In Britain, workplace injuries and new cases of work-related ill health arising from current working conditions cost an estimated £22.9 billion in 2023/24. Employers bore about £4.3 billion of that total. Government bore £5.2 billion. Individuals carried an estimated £13.4 billion. That last figure includes an economic valuation of pain, suffering and lost quality of life as well as direct financial losses, so it should not be mistaken for £13.4 billion literally taken from workers’ pockets. But the imbalance remains striking.
A company makes an economic decision and receives the profit. Yet the consequences of that decision may be shared by workers, families, the NHS, taxpayers, communities or people who had no part in making it. Who, then, should pay? And what happens when the cheapest decision for a company is not the safest, healthiest or fairest decision for everyone else? That leads to a larger question.
Is capitalism unethical?
There are two easy answers. One says yes: capitalism puts profit before people. The other says no: capitalism has produced extraordinary innovation, prosperity and improvements in living standards. Both contain some truth. Neither is enough. The harder and more useful question is this:
What happens when what is profitable and what is right point in different directions?
The history of health and safety offers an unusually clear way of investigating that question. But the same conflict appears in employment, consumer markets, healthcare, pollution and even the political rules that govern capitalism itself. To understand why, we first need to separate two ideas that are often allowed to blur together.
Price and value.
What capitalism can tell us
Capitalism is not one fixed model. Britain, Germany, Sweden, Japan and the United States are all capitalist economies, yet their institutions differ greatly. They nevertheless share familiar features: private ownership, investment seeking a return, wage labour, markets, prices, profit and loss. These mechanisms can do certain jobs extremely well. Prices communicate scarcity. Profit rewards successful ideas. Investment allows useful technologies to expand. Competition can force businesses to improve, cut prices or disappear.
The OECD notes that competitive and fair markets can bring consumers lower prices, more choice and higher-quality goods and services, while supporting efficiency, innovation and economic growth. But prices communicate only certain kinds of information. They do not automatically tell us whether a transaction is fair. They do not tell us whether somebody outside the transaction is being harmed. They do not tell us whether somebody desperately needs something but cannot afford it. And they cannot represent the interests of someone who has not yet been born. A market can answer:
What will someone pay?
Ethics asks:
What should we do?
Sometimes those answers line up neatly. Sometimes they do not.
When injury was cheap
Industrial Britain shows the problem in its starkest form. Factories brought enormous increases in production alongside dangerous machinery, child labour, exhausting hours and severe industrial injury. There were humane employers. Some factory owners supported reform. That matters, because the history is not simply a morality play in which every capitalist is cruel and every regulator enlightened. The deeper problem was structural.
Safer machinery and shorter hours could cost an employer money. An injured worker might lose wages or health. Their family might lose its income. Their community might end up supporting them. Those consequences were real, but they did not necessarily appear on the employer’s balance sheet. Early factory legislation repeatedly exposed that gap.
The Factory Act of 1833 created an inspectorate with enforcement powers. There were, however, only four inspectors responsible for roughly 4,000 mills. Parliament’s own history records that the inspectorate was far too small to enforce the legislation effectively and that the Act was widely evaded. Further legislation tightened restrictions and expanded enforcement. Then compensation law began to alter the economics more directly.
The 1897 Workmen’s Compensation legislation established employer liability in covered employment where a worker suffered personal injury by accident “arising out of and in the course of” employment, subject to the Act’s conditions. The significance was that qualifying compensation no longer depended on the worker first proving employer negligence in the ordinary way. The injury had not changed. The moral value of the worker had not changed.
Who paid had changed.
As more of the cost of an accident was pushed back towards the organisation creating the risk, the economic case for prevention became stronger. This is the central mechanism running through the history. A business does not have to want anybody hurt. It merely has to face a situation where preventing harm costs more than the harm costs the business itself. That difference between the private calculation and the wider human calculation is crucial.
Modern health and safety law can therefore be understood partly as an attempt to close the gap between the private cost of an unsafe decision and its real social cost.
Safety improved when the calculation changed
The history is not simply one of failure. Modern health and safety has achieved an enormous amount. HSE says the rate of fatal injury to workers has shown a substantial downward trend over the long term, although the rate of improvement has slowed in recent years. Its latest provisional figures record 126 workers killed in work-related accidents in Great Britain in 2025/26. That progress is real. And regulation does not merely accompany safer workplaces.
A natural field experiment compared 409 randomly inspected workplaces in California with 409 matched establishments that could have been inspected but were not. The inspected workplaces subsequently experienced a 9.4% decline in injury rates and a 26% reduction in injury costs, with no detectable evidence of job losses from the inspections. The inspections changed behaviour. People were injured less. Businesses adapted. Safer machinery, protective technologies and increasingly sophisticated safety-management systems became normal features of advanced economies.
But this does not mean markets naturally maximise safety. A business normally has an incentive to find an economically efficient level of safety. That level may be extremely high if an accident would close a plant, destroy valuable equipment, drive away skilled workers, increase insurance costs, lose customers or create enormous liability. In those circumstances, safety and profitability pull in the same direction. The harder cases begin when they do not.
When harm became harder to see
Industrial disease changed the nature of the problem. A severed hand leaves little room for disagreement over whether an injury occurred. Asbestos, silica, lead and industrial chemicals are different. Disease may emerge decades after exposure. Workers may have changed employers. Several causes may be possible. Exposure may be difficult to reconstruct. Scientists may genuinely disagree. That creates room for delay. But delay is not automatically dishonesty.
Some scientific uncertainty is real. Some regulation moves slowly because institutions are fragmented, cautious or under-resourced. Some failures are incompetence rather than conspiracy. The more interesting question is what happens when uncertainty itself acquires economic value.
Britain’s history of silicosis shows how compensation could depend on narrow definitions of qualifying disease and employment. Historical research found that restrictive compensation rules were significantly relaxed only after campaigning by labour organisations and persistent evidence of disease among workers who were not adequately covered by the existing system. Asbestos provides an even richer record.
Britain introduced an Asbestosis Scheme in 1931, but historical research into Turner & Newall, then a leading British asbestos producer, found an inequitable compensation system and significant limitations in who could qualify. Here the story becomes one of information as much as injury. The worker, employer, scientist, regulator and public do not necessarily learn about a danger at the same time. That gives us the first of three clocks.
Clock One: Knowledge
First credible warning → accepted danger
How long does society take to accept that something is harmful? The question becomes especially important where the company knows more than the worker or the public. A 2023 analysis of previously secret PFAS industry documents found evidence that manufacturers knew by 1970 that PFAS presented serious toxicity concerns. The researchers concluded that this knowledge preceded comparable public-health understanding by decades and documented suppression of unfavourable research and distortion of public discussion. This is not only a story about chemicals.
It is a story about who knew, who did not, and what happened in between.
When uncertainty became valuable
Tobacco showed how sophisticated that gap could become. A company does not always need to prove that a profitable product is harmless. Sometimes it is enough to prevent regulators and the public becoming sufficiently certain that it is harmful. A comparative study examining tobacco, coal, sugar, the pesticide atrazine and organisations involved in climate-science disputes identified 28 techniques used to manufacture doubt. Recurrent tactics included challenging research methods, misrepresenting evidence, deploying supportive authorities and attempting to influence regulation. The argument shifts. From:
There is no danger.
to:
The evidence is not conclusive.
That difference can be worth money. Five years of delay can mean five more years of sales. Five years before redesign. Five years before compensation. Five years before liability fully arrives. Permanent victory is not always required. Sometimes the commercially valuable answer is simply:
not yet.
And regulation does not necessarily end the contest. Research published in 2026 reviewed 308 sources spanning approximately 50 countries to examine tobacco-industry behaviour after policies had been adopted. It identified five recurring responses: pre-emptive adaptation before rules took effect, followed by disregard, token implementation, circumvention or attempts to influence implementation through other actors. Circumvention was the most frequently documented response across most policy areas. That gives us the second clock.
Clock Two: Decision
Accepted danger → decision to regulate
But even once a rule exists, people may still be waiting for protection.
When compliance became part of the problem
Modern capitalism sits inside elaborate safety systems. Products are tested. Buildings are certified. Companies conduct risk assessments. Auditors inspect procedures. Regulators supervise industries. Boards review safety data. This is genuine progress.
But once proof of compliance becomes necessary for commercial success, that proof acquires value of its own. Grenfell Tower demonstrates the danger. The Grenfell Tower Inquiry found systematic dishonesty among manufacturers of products used in the refurbishment. It described deliberate and sustained attempts to manipulate testing, misrepresent test data and mislead the market about combustible products. Certification bodies also failed to scrutinise the claims with the necessary rigour and independence. This was not simply an old-fashioned case of ignoring a safety rule.
Products moved through a world of tests, certificates, technical claims, designers, contractors, building control and regulators—and still ended up on people’s homes. The safety system itself had become part of the terrain on which commercial behaviour operated. Grenfell also reveals the third clock.
Clock Three: Protection
Decision to reform → people actually protected
The fire happened in June 2017. The Inquiry’s final Phase 2 report was published in September 2024. As of 10 September 2026, the government’s implementation timetable still placed primary legislation for the proposed Single Construction Regulator in 2027-29, when Parliamentary time allows, with the legislative framework due to be complete and the single regulator established in 2028-29. This does not mean nothing has happened since Grenfell. Substantial reform has occurred. It means something more important.
Knowing what failed and completing the repair are different events.
The repertoire expanded
There is no neat historical progression in which one form of dishonesty replaced another. Modern organisations can still deny obvious problems. Victorian employers could manipulate evidence. What changed was the size of the repertoire. A business or industry can deny a problem, minimise it, dispute causation, narrow definitions, demand more evidence, lobby against regulation, exploit loopholes, challenge enforcement, game testing or selectively present favourable statistics. Modern organisations are especially rich in numbers. That can improve accountability. It can also create new ways to shape appearances.
Researchers examining occupational safety reporting by Australia’s 50 largest listed companies coined the term “Safewash”. Their analysis of corporate annual reports found that publishing more injury measures did not necessarily produce more complete information. Reporting tended to emphasise frequent, lower-consequence injuries while making serious safety-system failures less visible to investors. Nothing has to be fabricated. The distortion can lie in what gets measured, what is omitted, which denominator is used and which figures receive prominence.
A workplace with impressive safety statistics is not necessarily a safe workplace. That may be one of the most important changes in modern safety dishonesty. The lie does not always have to be:
This didn’t happen.
It can be:
Look at everything else.
The same problem outside safety
Health and safety gives us the clearest historical spine. But once the underlying mechanism is visible, it appears elsewhere.
Labour
Employment is formally voluntary. An employer offers a wage. A worker accepts or rejects it. But the two sides may have very different bargaining power. The worker may need next month’s rent. The employer may be able to wait. The worker may have few realistic alternative employers. Labour markets can also become concentrated.
ILO research published in March 2026 found that greater labour-market concentration was associated with greater wage inequality, particularly in the upper half of the wage distribution. It also found that institutions such as trade unions, collective bargaining and minimum wages appear to mitigate that relationship. A market wage tells us what agreement was reached. It cannot, by itself, tell us whether that agreement was fair.
Consumers
Competitive capitalism depends heavily on consumer choice. Yet individual businesses can profit by shaping that choice. The UK’s Competition and Markets Authority has examined evidence around online choice architecture: defaults, misleading framing, scarcity claims, hidden information, drip pricing and friction around cancellation. Its evidence review found, for example, that drip pricing can lead consumers to underestimate total prices, spend more and compare alternatives less effectively. The contradiction is revealing. Capitalism works best when consumers make informed choices.
An individual seller may sometimes profit when those choices are less informed. Consumer protection therefore does not simply restrain markets. It helps create some of the conditions markets need to work properly.
Healthcare
Healthcare exposes another limit. Markets respond to effective demand: willingness and ability to pay. Human need is not distributed according to wealth. The latest WHO monitoring estimates that 2.1 billion people experienced financial hardship from out-of-pocket healthcare spending in 2022, including 1.6 billion people who were pushed into poverty or further impoverished. The distinction is straightforward.
Need is a moral category.
Purchasing power is an economic one.
They are not interchangeable.
The person missing from the transaction
Environmental damage makes the problem easiest to see. A factory makes a product. The company earns money. The customer gets something they want. Both benefit. But pollution damages somebody downstream. That person never agreed to the exchange. Neither the buyer nor seller necessarily sees that entire cost in the price.
Climate change extends the problem across borders and generations. The IMF describes climate change as a global externality: emissions from one country add to the atmospheric stock of greenhouse gases and affect other countries as well. Carbon pricing is one attempt to make part of that otherwise external cost visible in the economic calculation. But climate change exposes a deeper problem. Future generations cannot enter today’s market. They cannot refuse our transaction. They cannot demand compensation. They cannot choose another supplier.
They are absent from today’s price. They are not absent from its consequences. A market has no automatic mechanism through which someone born fifty years from now can bid against today’s profitable pollution. That judgement has to come from somewhere else.
When wealth can influence the rules
Capitalist economies do not exist independently of governments. Markets depend on property law, contract law, courts, currencies, company law and competition rules. So the real choice is rarely market or government. It is:
what rules govern the market, and who gets to influence them?
Lobbying is not inherently corrupt. Businesses have legitimate expertise and legitimate interests. So do unions, charities, professional organisations and campaigners. But access is not distributed equally. Money can purchase lawyers, consultants, research, advertising, lobbying staff and persistence. The OECD revised its Recommendation on Transparency and Integrity in Lobbying and Influence in May 2024 to reflect changes in how political influence works. Its 2024 integrity assessment warns that low transparency around lobbying increases the risk that policymaking can be unduly influenced.
WHO identifies a similar problem in public health. Its framework on the commercial determinants of health includes lobbying, political donations, misinformation, research influence, working conditions and production externalities among the mechanisms through which commercial activity can improve or damage health. WHO also explicitly recognises that the private sector can provide important health benefits. That creates a possible feedback loop:
economic power → political influence → rules → economic advantage → greater economic power
Democratic institutions can interrupt that cycle. But the question remains uncomfortable:
Can society reliably restrain concentrated economic power when concentrated economic power can also influence the restraints?
The case for capitalism
None of this proves that capitalism is inherently evil. Markets can produce enormous social benefit. Competition can drive down prices. Profit can reward innovation. Investment can scale useful technologies rapidly. Entrepreneurs can solve problems governments never identified. Companies can compete for workers by improving conditions. Businesses can profit from safer technology, cleaner energy and better medicines.
Open markets can expand access to goods, reduce prices and contribute to economic growth, although the gains are not necessarily shared equally. The history of workplace safety itself shows capitalism’s adaptability. Change the calculation and businesses respond. If accidents become expensive, prevention becomes valuable. If pollution has a price, cleaner production becomes more competitive. If deceptive selling becomes costly, honest competitors are less disadvantaged. If workers have stronger alternatives, employers must compete harder for them. This may be capitalism’s greatest strength:
It can harness self-interest to produce socially useful outcomes when private incentives and public goals are aligned.
The difficult political work lies in creating that alignment where it does not arise by itself.
Is capitalism unethical?
We can now answer the question more precisely. If unethical means capitalism requires people to behave immorally, then no. Owning a business, employing people, investing money or earning a profit is not inherently unethical. Capitalism can reward behaviour that produces enormous human benefit. But capitalism also contains no automatic principle saying:
human welfare must take precedence over profitable opportunity.
That is not what markets are built to decide. A price mechanism can ask:
What is scarce?
What will people buy?
Where can capital earn a return?
Can this be produced more cheaply?
It cannot, by itself, answer:
What level of danger is acceptable?
What makes a wage fair?
Should healthcare depend on wealth?
What do we owe someone harmed by our pollution?
What do we owe people not yet born?
Those are ethical and political questions. That leads to a conclusion more useful than either capitalism is ethical or capitalism is evil.
Capitalism is morally incomplete.
It is a powerful economic mechanism. But it requires institutions around it to decide where economic calculation must stop. Law. Democracy. Regulation. Unions. Professional ethics. Courts.
Public services. Consumer protection. Environmental rules. Human rights. These institutions are not proof that capitalism has completely failed. They are evidence that markets alone are not enough.
The Price of Waiting
Health and safety brings us back to where we began. Britain did not abolish capitalism to make workplaces safer. It changed the rules under which capitalism operated. Dangerous machinery was regulated. Employers acquired liability. Workers gained compensation rights. Inspectors gained powers. Products had to meet standards.
Businesses adapted. People became safer. But the underlying problem keeps returning in new forms. The machine becomes the chemical. The chemical becomes the scientific dispute. The scientific dispute becomes the regulatory argument. The regulation becomes the compliance exercise. The compliance exercise becomes the metric.
And through it all, three clocks keep running.
When did we know enough?
When did we decide to act?
When did people actually become safer?
Between those moments, people continue to carry the consequences. Capitalism can tell us what something costs to produce. It can tell us what somebody will pay. It can create extraordinary incentives to invent, compete and solve problems. What it cannot do by itself is determine who should bear the cost when profitable activity harms somebody else. For two centuries, much of social reform has involved answering precisely that question and forcing previously hidden costs back into the economic calculation. Health-and-safety law. Compensation.
Minimum wages. Consumer protection. Environmental regulation. Public healthcare. Competition law. Taxation. Each says, in a different way:
The market calculation is not the only calculation that matters.
So, is capitalism unethical? Not exactly.
Capitalism is not an ethical theory. It is an economic system.
The danger begins when we confuse those two things—when we assume that because something is profitable, voluntary or efficiently priced, it must therefore also be right. Sometimes it is. Sometimes it is not. And when the market cannot see the difference, somebody outside the transaction may end up paying it.
That is the price of waiting.