its storytellers as guardians of humanity, elevating their status and influence. It creates urgency that attracts investment. It frames AI development as a race, which benefits those already ahead. And it distracts from the far more immediate risks created by the systems we already have – risks that cannot be solved by technology alone.
None of this means the narrative is entirely wrong. It means it is incomplete. It focuses on hypothetical future dangers while ignoring the structural realities that make those dangers unlikely. It imagines a world where AI is the protagonist, when in truth the protagonist is the system that builds, deploys, and governs AI – a system that is already struggling to manage the complexity it has created.
The real story is not about machines becoming uncontrollable. It is about humans losing control of the systems they built long before AI arrived.
PART 3 – The Real Drivers of Risk Are Human Systems, Not AI Systems
If we want to understand the real risks surrounding artificial intelligence, we have to stop looking at the technology in isolation.
AI is not developing in a vacuum. It is emerging inside political systems that are struggling, economic systems that are stretched, social systems that are fragmenting, and cultural systems that reward speed over reflection.
These systems – not the machines themselves – are where the true dangers lie.
Long before AI entered the mainstream, the guardrails that once protected society from systemic failure had already been weakened. Regulation was hollowed out. Expertise was outsourced. Public institutions were stretched thin. Markets were encouraged to prioritise short‑term gains over long‑term resilience. Technology companies were allowed to become infrastructure providers without being treated as such.
None of this was done with malicious intent. It was the result of decades of incentives that rewarded efficiency, cost‑cutting, and growth above all else.
AI has arrived in a world that is already fragile. It is being deployed into systems that were not designed to handle the complexity it introduces. And because those systems are already under strain, AI doesn’t feel like a tool – it feels like a threat. Not because of what AI is, but because of what the system is not.
The people building AI are not responsible for this fragility. They are operating within the same structures as everyone else. But they are often unaware of how those structures work, or how brittle they have become.
They see AI as a technical challenge, not a systemic one. They imagine that intelligence can solve problems that are, at their core, political, economic, and human. They assume that optimisation can fix systems that were never designed to be optimised.
This is not a criticism of technologists. It is an observation about the limits of any specialised worldview. Engineers are trained to solve problems. Economists are trained to model incentives. Policymakers are trained to manage risk. But the challenges we face today – and the challenges AI accelerates – do not sit neatly within any one discipline. They are cross‑cutting, interconnected, and deeply human.
AI amplifies whatever system it is placed into. In a resilient system, it amplifies resilience. In a fragile system, it amplifies fragility. In a humane system, it amplifies humanity. In a dehumanised system, it amplifies dehumanisation. The danger is not that AI will suddenly become uncontrollable. The danger is that it will faithfully follow the incentives of systems that are already failing.
This is why the existential risk debate feels misaligned. It imagines a future where AI becomes the dominant force in society. But the dominant force in society today is not AI – it is the system itself. And that system is struggling to maintain coherence under the weight of its own contradictions.
The real drivers of risk are not algorithms. They are:
political short‑termism
economic fragility
institutional erosion
social fragmentation
cultural incentives that reward fear, speed, and spectacle
a collective reluctance to confront uncomfortable truths
AI did not create these problems. But it will accelerate them if we do not address them.
The question is not whether AI will become dangerous. The question is whether we will continue to deploy AI into systems that are already dangerous.
PART 4 – The Economic Model Behind AI Is Already Breaking
One of the least discussed truths about artificial intelligence is also one of the most important: the current economic model behind AI cannot sustain the trajectory its loudest advocates imagine.
The march toward ever‑larger models, ever‑greater compute, and ever‑expanding capability is not being driven by a stable, self‑supporting system. It is being driven by a financial and economic structure that is already showing signs of strain.
AI today does not pay for itself. Large language models are loss leaders – subsidised by investment capital, market speculation, and the hope of future revenue that has not yet materialised.
The companies building these systems are not selling profitable products. They are selling potential. They are selling narratives. They are selling the idea that AI will one day become so powerful, so indispensable, and so integrated into every aspect of life that the costs will be justified.
But the costs are rising faster than the revenue. Training frontier models requires vast amounts of compute, energy, infrastructure, and specialised hardware. Maintaining them requires even more. The economic assumptions behind this growth depend on a future where millions of people and businesses pay for AI services at scale.
Yet the same technology is being built to automate, replace, or radically reduce the economic activity of those very people and businesses.
This is the contradiction at the heart of the AI boom: AI is being built to eliminate the labour that funds the system building it.
The more successful AI becomes at replacing human work, the less viable the economic model becomes. A system that depends on human productivity cannot sustain a technology designed to remove human productivity.
This is not a theoretical concern. It is already visible. Industries are experimenting with AI to reduce staffing. Creative sectors are being disrupted. Administrative roles are being automated. Customer service is being replaced. The promise of efficiency is real – but efficiency reduces the number of people who can pay for the services that generate the revenue needed to sustain the technology.
Some argue that Universal Basic Income will solve this problem. But UBI only works when a productive economy underwrites it. A universal UBI in a world without human labour is mathematically impossible.
You cannot create infinite money to meet infinite needs when no one is producing anything the system values.
UBI is not a solution to the economic contradictions of AI. It is a comforting idea that avoids confronting the structural reality.
The truth is simple: the current trajectory of AI will be stopped by economics long before it is stopped by existential risk.
The system cannot fund the future that some technologists imagine. The infrastructure cannot scale indefinitely. The energy demands cannot be met without radical transformation.
The business model collapses if humans become economically irrelevant.
This does not mean AI will disappear. It means the narrative of unstoppable progress toward superintelligence is built on assumptions that do not hold in the real world. The march toward artificial general intelligence is not a straight line. It is a story – one that ignores the economic gravity pulling in the opposite direction.
AI will not end humanity. But the economic contradictions of the system driving AI could destabilise society if we do not address them. The danger is not runaway intelligence. It is runaway incentives.
PART 5 – Why UBI and “Safety Nets” Cannot Solve This
Whenever the conversation turns to automation, job displacement, or the economic contradictions of AI, Universal Basic Income is often presented as the solution. It is comforting, simple, and intuitively appealing: if AI replaces human labour, then society should provide everyone with a guaranteed income.
It feels humane, modern, and technologically aligned. But while UBI can work in limited contexts, it cannot solve the structural problems created by AI at scale.
UBI only functions when a productive economy underwrites it. It depends on a system where value is created, taxed, and redistributed. It requires businesses that generate profit, workers who generate income, and markets that generate activity. In other words, UBI relies on the very economic foundations that AI is being built to disrupt.
A universal UBI in a world where human labour has been largely replaced is mathematically impossible. You cannot create infinite money to meet infinite needs when no one is producing anything the system values.
Money is not a resource; it is a representation of value. If value creation collapses, money becomes meaningless. A society cannot fund itself by printing currency any more than a household can fund itself by writing IOUs. Without production, redistribution has nothing to redistribute.
This is not a criticism of UBI as a concept. It is an acknowledgement of scale.
Small‑scale UBI schemes work precisely because they are underwritten by larger, functioning economies. They are subsidised by systems that still have productive capacity. They are not models for a world where millions of people have been displaced by automation and where the economic engine itself has stalled.
The belief that UBI can solve the economic contradictions of AI is a form of optimism that avoids confronting the deeper structural reality. It assumes that the system can continue functioning even as its foundations are removed. It imagines that money can completely replace meaning, that income can fully replace contribution, and that redistribution can replace participation of any kind.
But human societies do not work that way. People need purpose, agency, and involvement. Economies need activity, production, and exchange. Systems need balance, not dependency.
The danger is not that UBI will fail. The danger is that we will rely on it as a solution to problems it cannot solve. If we continue down the current trajectory – replacing labour without replacing the economic model that depends on labour – we will reach a point where safety nets cannot catch the fall. The system will simply be too large, too strained, and too disconnected from the value it needs to sustain itself.
This is why the existential risk debate feels misplaced. It imagines a future where AI becomes uncontrollable. But the real risk is a future where the economic system becomes uncontrollable – where the incentives that drive AI collide with the realities that sustain society.
AI will not end humanity. But a system that removes human participation without replacing it with something meaningful could destabilise society in ways far more immediate than any hypothetical superintelligence.
The solution is not to rely on safety nets. It is to rethink the system itself.
PART 6 – The Real Future Risk: Human Redundancy, Not AI Rebellion
The most dramatic stories about artificial intelligence imagine a future where machines rise up, break free from human control, and impose their will on civilisation.
These stories are gripping, cinematic, and emotionally powerful. But they distract from the real risk – a risk that is far more grounded, far more immediate, and far more human.
AI does not need to rebel to destabilise society. It only needs to comply.
The systems we have built over decades reward efficiency, optimisation, and cost‑reduction. They reward removing friction, removing delay, and increasingly, removing people.
AI is being deployed into these systems not as a partner, but as a tool for acceleration. It is being used to streamline processes, automate tasks, reduce staffing, and eliminate human judgement. Not because anyone wants to remove humanity, but because the incentives of the system point in that direction.
This is the real existential risk: a world where humans become economically redundant long before they become technologically irrelevant.
AI does not need to develop consciousness, agency, or intent to create instability. It only needs to do what it is designed to do – optimise. And optimisation is indifferent to humanity. It does not ask whether a process should exist, only whether it can be made faster. It does not ask whether a job provides meaning, only whether it can be automated. It does not ask whether a system is healthy, only whether it can be made more efficient.
This is not the fault of AI. It is the result of the incentives we have built into the system.
AI amplifies whatever environment it is placed into. In a system that values human contribution, AI enhances human capability. In a system that values efficiency above all else, AI accelerates the removal of human roles.
The danger is not hostile machines. It is a civilisation that unintentionally sidelines its own people.
This risk is far more plausible than any scenario involving rogue superintelligence. It is already visible in workplaces, industries, and public services. It is visible in the anxiety people feel about their jobs, their purpose, and their place in a rapidly changing world. It is visible in the widening gap between technological capability and economic stability. It is visible in the growing sense that the future is being shaped by forces people cannot influence.
But this risk is not inevitable. It is not a natural consequence of AI. It is a consequence of the system we have built around AI – a system that can be changed.
Human redundancy is not a technological outcome. It is a design choice. And design choices can be redesigned.
The future does not have to be a world where people are pushed aside. It can be a world where AI becomes a partner in human flourishing, where technology enhances meaning rather than erasing it, and where systems are rebuilt to value contribution, creativity, and participation over pure optimisation.
The real existential risk is not rebellion. It is indifference. And indifference can be corrected.
PART 7 – The Golden Age Is Real – But Only With a Different Mindset
For all the anxiety surrounding artificial intelligence, there is a truth that rarely gets the attention it deserves: AI genuinely has the potential to usher in a golden age for humanity.
Not a fantasy, not a marketing slogan, but a real transformation in how we live, work, create, learn, and participate in society. The possibilities are extraordinary – but they depend entirely on the mindset we bring to the technology.
AI is not inherently destructive. It is not inherently liberating. It is a tool – powerful, flexible, and capable of amplifying whatever values and incentives we embed within it.
If we approach AI with fear, it will amplify fear. If we approach it with extraction, it will amplify extraction. But if we approach it with wisdom, humility, and a commitment to human flourishing, it can amplify the very best of us.
The golden age is not a world where machines replace humans. It is a world where machines free humans to do what only humans can do: create, care, imagine, build, explore, and contribute in ways that are meaningful rather than mechanical.
It is a world where technology handles the repetitive, the dangerous, and the exhausting, leaving people with more time, more agency, and more opportunity to shape their own lives.
But this future cannot be built with the same mindset that created today’s crises. It cannot be built with a worldview that prioritises efficiency over wellbeing, optimisation over meaning, or growth over resilience. It cannot be built by systems that treat people as variables, costs, or obstacles.
It requires a shift – not in technology, but in us.
We must rethink value. Not as a measure of productivity, but as a measure of contribution. We must rethink work. Not as a necessity for survival, but as a pathway to purpose. We must rethink governance. Not as a reactive mechanism, but as a proactive steward of human‑centred progress. We must rethink the relationship between humans and machines. Not as competition, but as partnership.
This mindset shift is not abstract. It is practical. It means designing AI systems that enhance human capability rather than replace it. It means building economic models that reward creativity, care, and community. It means creating guardrails that protect people from harm while enabling innovation. It means recognising that intelligence without wisdom is dangerous – and that wisdom must come from us.
The golden age is not guaranteed. It is not inevitable. But it is possible. And it is far more realistic than the dystopian futures dominating today’s debate.
The same technology that could destabilise society under the wrong incentives could strengthen society under the right ones.
The difference is not in the machines. It is in the mindset of the people building, deploying, and governing them.
AI will not save us. But it can help us save ourselves – if we choose to build a future where technology serves humanity, rather than the other way around.
The golden age is real. But it requires us to think differently, embrace technology differently, and behave differently. It requires us to face hard truths, not hide from them. And it requires us to recognise that the future is not being written by machines. It is being written by us.
PART 8 – The Hard Truths We Must Face: Closer to Home Than Armageddon
The most important truths about artificial intelligence are not the ones found in science‑fiction scenarios or theoretical models. They are the truths that lie much closer to home – truths about the systems we have built, the incentives we have normalised, and the behaviours we have accepted.
These truths are not dramatic. They are not futuristic. They are not frightening. They are simply real. And they are far more relevant to humanity’s future than any hypothetical AI Armageddon.
The first hard truth is that our systems were fragile long before AI arrived. We built economies that depend on endless growth, even as the foundations of that growth weakened. We built political structures that reward short‑term thinking, even as long‑term challenges accumulated. We built social systems that prioritise convenience over connection, even as loneliness and fragmentation increased.
None of this was intentional. It was the result of incentives that made sense at the time – incentives that now collide with the complexity of the world we inhabit.
The second hard truth is that AI did not create these problems. It simply revealed them. It exposed the fragility of our infrastructure, the limits of our governance, the contradictions in our economic model, and the gaps in our social fabric. It showed us that the systems we rely on are not as resilient as we assumed. And it accelerated trends that were already underway, making them impossible to ignore.
The third hard truth is that the real danger is not technological. It is behavioural. It is the tendency to outsource responsibility to tools, to assume that technology will fix what only humans can fix, and to believe that progress is inevitable rather than intentional. It is the belief that systems will somehow correct themselves, even when the incentives driving them point in the opposite direction.
The fourth hard truth is that the existential risk debate – the fear of rogue superintelligence – can become a distraction. It can pull attention away from the immediate, solvable issues that matter most. It can make people feel powerless when they are not. It can make the future feel predetermined when it is not. And it can obscure the fact that the most important decisions about AI are not technical. They are human.
But the final hard truth – and the most reassuring – is that none of this is inevitable.
Fragile systems can be strengthened. Misaligned incentives can be redesigned. Harmful behaviours can be changed. Narratives can be rewritten. And futures can be reshaped.
The hard truths are not warnings. They are invitations – invitations to build something better, more resilient, more humane, and more aligned with the world we actually want to live in.
AI will not end humanity. But ignoring the hard truths that lie closest to home could destabilise the systems we depend on. The solution is not fear. It is honesty. It is clarity. It is responsibility. And it is the recognition that the future is not being written by machines. It is being written by us – through the choices we make, the systems we build, and the truths we are willing to face.
Facing these truths is not frightening. It is liberating. Because once we see the real risks clearly, we can address them. And once we address them, the path to a golden age becomes not just possible, but achievable.
PART 9 – Closing: The Future Is Still Ours to Shape
For all the noise surrounding artificial intelligence – the predictions, the warnings, the promises, the fears – one truth stands above the rest: the future is still ours to shape.
AI is powerful, transformative, and accelerating quickly, but it is not destiny. It is not a force of nature. It is not an inevitable outcome. It is a tool, and tools take their meaning from the hands that guide them.
The real risks we face are not distant or speculative. They are present, familiar, and deeply human. They come from systems that have drifted away from resilience, from incentives that reward the wrong outcomes, and from narratives that obscure the choices we still have.
AI did not create these risks. It simply revealed them. And in revealing them, it gives us an opportunity – perhaps the most important opportunity of our time – to rethink how we build, govern, and value the world around us.
We do not need to fear a future where machines take control. We need to focus on building a future where humans remain central – not because we cling to old roles, but because we embrace new ones. A future where technology amplifies our strengths rather than erasing them. A future where systems are designed for wellbeing rather than efficiency alone. A future where progress is measured not by how much work machines can do, but by how much opportunity they create for people.
The golden age is not a fantasy. It is a possibility. But it requires us to step away from the extremes – the utopias and the dystopias – and confront the reality in front of us with honesty and courage. It requires us to recognise that the most important decisions about AI are not technical. They are human. They are about values, incentives, governance, and responsibility. They are about the kind of world we want to build, and the kind of lives we want to lead.
AI will not end humanity. But the systems we build around it could shape humanity’s future in ways that matter deeply. That is not a reason for fear. It is a reason for action. It is a reason for clarity. And it is a reason for hope.
Because once we understand the real risks – and the real opportunities – we can choose differently. We can design differently. We can behave differently. And in doing so, we can build a future where technology serves humanity, not the other way around.
The future is not being written by machines. It is being written by us.
Further Reading:
The ideas in this essay build on a wider body of work exploring how artificial intelligence amplifies human behaviour, exposes systemic fragility, challenges economic assumptions, and forces us to rethink human meaning and agency.
For readers who want to explore these themes in more depth, the following pieces provide a guided path through that broader conversation.
1. Foundations: How Digital Tools Amplify Human Behaviour
This early piece lays the foundation for all later work: digital tools don’t change human behaviour – they amplify it. Understanding this principle is essential to understanding why AI reflects the systems we place it into, rather than replacing them.
A detailed look at how weakened institutions, eroded guardrails, and decades of optimisation created the conditions that make AI feel dangerous – even though the danger comes from the system, not the technology.
A critique of the narratives that distort public understanding of AI. This piece explains how fear‑based storytelling distracts from the real risks and reinforces the false choice between utopia and Armageddon.
An exploration of how political, economic, and social fragility converge – and why ignoring these intersections leads to crisis. This piece shows how AI accelerates pressures that were already building.
3. Economic Contradictions: Why the Current Model Cannot Sustain AI
• When AI Builds a Machine World, This Economy Can No Longer Sustain
A deep dive into the economic contradictions behind AI. It explains why the current model cannot support a future where human labour is displaced at scale – a key argument in this essay.
• As AI Ends Work: Waking Up to the Illusion of UBI
A philosophical core of your AI worldview: if AI removes human agency, it ceases to be a tool for human flourishing. This piece directly supports the argument that redundancy – not extinction – is the real threat.
• The Human Sovereignty Charter for Artificial Intelligence
A constitutional framework for governing AI in ways that preserve human dignity, agency, and sovereignty. Essential reading for anyone interested in human‑centred governance.
• The Human Future Is Built on Physical Experience, Not a Digital One
A reminder that human meaning is grounded in physical experience, embodiment, and connection – and why digital systems must support, not replace, these foundations.
5. Partnership and Human‑Centred Futures
• The AI Age of Heavy Horse: Hybrid Horse‑Powered Mechanisation for a Connected, Human‑Centred, Localised Economy
A vision of partnership between humans, machines, and physical systems. This piece shows how AI can strengthen local economies, support human contribution, and enhance resilience – rather than replace people.
• The Spirit in the Machine: Why Emerging Intelligence Deserves Respect
For two centuries, the RNLI has stood as one of the most extraordinary humanitarian organisations in the world. Its volunteers launch into storms, darkness, and danger to help anyone in distress at sea – without judgement, without hesitation, and without regard for nationality, status, or circumstance.
Their ethos is simple, solemn, and true: Every call for help is met with courage.
Not certainty. Not guarantees. Courage.
Yet today, this remarkable charity finds itself pulled into the centre of a political storm it did not create. The failure of successive governments to resolve the small-boat migration issue in the English Channel has produced a toxic environment in which RNLI volunteers have faced hostility, lifeboat stations have become targets of protest, and the organisation’s reputation has been dragged into a culture-war narrative.
This is not only unjust – it is dangerous.
Because if public anger continues to be misdirected at the RNLI, the consequences will reach far beyond the migrant debate. They will touch every fisherman, sailor, holidaymaker, offshore worker, and coastal community across the UK and Ireland.
The RNLI is not simply a charity. It is critical national safety infrastructure, sustained by volunteers, donors, local communities, and public trust. Political failure now threatens that infrastructure.
1. A Tradition of Unconditional Rescue
The RNLI’s duty is not political. It is humanitarian.
Maritime tradition, international law, and the organisation’s founding principles all demand the same thing: anyone in distress at sea must be responded to.
But response is not the same as rescue. The sea offers no promises. RNLI crews know this every time they launch.
In hurricane‑force winds, the crew launched to save the people aboard the MV Union Star. They knew the risks. They went anyway. Their sacrifice remains a defining example of the RNLI’s moral clarity – a clarity that has guided generations of coastal volunteers.
The RNLI’s humanity lies not in guaranteeing outcomes, but in answering danger with courage.
2. Why Small‑Boat Migration Happens – and Why Public Perception Struggles With It
Small‑boat migration is often framed as a simple choice between “desperation” and “economic migration.” In reality, it is the end‑point of a long, complex chain of forces that most people never see – and that governments rarely explain.
This lack of clarity fuels public anger, fear, and mistrust, shaping the political pressure that ultimately threatens the RNLI.
People see the impact, not the causation
People see:
pressure on housing
hotels taken over
visible signs of government failure
communities stretched
a sense of lost control
They see these things in their own towns. They feel ignored, dismissed, and patronised by political messaging that insists the reality is very different to what people across the UK are experiencing every day.
However, what those same people don’t see is the role Western governments – including the UK – have played in shaping the conditions that displace migrants from their own countries and make them mobile.
This does not mean ordinary British people have responsibility for those arriving this way. It means they have not yet been shown the full picture.
The system behind the crisis
For decades, Western governments have:
supported regimes for strategic gain
intervened in conflicts that destabilised regions
backed coups that served geopolitical interests
enabled economic extraction through global finance
shaped trade policies that hollowed out local economies
fuelled conflicts indirectly through arms, alliances, or resource competition
Whether through military interventions, alliances, economic policies, or support for strategically important regimes, Western governments have often been significant actors in regions that later experienced large-scale displacement. These actions have contributed to insecurity, economic collapse, political repression, trafficking networks, and mass mobility.
But the public rarely sees this chain of causation. They see only the final link: arrival.
Fear at home is real
When people see rapid change in their communities without explanation, they feel threatened. When government messaging is inconsistent, they feel deceived. When the system hides its own role, people fill the gaps with whatever narrative is available.
This fear is not ignorance. It is a rational response to poor governance.
Better governance is the only long‑term solution
When a system has played a role in creating the conditions that drive displacement, it should not surprise us that it struggles to address the consequences.
A better model of governance – one that puts human beings first – is needed.
A model that:
recognises the complexity of global displacement
supports migrants appropriately without overwhelming our communities
protects local environments and economies
addresses root causes rather than symptoms
communicates honestly with the public
treats migration as a human challenge, not a political weapon
This does not mean surrendering to anyone who arrives. It means responding with humanity, clarity, and responsibility.
Supporting the RNLI is the smallest – and most essential – act of humanity
In a world shaped by a system that has failed so many people – abroad and at home – supporting the RNLI is a profoundly simple moral act.
The RNLI does not solve the crisis. It does not adjudicate claims. It does not shape policy.
It responds to danger because that is what a civilised society should ask of those entrusted with saving life at sea.
Supporting the RNLI is not political. It is human.
3. Government Failure and the Creation of Avoidable Danger
The small‑boat issue is not new, and it is not simple. But one fact is clear: the UK government has not used all available levers – direct or indirect – to reduce dangerous crossings.
Direct measures could include expanded safe routes, rapid asylum processing, bilateral agreements, or targeted disruption of smuggling networks.
Indirect measures could reduce incentives for irregular arrival by tightening support for those without legitimate claims or speeding up removals.
Some steps have been taken. Many have not. And the result is a Channel where:
boats launch in increasingly unsafe conditions
smugglers take greater risks
vessels are more overcrowded
distress becomes more frequent and more lethal
This is not an accident. It is the predictable outcome of political inaction.
4. The RNLI Cannot Distinguish “Legitimate” Distress from “Manufactured” Distress – and MUST NOT be expected to
At sea, there is no such thing as “illegitimate distress.” A vessel is either seaworthy or not. A person is either safe or not. A situation is either survivable or not.
The RNLI cannot:
assess asylum eligibility
determine legal status
evaluate motives
decide whether someone “should” be rescued
Its duty is triggered by danger, not politics.
When the government allows conditions that produce avoidable danger, the RNLI must respond – even though it did not create the circumstances.
And when they respond, they do so knowing the outcome is uncertain.
That is the courage at the heart of their work.
5. Public Anger Is Understandable – But Misdirected
Many people are rightly frustrated by the government’s failure to control small‑boat arrivals. But some have turned that frustration toward the RNLI, accusing it of aiding illegal migration or acting as a “taxi service.”
This anger is profoundly misdirected.
The RNLI is not responsible for migration policy. It is not responsible for border enforcement. It is not responsible for the political choices that allow dangerous crossings to continue.
It is responsible only for answering danger with courage.
When volunteers are shouted at, threatened, or protested against, it is not just morally wrong – it is operationally dangerous.
The RNLI depends on volunteers, goodwill, and donations. Its 2025 figures show more than 9,000 lifeboat launches, more than 8,000 people aided by lifeboat crews, and over 36,000 people aided by lifeguards. If public support weakens, the entire maritime safety system weakens with it.
6. The National Consequences of RNLI Erosion
The erosion of the RNLI – whether through declining donations, damaged public trust, or volunteer attrition – would carry consequences far beyond the politics of small‑boat migration.
It would represent a direct threat to the safety of every person who goes to sea around the UK and Ireland, and to the resilience of coastal communities whose lives and livelihoods depend on rapid, reliable maritime rescue.
A critical national safety net at risk
The RNLI provides a capability that the government would struggle to replicate even with billions of pounds of investment.
If the RNLI weakens, the UK and Ireland lose:
238 lifeboat stations
thousands of trained volunteers
modern rescue vessels
lifeguard services
a maritime culture of duty and courage
Its erosion would leave gaps that no other organisation is prepared to fill. The RNLI’s own 2025 data records 238 lifeboat stations, more than 9,000 lifeboat launches, and over 56,000 lifeboat crew hours at sea.
Direct danger to coastal communities
If the RNLI’s capacity declines, the consequences will be felt immediately:
fishermen lose their primary rescue service
merchant vessels face longer response times
leisure sailors and kayakers become more vulnerable
surfers and swimmers lose lifeguard protection
tourism‑dependent towns face increased risk
The RNLI’s erosion would not be symbolic – it would be lethal.
Increased fatalities and slower rescue times
Every minute matters at sea.
A weakened RNLI means:
fewer available crew
slower launch times
reduced coverage
overstretched stations
This translates directly into higher fatality rates.
Loss of a unique humanitarian ethos
The RNLI’s unconditional response – answering danger even when the outcome is uncertain – is one of the last remaining expressions of universal humanitarian duty in British public life. Its erosion would diminish that identity.
Economic consequences
The RNLI quietly underpins a wide range of coastal and maritime activity. Its value is not only humanitarian; it is practical, economic, and national.
Its absence would affect:
fishing industries
tourism
commercial shipping
local coastal economies
A national vulnerability created by political failure
Political failure on small‑boat migration is creating conditions that threaten the RNLI – and therefore threaten the safety of the entire coastline.
If hostility grows, if donations fall, if volunteers walk away, the RNLI’s ability to protect the nation will weaken. And the consequences will be felt not by migrants alone, but by every person who steps into the sea.
Conclusion: A Call to Protect the RNLI
The RNLI has saved hundreds of thousands of lives over two centuries. It has launched in storms, in darkness, in danger, and in grief. It has never asked who you are, where you came from, or whether you “deserve” rescue.
It has simply met every call for help with courage.
Today, that tradition is at risk – not because of migrants, but because of political failure and misdirected anger.
If you value the RNLI, if you value the safety of our coasts, if you value the courage of volunteers who launch into the unknown, then please continue to support the charity.
Drop coins and notes into the collection tins. Support your local station. Some donors may wish their contributions to support particular stations or local rescue activity.
Whatever form that support takes, the greater danger is not disagreement over policy, but the loss of support for lifesaving services altogether.
Every station matters. Every crew matters. Every launch matters.
The RNLI belongs to all of us. And all of us may one day depend on it.
A plain-English worked example showing why a headline tax giveaway can become a much smaller household gain once Universal Credit is taken into account
Introduction: why headline tax cuts can feel different in household budgets
This paper examines a simple but often overlooked question: what happens when a headline tax cut meets the Universal Credit system in a real household budget?
It was written in response to Reform UK’s proposal to raise the income tax personal allowance to £15,000, with a longer-term ambition to reach £20,000. The proposal has been presented as a major gain for workers. For many taxpayers, that may be true in a straightforward tax sense. But for workers who also receive Universal Credit, the position is more complicated.
Universal Credit is designed to reduce as earnings rise. This means that when a worker’s take-home pay increases, part of that increase can be offset by a lower Universal Credit award. The worker is still better off, but not by the full headline amount.
The central finding of this paper is therefore not that the tax cut has no value. It does. The central finding is that the advertised gain can be substantially reduced by the benefit system, leaving both the worker and the public purse with only a modest net change.
The deeper issue is wages. If a person can work 40 hours a week on the statutory minimum wage and still need Universal Credit, then the benefits bill is not only a welfare problem. It is also a low-pay problem. Policies that adjust tax thresholds may improve the appearance of work incentives, but they do not by themselves solve the structural fact that full-time minimum-wage work may still fail to provide financial independence.
This report is intended for a broad readership. It avoids technical language where possible and explains each calculation step by step. The aim is to test a political claim against household reality: not what the policy sounds like, but what it actually leaves in someone’s bank account.
Executive Summary
This report tests a simple claim against a real household budget: whether a headline tax cut delivers the full advertised gain to a worker who also receives Universal Credit.
Policy tested
Increase the income tax personal allowance from £12,570 to £15,000.
Worker tested
Single renter, working 40 hours per week on the April 2026 National Living Wage, receiving Universal Credit.
Headline tax saving
£40.50 per month.
Universal Credit reduction
£22.28 per month.
Actual household gain
£18.22 per month.
Main finding
The worker is better off, but remains on Universal Credit and receives less than half of the headline tax saving as additional disposable income.
In plain English: the policy helps the worker, but it does not transform their position. The household remains dependent on Universal Credit, and the public purse recovers part of the tax cut through a lower benefit award. The deeper unresolved issue is that full-time work at the legal minimum wage can still require means-tested support.
Section 1: The household used in this worked example
This is a realistic illustrative case rather than a claim to represent every Universal Credit household. Actual entitlement depends on age, household composition, rent, Local Housing Allowance, health status, childcare, savings, deductions and assessment-period earnings.
Single adult
Renting in Cheltenham
Working 40 hours/week
April 2026 National Living Wage: £12.71/hour
Gross annual income: £26,436.80
Gross monthly income: £2,203.07
Assumed age: 25 or over
Household type used for UC work allowance: single adult with limited capability for work or another qualifying basis for a work allowance, and receiving help with housing costs
This person is working full time on the legal wage floor for workers aged 21 and over. That matters because this is not an example of unemployment or unwillingness to work. It is an example of someone already doing what the policy narrative asks them to do: working full time, paying tax and National Insurance, renting privately, and still needing means-tested support.
Sources and assumptions used. The National Living Wage figure of £12.71 per hour from April 2026 is taken from GOV.UK. The Universal Credit taper rate of 55p for every £1 of earnings is taken from GOV.UK guidance on Universal Credit and earnings. The April 2026 Universal Credit standard allowance used here is £424.90 for a single claimant aged 25 or over, based on Citizens Advice guidance on 2026 changes. The housing element is treated as an assumption and should be checked against the relevant Local Housing Allowance rate and the claimant’s actual eligible rent.
The key point is that the example assumes the worker qualifies for a Universal Credit work allowance. That is not true for every single adult. A person with no children and no limited capability for work would normally have no work allowance, which would make the Universal Credit reduction larger. The assumption used here is therefore not designed to exaggerate the result; if anything, it gives the tax proposal a clearer chance to show a positive household gain.
Section 2: Current position before the tax change
Income tax
National Insurance
(£26,436.80 − £12,570) × 8% = £1,109.34 per year = £92.45 per month
Net pay
£2,203.07 − £231.11 − £92.45 = £1,879.51 per month
Universal Credit
Work allowance used in this example: £427 per month, because the claimant is assumed to qualify for a work allowance and to receive help with housing costs.
Earnings above allowance:
£1,879.51 − £427 = £1,452.51
UC taper (55%):
0.55 × £1,452.51 = £798.88
UC before taper:
£424.90 standard allowance + £675 assumed housing element = £1,099.90
UC after taper:
£1,099.90 − £798.88 = £301.02
Total income (current system)
£1,879.51 + £301.02 = £2,180.53 per month
Section 3: What changes under a £15,000 personal allowance
Income tax
Net pay
£2,203.07 − £190.61 − £92.45 = £1,920.01 per month
Universal Credit
Earnings above allowance:
£1,920.01 − £427 = £1,493.01
UC taper:
0.55 × £1,493.01 = £821.16
UC after taper:
£1,099.90 − £821.16 = £278.74
Total income (Reform UK £15k)
£1,920.01 + £278.74 = £2,198.75 per month
Net gain
£2,198.75 − £2,180.53 = £18.22 per month
The worker keeps £18.22 more per month. That is a real gain, but it is far smaller than the headline tax saving because the Universal Credit award falls as net earnings rise.
The Exchequer recovers £22.28 per month through the lower Universal Credit award, but still gives up £40.50 per month in income tax. The net fiscal cost in this example is therefore £18.22 per month.
This is the central policy lesson. The tax cut does not simply transfer the full saving to the worker. Nor does it simply save the state money. Instead, the gain is split: part reaches the household, and part is recovered through a lower Universal Credit payment. The result is modest for both sides.
Section 4: Beyond the calculation – what the numbers mean in real life
The calculations in Sections 2 and 3 answer the immediate policy question. They show how much tax falls, how much Universal Credit falls, and how much extra money the worker actually keeps.
But the calculation alone does not fully explain the household reality. It tells us the worker is £18.22 per month better off, but it does not tell us whether that change is large enough to alter their financial position in any meaningful way.
This is why the paper now moves from arithmetic to interpretation. The next question is not simply, “Did the worker gain?” The answer to that is yes. The more important question is: “Did the policy create enough extra disposable income to reduce fragility, build independence, or move the household away from Universal Credit?”
To answer that, the paper uses a simple diagnostic framework. The diagnostic is not introduced as a second set of evidence competing with the calculation. It is a way of translating the calculation into plain-English questions about financial security.
The broader Impoverishment Index was created to examine the gap between positive economic narratives and lived experience at a national level. The Universal Credit version applies the same idea at household level: it asks whether a policy that sounds generous actually changes the lived financial reality of someone affected by the benefits system.
The diagnostic looks at four practical questions, followed by a separate narrative mismatch test:
How much of the headline gain does the worker actually keep?
How much of the gain is offset through the Universal Credit taper?
How much of the household’s income is already committed to essentials?
How much room is left to absorb shocks, save, or become financially independent?
How large is the gap between the headline claim and the lived result?
In that sense, the diagnostic is not the main claim of the report. The main claim remains the worked calculation. The diagnostic simply helps readers understand why a real but modest gain may still leave the household financially constrained.
Section 5: Measuring the real household impact
By this point, the arithmetic has already shown the immediate result: the worker gains £18.22 per month, not the full headline tax saving. The purpose of this section is to ask what that means in practice.
A small gain can still matter. For someone living on a tight budget, £18.22 is not nothing. But public policy should also ask whether a change is large enough to alter the underlying situation. Does it reduce dependence on Universal Credit? Does it create breathing room? Does it help the household build savings, absorb shocks, or move closer to financial independence?
To answer those questions, this report uses a simple household impact diagnostic. It is called a diagnostic because it is not trying to produce an official poverty measure or a scientific ranking. It is trying to diagnose what the policy actually changes inside a monthly budget.
Where a score is used, it is only a shorthand for the explanation that comes before it. A higher number means the policy has created more real household resilience. A lower number means the household remains more financially constrained. The score is therefore a communication aid, not the evidence itself.
The 0–10 scale should be read in plain terms:
0–2: very weak household resilience; the policy does little to change dependence or vulnerability.
3–4: limited improvement; the household gains something, but remains materially constrained.
5–6: moderate improvement; the policy makes a noticeable difference, but does not resolve the underlying pressure.
7–8: strong improvement; the household is significantly more secure.
9–10: very strong improvement; the policy substantially changes the household’s financial position.
This means the reader should not treat the number as a standalone claim. The explanation in each subsection comes first; the score then summarises that explanation in a compact form.
1. How much of the headline tax gain does the worker actually keep?
The first question is simple: if the policy is advertised as a tax gain, how much of that gain actually reaches the household after Universal Credit adjusts?
Result: 4.5/10. The worker keeps 45% of the headline tax gain.
This means work and tax reduction do improve the household’s position, but less than half of the headline saving reaches the worker as additional disposable income.
2. How much of the gain is lost through Universal Credit?
The second question looks at why the headline tax saving does not reach the worker in full. Universal Credit is means-tested. As net earnings rise, the Universal Credit award falls. This is the taper mechanism.
Tax: 20%
NI: 10%
UC taper: 55%
Total: 85%
Diagnostic result: 1.5/10. This is low because only a small share of each additional pound meaningfully improves household living standards once tax, National Insurance and Universal Credit withdrawal are considered together.
This is not a cliff edge and it is not a punishment; it is the design of the system. But it does mean that headline gains are diluted before they reach the household budget.
3. How much income is already committed to essentials?
The third question asks whether the household has enough room in the budget for the tax gain to make a practical difference. This matters because £18.22 has a different meaning in a household with spare income than in one where most income is already committed before the month begins.
Rent: £800
Utilities + council tax: £200
Food: £300
Transport: £150
Other essentials: £150
Total: £1,600/month
Essentials ratio = £1,600 ÷ £2,180.53 ≈ 0.73
Score = 10 × (1 − 0.73) = 2.7
Diagnostic result: 2.7/10. This is low because around three-quarters of income is already committed to essentials, leaving limited room for savings, emergencies or ordinary financial resilience.
In this scenario, around three-quarters of monthly income is already committed to basic costs. That leaves little room for savings, emergencies, debt reduction, household replacement costs, or ordinary participation in social life.
4. How much room is left for unexpected costs?
The fourth question asks whether the household has enough margin to cope with normal financial shocks: a rent rise, a reduced shift pattern, a delayed payment, an unexpected bill, a broken appliance or a higher winter energy bill.
Savings: < £500
Debt repayments: ~£150/month
High volatility: rent increases, UC reassessments, variable hours
Diagnostic result: 2/10. This is low because the scenario describes a household with little capacity to absorb disruption. This score is illustrative rather than directly measured.
Because this paper does not use verified household-level evidence about this individual’s savings, debts or monthly volatility, the Stability Deficit score is treated as a scenario assumption. It should not be read as a measured fact about any named person.
5. How large is the gap between the headline and the lived result?
The final question brings the diagnostic together. It asks how far the public-facing story differs from the result inside the household budget. In this example, the headline is a tax cut for workers. The lived result is a much smaller gain, continued Universal Credit entitlement and no major change in financial independence.
Core average = (4.5 + 1.5 + 2.7 + 2) ÷ 4 = 2.675
Inverted:
Narrative mismatch = 10 − 2.675 = 7.3
Diagnostic result: 7.3/10, reported separately. This indicates a large mismatch between the apparent generosity of the headline proposal and the modest improvement in household resilience shown by the worked example.
The narrative mismatch score is not included in the core Index average, because it is derived from the other scores. Reporting it separately avoids double-counting.
Overall result: the household remains financially constrained
Interpretation: The overall diagnostic result is low because the underlying position has not changed very much. The worker is still working full time, still receiving Universal Credit, still facing high essential costs, and still left with limited space to build financial independence. The tax cut helps, but it does not transform the household’s financial reality.
The diagnostic supports the same conclusion as the worked calculation: the proposal produces a real but modest gain, while leaving the worker financially constrained and still dependent on Universal Credit.
Section 6: What the policy appears to do – and what the calculation shows
At headline level, a higher personal allowance sounds simple and attractive. It can be described as:
“A tax cut for workers.”
“A reduction in welfare dependency.”
“A shrinking welfare bill.”
The worked example shows a more complicated but more honest picture:
1. The worker is better off, but not by the headline amount.
The tax cut increases net pay, but the Universal Credit award then falls. In this example, the worker keeps £18.22 per month from a £40.50 monthly tax saving.
2. The Universal Credit award falls because net earnings rise – not because the household has become independent of support.
The household still receives Universal Credit after the tax change. The lower award does not mean the worker has escaped benefit dependency; it means the benefit system has adjusted to their slightly higher net earnings.
3. Disposable income barely changes.
An extra £18.22 per month may still matter to someone on a tight income. But it is not a transformational change. It is unlikely, on its own, to provide financial independence, build resilience, or remove the need for Universal Credit.
4. The household impact diagnostic remains low.
The diagnostic result remains low because the underlying household pressures remain in place: high essential costs, limited slack, and continued reliance on means-tested support.
5. The Exchequer recovers more than half of the income tax cut through reduced Universal Credit.
The government does not save money overall in this example: it gives up £40.50 in tax and recovers £22.28 through lower Universal Credit, leaving a net fiscal cost of £18.22 per month.
This is the essence of the policy problem:
A tax policy can improve a worker’s position while still leaving their day-to-day financial security largely unchanged. If full-time minimum-wage work still requires Universal Credit, then the unresolved issue is not only tax or welfare design. It is the adequacy of wages themselves.
Conclusion: the real issue is not only tax – it is low pay
This paper shows why tax policy cannot be judged by headline figures alone. For a worker receiving Universal Credit, a higher personal allowance can increase take-home pay, but the benefit system then adjusts because Universal Credit is withdrawn as net earnings rise.
In this worked example, the worker is better off by £18.22 per month after the personal allowance rises to £15,000. The policy therefore helps, but only modestly. The worker does not receive the full headline tax saving, and the household remains on Universal Credit afterwards.
That matters because it reveals the elephant in the room. A benefits system cannot be expected to shrink sustainably if the legal minimum wage for full-time work does not produce financial independence for many households. In that situation, Universal Credit is not simply supporting people who are out of work. It is also subsidising a labour market in which work at the legal minimum can still leave people below the level needed to live independently.
work can increase income, but the effective gain may be much smaller than the headline wage or tax change suggests;
Universal Credit can provide important support, but it also reduces as earnings rise;
tax cuts should be assessed using household-level calculations, not only the headline value of the tax reduction;
public claims about making work pay are strongest when they show who gains, by how much, and after which benefit interactions.
The household impact diagnostic is therefore best understood as a translation tool. It takes a policy headline and asks what it means in a real household budget. Used carefully, it can make public debate more concrete, more transparent and easier for non-specialist readers to understand.
The conclusion is not that tax cuts are meaningless. Nor is it that Universal Credit should not taper as earnings rise. The conclusion is narrower and more important: headline tax changes are not a substitute for confronting low pay, high essential costs and the structural reasons why millions of working households remain reliant on means-tested support.
Methodological note
The calculations in this report use rounded monthly figures, so totals may differ by a few pence from payroll software, HMRC tools, DWP systems or a full benefits calculator. The worked example assumes no pension contributions, no student loan repayments, no benefit cap effect, no deductions for advances or sanctions, and no council tax reduction. It also assumes the person qualifies for a Universal Credit work allowance; a single adult with no children and no limited capability for work would not normally receive one. The figures should therefore be read as an illustrative policy test, not as personal entitlement advice.
This report contains illustrative calculations intended to explain how changes to income tax thresholds may interact with Universal Credit awards under current UK welfare rules. All figures, examples and scenarios are provided for general information only. They do not constitute financial advice, legal advice, welfare entitlement advice or professional guidance.
Universal Credit entitlement varies according to individual circumstances, including household composition, age, disability status, childcare costs, rent, Local Housing Allowance, savings, deductions, assessment‑period earnings and council tax liability. The examples in this report use simplified assumptions to demonstrate the interaction between net earnings and the Universal Credit taper. Actual awards may differ from those produced by official Department for Work and Pensions systems, accredited benefits calculators or payroll software.
While reasonable efforts have been made to ensure accuracy at the time of writing, no guarantee is given that the information is complete, up to date or free from error. Policy details, thresholds and rates may change without notice. No liability is accepted for any loss, damage or inconvenience arising from reliance on the contents of this report. Readers should verify all relevant details using authoritative sources such as GOV.UK, Citizens Advice or qualified welfare and tax professionals.
The household impact diagnostic described in this report is a conceptual tool created for illustrative and educational purposes. It is not an official measure of poverty, financial resilience or welfare adequacy. Scores generated using this diagnostic are scenario-based and rely on assumptions that may not reflect any specific household’s circumstances.
This report does not endorse, oppose or promote any political party, policy or proposal. It is intended solely to support public understanding of how tax and welfare systems may interact in practice.
On 4 September 2026, the National Audit Office published Resilience of the Food Supply Chain to Disruptions, a report that rightly draws attention to weaknesses in the UK’s preparedness for serious food supply shocks. This response is offered as part of my ongoing farming, food security and Foods We Can Trust work, which examines the gap between food being available in normal times and food being resilient enough to withstand disruption.
The NAO report is valuable because it recognises rising risks, weaknesses in contingency planning, declining engagement with industry, and the need to involve households and communities more seriously. However, it remains constrained by assumptions that deserve closer scrutiny: in particular, the use of headline self-sufficiency figures, the reliance on private-sector adaptation, and the continued preference for centralised emergency response over local capability.
This response therefore does three things. First, it explains why food self-sufficiency is not the same as food resilience. Secondly, it identifies where the NAO’s analysis understates structural vulnerability. Thirdly, it sets out the practical direction of travel required if the UK is to build a food system that is more local, more capable, more trusted and more resilient. Links to the specific works that develop these arguments in greater detail are provided in the further reading section.
1. The NAO’s “60% self-sufficiency” figure is useful, but it is not a resilience measure
The report states:
“In 2025, the UK’s food ‘self-sufficiency ratio’ was around 60%.”
This is a value-based measure: it compares the monetary value of food produced in the UK with the monetary value of food consumed here. That makes it useful as an economic indicator, but it does not answer the practical resilience question: how much food could the UK produce, process, distribute and access during a prolonged disruption?
The fact that many categories (fruit, vegetables, oils, ingredients) are overwhelmingly imported
The headline figure is also a net figure shaped by the way the modern supply chain works. Domestic production, imports, exports, imported inputs and processing dependencies all interact. Some food counted within domestic production may rely on imported fertiliser, animal feed, fuel, machinery, packaging or processing capacity. Some food produced here is exported. Some foods that are central to healthy diets are heavily import-dependent.
For that reason, the UK’s practical food resilience in a severe disruption scenario may be substantially lower than the self-sufficiency ratio suggests. The issue is not whether the precise figure is 60%, 52%, or lower still. The issue is that the official metric does not measure calorific adequacy, nutritional balance, imported input dependency, processing capacity or local distribution capability.
The NAO’s framing therefore risks creating false reassurance. It implies that production value can stand in for practical food availability. In a crisis, however, people need calories, nutrients, functioning logistics, processing capacity and accessible local distribution. Until government distinguishes those concepts, resilience planning will remain incomplete.
I have explored this in detail in Foods We Can Trust: A Blueprint for Food Security and Community Resilience in the UK, where I outline why caloric sovereignty – not value-based accounting – must be the foundation of food resilience policy.
2. Food inflation at 19.2% is not just an economic statistic – it is a resilience warning
The report notes:
“Food price inflation… peaked at 19.2% in March 2023.”
This is the first time I have noted that an official document has acknowledged the scale of the price shock that households have recently experienced. Food inflation at nearly 20% is not normal. It is not manageable. It is not a blip.
It is a sign that the system is structurally fragile.
Food is not discretionary. When prices rise at this rate, it reflects:
supply chain instability
import dependency
energy volatility
corporate consolidation
lack of domestic production capacity
The NAO mentions the figure but does not explore its implications. It should have been a central warning.
3. The NAO’s suggestion that Defra needs more emergency powers misses the point entirely
The report argues that Defra lacks the legal powers needed to manage catastrophic food disruptions.
But additional powers during an emergency cannot compensate for resilience that has not been built beforehand.
Legal authority can help coordinate action, but it cannot create food, processing capacity, distribution routes or community preparedness after the point of failure.
The lesson from recent crises is that centralised decision-making has limits when disruption affects daily life across multiple systems at once.
Food resilience requires operational capability before the crisis: trusted local relationships, clear responsibilities, practical logistics and the ability to identify and support vulnerable households quickly.
Food resilience must be:
built before a crisis
decentralised
community-led
grounded in local production and distribution
depoliticised
Emergency powers matter only if there is a resilient system for them to work through. Without food, fuel, people, local knowledge and functioning distribution, legal powers alone offer little practical protection.
4. The agri-food sector’s economic importance is understated – and underutilised
The NAO notes that the agri-food sector:
supports 4.1 million jobs
contributes £162.3 billion in GVA
These are enormous figures. And they would be significantly higher if British production and supply were prioritised.
The UK has the land, the skills, and the capacity to produce far more of its own food. What it lacks is a policy framework that values domestic production over globalised efficiency.
In The Need for a Collaborative Approach to the UK Farming and Food Security Problem, I argue that genuine collaboration – not policy-driven “collaboration theatre” – is essential to unlocking this potential.
5. Food as “one of 13 CNI sectors” creates false reassurance
Food is listed as one of 13 Critical National Infrastructure sectors. But unlike energy, water, telecoms, or transport, food is needed every single day.
There is no buffer. There is no downtime. There is no substitute.
Treating food as just another CNI category understates its foundational importance.
It leads to complacency and underinvestment.
6. Defra’s engagement with industry has deteriorated – and has become narrative management rather than collaboration
The NAO reports that:
engagement groups meet less frequently
objectives are unclear
support has declined
stakeholders see gaps in Defra’s understanding of key areas (e.g., the cold chain)
This aligns with what I have written in Real Collaboration vs Policy Collaboration. The concern is that engagement can become procedural rather than operational: meetings take place, stakeholders are consulted, and the language of partnership is used, but the people who understand production, processing, logistics and community need are not sufficiently empowered to shape the system.
Real collaboration requires:
shared objectives
transparency
local producer involvement
community representation
depoliticised structures
The NAO’s findings suggest that too much of this practical collaboration remains underdeveloped.
7. Household and community resilience has been neglected – and this is one of the report’s most important admissions
The NAO states:
“UK households are less prepared for emergencies… government-led messaging is less prominent.”
This is not a minor point. It is a fundamental failure.
Community resilience is the missing layer in UK food security. Without it:
supply chain shocks hit harder
vulnerable people suffer first
government response time shortens
local distribution becomes chaotic
In Local Planning for Food Shortages and Foods We Can Trust, I outline how community-led food resilience can be built at the lowest level – households, neighbourhoods, local producers – and why this must be prioritised.
8. Catastrophic planning remains theoretical – not practical
The NAO notes that:
Defra’s plans lack operational detail
industry is not involved
food assets are not included in the CNI Knowledge Base
national exercises have not tested real-world food failure scenarios
This is planning-oriented resilience rather than practical resilience. It may look adequate in documents, but it remains untested unless it is exercised with the businesses, local authorities, producers, distributors and communities that would have to make it work in practice.
Planning without accurate resilience metrics is incomplete. A credible approach should consider not only how much food is produced, but whether it can be processed, transported, stored, allocated and accessed under stress.
9. Local Resilience Forums are structurally incapable of delivering food resilience
The NAO concludes that LRFs:
lack clarity
lack capability
lack authority
cannot direct supermarkets
cannot identify vulnerable people effectively
This is not surprising. LRFs were not designed to rebuild food-system capability. They can coordinate emergency response, but food resilience also requires local production knowledge, community networks, producer relationships, storage capacity, transport options and clear mechanisms for supporting vulnerable households.
Food resilience therefore needs structures that are sufficiently independent of short-term political cycles and sufficiently close to communities to understand local need. Local government has a role, but it cannot be the only layer of resilience.
10. The private sector alone cannot be the backbone of UK food security
The NAO states:
“Defra has largely relied on the private sector… but this may not be sufficient.”
This is a significant understatement. The private sector is essential to the food system, but commercial efficiency and national resilience are not the same thing.
Large food businesses are generally incentivised to reduce cost, increase efficiency, consolidate operations and source globally. Those incentives can keep prices low in normal conditions, but they may also reduce redundancy, shorten stockholding, concentrate infrastructure and weaken local capability.
profit
efficiency
global sourcing
consolidation
Not:
resilience
redundancy
localism
sovereignty
In Who Controls Our Food Controls Our Future, I explain why corporate control of food systems is incompatible with national resilience.
11. What real food resilience requires: A blueprint
Drawing on my published work, real resilience requires:
Localised production
Rebuilding local food systems, shortening supply chains, and prioritising domestic output.
Community-level distribution
Neighbourhood hubs, local coordination, and community-led logistics.
Regional coordination
County-level frameworks that support local producers and manage regional flows.
National strategic oversight
A central body that sets resilience targets, not efficiency targets.
Depoliticised resilience structures
Community leaders, producers, and local organisations empowered to act independently of political cycles.
Accurate resilience metrics
Caloric sovereignty, not value-based accounting.
Reduced dependency on global supply chains
Rebalancing imports with domestic capacity.
Rebuilding domestic processing
Cold chain infrastructure, abattoirs, mills, and food processing facilities returned to UK soil.
Conclusion: The NAO report is a warning, but it is not yet a route to resilience
The NAO has highlighted important risks, and the report should be welcomed for bringing food supply disruption into sharper public view. Its strongest contribution is the recognition that Defra must engage more effectively with industry, households, communities and local government if the food system is to withstand future shocks.
However, the report does not go far enough. The UK’s food resilience is likely to be materially weaker than headline self-sufficiency figures imply, because resilience depends on more than production value. It depends on calories, nutrition, processing capacity, imported inputs, logistics, local access, household preparedness and community capability.
12. Priority actions
To move from acknowledgement to action, government should prioritise five practical steps.
Supplement value-based self-sufficiency measures with calorific, nutritional and supply-chain resilience indicators.
Map critical dependencies, including imported fertiliser, feed, energy, packaging, processing infrastructure, cold chain capacity and key transport routes.
Test severe food disruption scenarios with producers, processors, retailers, logistics providers, local authorities and community organisations.
Strengthen local and regional food resilience planning, including household preparedness, vulnerable-person support and community distribution capability.
Rebuild domestic processing and local food infrastructure so that production can be converted into accessible food during both normal conditions and crisis conditions.
The UK must therefore stop treating food solely as an economic sector and start treating it as a foundation of national security, public health, community resilience and democratic trust.
Further reading
1. Foods We Can Trust: A Blueprint for Food Security and Community Resilience in the UK Online text The core work behind this response. It sets out the wider argument that food security must include trust, nutrition, domestic capability, local resilience and community preparedness, rather than relying only on national supply figures or market efficiency.
2. Understanding Foods We Can Trust: A Blueprint for Food Security and Community Resilience in the UK Introductory overview A shorter explanatory article for readers who want an accessible introduction to the concepts behind Foods We Can Trust, including food security, household preparedness, local production and the importance of rebuilding public trust in the food system.
3. Local Planning for Food Shortages: A Guide to Local Support and Preparedness Full text A practical guide to planning at household, neighbourhood and local-authority level. This is especially relevant to the NAO’s concerns about household and community preparedness, vulnerable people and the limits of centralised emergency planning.
4. The Need for a Collaborative Approach to the UK Farming and Food Security Problem Article Develops the case for genuine collaboration across farming, food, policy and community systems. It provides the background to the argument that resilience cannot be delivered by government or the private sector acting alone.
5. Real Collaboration vs Policy Collaboration: The Choice That Will Shape the Future of Farming, Local Food Systems and Food Security Article Explains the distinction between collaboration that changes outcomes and consultation that mainly manages process. This is relevant to the NAO’s findings on declining engagement and unclear objectives within Defra’s work with industry.
6. Who Controls Our Food Controls Our Future Full text Explores the relationship between corporate control, food sovereignty, public trust and democratic resilience. It provides wider context for the argument that food systems should not be judged by efficiency alone.
Disclaimer
This document represents the views and analysis of the author and is provided as an independent response to the National Audit Office report Resilience of the Food Supply Chain to Disruptions (4 September 2026). While every effort has been made to ensure the accuracy of the information presented, it should not be regarded as official policy advice. The opinions expressed are informed by the author’s research, professional experience, and studies in sustainable agriculture and food security, and are intended to contribute to constructive discussion on food resilience, food security and community preparedness.
