An interactive primer · Hannah Ritchie · 2024

Not the End of the World

The world is awful. both true
The world is much better. at once
The world can be much better. and that's the point
Scroll to begin · ~25 minute read · interactive throughout
00 The framework

No generation has ever been sustainable

Hannah Ritchie is a data scientist at Our World in Data and the University of Oxford. She grew up in Scotland convinced she was living through the end times — and almost abandoned environmental science in despair. Then she encountered the work of Hans Rosling, and the data itself changed her mind.

Her argument begins with a definition most of us misuse. The UN defines sustainability as having two halves: meeting the needs of the present generation, without compromising the ability of future generations to meet theirs. Hold both halves in your head and something startling falls out: humanity has never once achieved sustainability.

Our ancestors had a tiny environmental footprint — but at the cost of failing the first half catastrophically. For most of history, roughly half of all children died before adulthood. Life expectancy hovered around 30. Famine was routine. The "harmonious past" environmentalists romanticise was sustainable only because nature kept killing people. Today we've flipped the failure: we are finally meeting present needs for billions — while borrowing heavily against the future.

Drag the timeline below and watch the two halves trade places. The opportunity — unique to our generation — is to be the first to hold both at once.

The two halves of sustainability, 1800 → today
Blue: share of children dying before age five (present needs). Red: global CO₂ emissions (future needs). Drag the year.
Year
1800
Children dying before 5
43%
CO₂ emissions
0.03 Gt/yr
Stylised from Our World in Data — child mortality (Gapminder/UN IGME) & Global Carbon Project. Figures approximate.

Why doom is not just wrong — it's useless

Ritchie's second move is psychological. Doom feels serious, but it functions exactly like denial: both tell you there's no point acting. If collapse is certain, why decarbonise? Surveys she cites found more than half of young people worldwide believed "humanity is doomed". Yet apocalyptic predictions have a long, unbroken record of being wrong — not because the problems weren't real, but because humans responded.

1798 — The prophecy
Malthus: population will outrun food, forever

Thomas Malthus argued food grows linearly while population grows exponentially, making mass famine a mathematical certainty.

1800s–1900s — The response
Fertiliser, crop science, the Haber–Bosch process

Synthetic nitrogen alone now feeds roughly half the people alive. Food per person rose as population octupled.

1968 — The prophecy
Ehrlich's Population Bomb: "hundreds of millions will starve in the 1970s"

Paul Ehrlich declared the battle to feed humanity already lost, whatever we did.

1960s–1990s — The response
Borlaug's Green Revolution

High-yield wheat and rice tripled cereal yields. Famine deaths per capita collapsed to historic lows even as population doubled again.

Today — The prophecy
"It's too late. We're doomed."

The same logic, the same certainty — and the same blind spot: predictions of doom assume we won't act. We always have.

The alternative she proposes is urgent optimism — not the complacent belief that things will work out by themselves, but the evidence-based conviction that things can work out if we push. Optimism, in this framing, is what makes action rational. The rest of the book applies that lens to seven problems, and each chapter asks the same four questions: What's the real story in the data? What's working? What should we do more of? And what popular "solutions" are mostly noise?

0
of children died before age 5 in 1800. Today: 3.7% — still too high, 10× better.
0×
halves of sustainability. We've only ever achieved one at a time.
0
problems ahead: air, climate, forests, food, wildlife, plastics, fish.
01 Air pollution

The sky was dirtier when your grandparents were young

Air pollution kills millions every year — and it is one of humanity's clearest, least-told success stories. The cities of the rich world were once vastly filthier than Delhi or Beijing today, and the path they took down is now being repeated, faster, everywhere.

In December 1952, a cold fog settled over London. Coal smoke had nowhere to go. Visibility fell so low people couldn't see their own feet; cinemas closed because audiences couldn't see the screen. The Great Smog killed an estimated 10,000–12,000 people in a matter of weeks. It was the catastrophe that finally forced action: the Clean Air Acts followed, and London's air — which had been deteriorating for centuries — began its long descent towards breathable.

London's air over three centuries
Suspended pollution (SO₂, indicative µg/m³) — rising for 200 years, then a dramatic clean-up
Stylised after Fouquet (2011) via Our World in Data. London's air is now cleaner than at any time since at least the 1600s.

This arc — pollution rises as a country industrialises, peaks, then falls as it gets richer and regulates — is the environmental Kuznets curve, and air pollution follows it almost everywhere. The crucial update from the data: countries no longer have to wait until they're rich. Because clean technology already exists, China's air pollution peaked at a fraction of the income level (and the filth) at which London's did, and has fallen dramatically since 2013. Late developers can cut the corner.

The deadliest air pollution is also the least photographed: indoor smoke from burning wood, dung and charcoal to cook, which disproportionately kills women and children in poor countries. The fix is "dirty" by activist standards — gas stoves, electrification — and it is saving lives at scale as energy access spreads. Death rates from air pollution have roughly halved globally since 1990.

Proof we can do this · acid rain

In the 1980s, acid rain was the apocalypse headline: forests dying, lakes sterilised. Scientists identified sulphur dioxide, governments built cap-and-trade schemes and treaties, scrubbers went on smokestacks — and European SO₂ emissions fell by more than 80%. A global environmental crisis, identified, coordinated on, and largely solved. Almost nobody talks about it, which is precisely Ritchie's point: solved problems vanish from the news.

What actually works
  • Stop burning things — coal first, then oil; combustion is the common root of most air pollution and climate change, so one fix pays twice.
  • Regulation with teeth — clean air acts, vehicle standards, sulphur scrubbers. Every major clean-up was policy-led.
  • Clean cooking access — LPG and electric stoves for the ~2 billion still cooking on solid fuels.
  • Electrify transport — EVs remove tailpipe pollution where people actually live.
Mostly noise
  • Assuming it's hopeless for poor countries — they're cleaning up decades faster than the rich world did.
  • Treating outdoor smog as the whole problem — indoor air pollution has long been a comparable killer.
  • Nostalgia for pre-industrial air — homes full of open-fire smoke were lethal.
~0M
deaths per year from air pollution — still one of the world's biggest killers.
0
fall in global death rates from air pollution since 1990.
0
drop in Beijing-region PM2.5-driving SO₂ emissions in roughly a decade after China's 2013 "war on pollution".
02 Climate change

Peak per-person emissions is already behind us

Climate change is the biggest problem in the book — and the one where the data most sharply contradicts both the deniers and the doomers. The world is not on track for 1.5°C. It is also no longer on track for the 4–5°C catastrophe scenarios that defined the despair of a decade ago.

Start with a fact almost nobody knows: the average person's carbon footprint has already peaked. Global CO₂ emissions per capita topped out around 2012 at roughly 4.9 tonnes and have been drifting down since. Total emissions are flattening as that decline begins to outpace population growth — meaning the world is at, or very near, peak emissions. That doesn't solve warming (the planet responds to cumulative carbon, so emissions must fall towards zero, not merely plateau), but it marks the turning of an enormous ship.

The engine of that turn is decoupling. For two centuries, more prosperity meant more carbon. That link has now been broken — not in projections, in the historical record. Dozens of rich countries have grown their economies while cutting emissions, and crucially the cut survives even when you count consumption-based emissions (the carbon embedded in imports), so it isn't just offshoring to China.

Decoupling: the United Kingdom since 1990
Indexed, 1990 = 100. GDP per capita vs consumption-based CO₂ emissions per capita
Stylised from Global Carbon Project / World Bank via Our World in Data. The UK's economy grew ~50% per person while consumption-based emissions fell by more than a third.

The second engine is price. Ritchie calls the collapse in clean-energy costs the most underrated story of our era: within roughly a decade, solar power went from the most expensive electricity on the market to the cheapest electricity in human history, with wind and batteries close behind. This changes everything about the politics: poor countries no longer face a choice between development and decarbonisation, because the clean option is increasingly simply the cheap option.

The price of solar fell off a cliff
Solar photovoltaic module price, US$ per watt (log scale) — every doubling of capacity has cut prices ~20%
Stylised from Lafond et al. / IRENA via Our World in Data. From ~$106/W in 1976 to under $0.30/W — a >99% decline.

Then there's the chapter's most confronting chart. People fear nuclear power and accept fossil fuels, yet measured in deaths per unit of energy — accidents plus air pollution — the ranking is almost exactly upside-down. Coal is hundreds of times deadlier than nuclear or solar. Even hydropower, including its worst dam disasters, is far safer than any fossil fuel. Ritchie's rule: the safest energy sources are also the cleanest, so safety and climate point the same way.

Deaths per terawatt-hour of electricity
Accidents + air pollution. One TWh ≈ the annual electricity of ~150,000 Europeans. Tap a bar for detail.
Markandya & Wilkinson (2007); Sovacool et al. (2016) via Our World in Data. Nuclear figure includes Chernobyl and Fukushima.
Myth check · "it's too late / we're locked into catastrophe"

A decade ago, "business as usual" pointed at 4°C+ of warming. Because of policy and plummeting clean-tech costs, current-policy trajectories now cluster around 2.5–3°C — and stated pledges nearer 2°C. That is still genuinely bad: every tenth of a degree matters, and 1.5°C will very likely be passed. But the gap between "bad and worth fighting over" and "civilisational endgame" is the whole argument. There is no cliff-edge after which action is pointless; it gets harder, never hopeless.

What actually works
  • Stop burning fossil fuels — the root cause; everything else is detail.
  • Electrify everything — cars, heating (heat pumps), industry — and make the electricity clean (solar, wind, hydro, nuclear).
  • Make low-carbon the cheap option — carbon pricing, clean-tech subsidies, innovation. Price collapses drove the progress so far.
  • Eat less beef and lamb — food, especially ruminant meat, is the other big lever (next chapters).
  • Fly and drive less, drive electric — among personal choices, transport genuinely registers.
Mostly noise
  • Obsessing over plastic straws and bags — climate impact near zero.
  • Unplugging chargers — a phone's annual charge is worth pennies of carbon.
  • Skipping the dishwasher — full modern dishwashers usually beat hand-washing.
  • Rejecting nuclear on safety grounds — it's among the safest sources we have.
  • "Personal purity" guilt — systems and prices move gigatonnes; perfection theatre doesn't.
~0
the year per-capita CO₂ emissions peaked globally.
0
fall in solar module prices since 1976.
0×
more deaths per unit of energy from coal than from nuclear power.
03 Deforestation

The world passed peak deforestation decades ago

Half of the world's forests are gone — most of that loss centuries old. But global deforestation peaked in the 1980s and has roughly halved since, and a growing list of countries has flipped from losing forest to regrowing it.

Rich countries tell a forgotten story. England, France and the United States tore down their forests long before satellites existed — for farmland, fuel and ships. Then, one by one, they hit a forest transition: agriculture intensified, populations urbanised, and trees came back. France has roughly doubled its forest cover since 1830. The tragedy and the hope sit together: today's tropical deforestation is countries running the same script — and the opportunity is to help them reach the turning point with far less forest lost.

Global deforestation by decade
Million hectares of forest lost per decade — the peak is behind us
Stylised from FAO / Williams (2003) via Our World in Data. 2010s loss ran at roughly a third of the 1980s rate.

What drives the loss that remains? Overwhelmingly, food — and overwhelmingly, beef. Pasture and feed expansion (especially in the Amazon, where cattle ranching dominates) dwarf every other cause; most soy is grown to feed animals, not to make tofu. Palm oil, the internet's favourite villain, is genuinely implicated in Southeast Asian losses — but the data delivers a twist that has become the book's signature style of argument.

The palm oil paradox: yield per hectare
Tonnes of vegetable oil produced per hectare of land — palm is in a league of its own
Stylised from FAO via Our World in Data. Replacing palm with alternative oils would need several times more land.
Myth check · "boycott palm oil"

Palm produces around 36% of the world's vegetable oil on under 10% of the land devoted to oil crops. Boycott it, and demand shifts to soybean, rapeseed or coconut — which need four to ten times more land per tonne of oil, pushing deforestation elsewhere. The data-driven ask isn't "no palm oil"; it's deforestation-free, certified palm oil and pressure on supply chains. Counter-intuitive, measurable, and more effective than the boycott that feels righteous.

What actually works
  • Eat less beef — the single biggest driver of tropical deforestation.
  • Raise crop yields — intensification spares land; the Green Revolution saved forests the size of continents from the plough.
  • Deforestation-free supply chains — like the Brazilian soy moratorium, which decoupled soy from Amazon clearing.
  • Pay countries to keep forests standing — support forest transitions before the loss, not after.
Mostly noise
  • Blanket palm-oil boycotts — risk increasing total land use.
  • Paper-bag guilt — paper and timber mostly come from managed or replanted forests; food is the real driver.
  • Tree-planting as a licence to keep emitting — regrowth helps, but old-growth loss can't be offset tree-for-tree.
~0
fall in global deforestation rates since the 1980s peak.
0
of vegetable oil from palm — grown on under 10% of oil-crop land.
France's forest cover compared with 1830 — forest transitions are real.
04 Food

How we feed ourselves is the master lever

Food touches every other chapter. It is the leading driver of deforestation, biodiversity loss and freshwater use, and a quarter of greenhouse gas emissions. Get food right and you ease all of them at once. The headline data point: half of the world's habitable land is farmed — and most of it goes to animals.

Where the world's land actually goes
Follow the cascade: habitable land → farmland → livestock vs crops → the calorie payoff
Stylised from Poore & Nemecek (2018) / FAO via Our World in Data. Livestock uses ~77% of farmland but supplies ~18% of calories and ~37% of protein.

Read that cascade slowly, because it contains the most important asymmetry in the book. Livestock occupies roughly 77% of all agricultural land, yet provides only about 18% of the world's calories. Animals are a spectacularly land-hungry way to make food. This single fact reframes nearly every food debate — and it explains why what you eat swamps almost everything else about your diet's footprint.

Ritchie is emphatic and counter-intuitive here: for the climate, what you eat matters far more than where it comes from. "Food miles" are typically a tiny slice of a food's emissions — most comes from land-use change and on-farm processes, not transport. So a plant-based food flown across the world usually still beats local beef. The hierarchy of impact is stark.

Greenhouse gas emissions per 100g of protein
Kg CO₂-equivalent. Beef and lamb are in a category of their own; transport is a sliver of the total
Poore & Nemecek (2018) via Our World in Data. Beef ≈ ~25× the emissions of legumes per gram of protein.

The good news layered underneath: we are getting radically more efficient. Thanks to rising yields, the world produces far more food on barely more land than decades ago — global agricultural land use may have already peaked. That "land sparing" is the quiet hero that lets forests and wildlife recover. The levers that help most are unglamorous: cut beef and lamb, cut food waste (roughly a quarter of food is lost or wasted), and keep raising yields so we farm less land, not more.

What actually works
  • Shift away from beef & lamb — by far the highest-impact dietary change; even swapping to chicken or pork helps a lot.
  • Cut food waste — about a quarter of food is wasted; halving it is one of the largest single levers.
  • Raise yields / land sparing — more food per hectare frees land for nature.
  • Eat more plants — legumes, grains and nuts are the lowest-impact protein.
Mostly noise
  • Buying local to cut carbon — food miles are usually a tiny fraction of emissions.
  • Avoiding all packaging — packaging is a sliver; it often prevents far costlier spoilage.
  • Organic-as-always-greener — lower yields can mean more land per unit of food.
  • Almond-milk panic — nearly every plant milk beats dairy on land, water and carbon.
⚖️ Play: Big lever or mostly noise?
Ritchie's recurring trick is separating actions that move gigatonnes from ones that just feel virtuous. Decide for each card.
🥩
Loading…
Score: 0 / 0
05 Biodiversity

The hardest chapter to feel hopeful about

Ritchie is honest that wildlife is where the data is most sobering. Extinction rates are far above natural background levels, and the headline statistics are real. But she insists on precision, because the scariest numbers are routinely misread — and misreading them leads to the wrong fixes.

Take the famous figure that wildlife populations have fallen ~69% since 1970 (the WWF Living Planet Index). It's often reported as "we've lost 69% of all animals," which is false. It's an average of population trends — and averages can be dominated by a handful of collapsing populations while most are stable. A small number of catastrophic declines drags the whole index down. The losses are serious; the specific claim people repeat is not what the number says. Getting this right matters, because despair built on a misread statistic is fragile.

The real drivers, in order, are clear from the data — and they put humans' biggest impact somewhere surprising: not poaching, not plastic, but land.

What actually threatens species
Share of threatened species affected by each driver (species can face several)
Stylised from IUCN Red List analyses via Our World in Data. Habitat loss from agriculture dominates — which loops straight back to food.

Because habitat loss leads, the most powerful thing we can do for wildlife is the same lever as the food chapter: use less land — eat less beef, raise yields, halt deforestation, and let nature reclaim what's spared. Ritchie also pushes back on a romantic instinct: invasive-species control, well-run protected areas and even some "unnatural" interventions are how conservation success actually happens. And successes exist — humpback whales, mountain gorillas, and many bird species have recovered from the brink through deliberate action. Recovery is possible; it just isn't automatic.

Proof we can do this · the ozone layer

In the 1980s a hole in the ozone layer threatened life-shielding protection from UV radiation. The world identified CFCs, signed the Montreal Protocol (1987) — the only UN treaty ever ratified by every country on Earth — phased the chemicals out, and the ozone layer is now healing, on track to recover this century. A genuinely global environmental catastrophe, coordinated on and fixed. It is the template for what the harder problems could become.

06 Ocean plastics

The right problem, often the wrong target

Plastic in the ocean is a real harm — to turtles, seabirds and ecosystems. But it is also the chapter where rich-world environmental guilt is most spectacularly misdirected, and where the data points to fixes very different from the ones we reach for.

First, the geography. The overwhelming majority of plastic entering the ocean comes from a relatively small number of rivers in regions without developed waste-management systems, concentrated in Asia and Africa. The plastic in your kitchen bin in Melbourne or Manchester, collected and managed, is very unlikely to end up in the sea. The problem is mismanaged waste, not consumption per se — which means the solution is waste infrastructure in fast-developing countries, not primarily guilt in rich ones.

Why "recycle harder at home" misses the target
Indicative share of ocean plastic, by waste management of the source region
Stylised from Meijer et al. (2021) / Our World in Data. Ocean plastic is overwhelmingly a waste-infrastructure problem.

Second, the trade-offs nobody mentions. Plastic is light and cheap, so its climate footprint per use is often far lower than the "natural" alternatives we feel good about. A cotton tote bag must be reused hundreds of times to beat a plastic bag on total environmental impact; a paper bag can carry several times the carbon of the plastic one it replaced. Plastic packaging also dramatically cuts food waste — and wasted food, as the food chapter showed, carries a heavy footprint of its own. Plastic is a genuine ocean problem and an over-blamed climate scapegoat at the same time.

Myth check · "ban plastic bags, save the planet"

For ocean litter, reducing single-use plastic in places with poor waste collection helps. For climate, swapping plastic for cotton or paper usually makes things worse unless you reuse the replacement many times over. The two goals can point in opposite directions — which is exactly why "plastic = bad, alternatives = good" is too crude. The high-leverage move is building waste-management capacity where the leakage actually happens, not performing zero-waste rituals where it doesn't.

07 Overfishing

Where good management already turned the tide

Overfishing is serious: a third of assessed fish stocks are exploited beyond sustainable limits, and global wild catch has plateaued because the ocean can't give more. Yet this is also a problem with a proven cure — and places that applied it have watched their fisheries come back.

The pattern is consistent. Where fisheries are well-managed — science-based catch limits, enforcement, rights-based systems that give fishers a stake in long-term stocks — populations are stable or recovering. The collapse stories come from places where management is weak or absent. The United States, much of Europe, Australia, New Zealand and others have rebuilt depleted stocks through exactly this kind of governance. Management, not abstinence, is the lever.

Fisheries: managed vs unmanaged
Schematic of stock health where strong management is applied vs where it isn't
Stylised from Hilborn et al. (2020) via Our World in Data. Intensively managed fisheries are, on average, at or above sustainable levels.

There's a second twist that links back to food. Aquaculture — fish farming — now provides about half the seafood we eat, and farmed fish like carp, tilapia and even some salmon can have a markedly lower environmental footprint than land-based meat, especially beef. Sustainable seafood, wild or farmed, can be part of a low-impact diet. The villain isn't fishing; it's unmanaged fishing.

What actually works
  • Science-based catch limits + enforcement — the core of every recovery story.
  • Rights-based management — giving fishers long-term stakes aligns incentives with conservation.
  • Low-impact aquaculture — well-run fish farming eases pressure on wild stocks and beats most land meat.
  • Cut bottom-trawling & bycatch — the most destructive practices, addressable by gear and rules.
Mostly noise
  • "Never eat fish" — well-managed and farmed seafood can be among the lower-impact proteins.
  • Assuming all fisheries are collapsing — managed ones are recovering; it's a governance gap, not a doom.
  • Ignoring aquaculture — it's already half of seafood and a big part of the solution.
What actually matters

The big levers, and the noise

Across all seven problems, the same shape recurs: a few choices move the planet, and a crowd of well-meant gestures barely register. Ritchie's gift is the willingness to say which is which, using numbers — even when the answer is unpopular. Here is the whole book's verdict on one screen.

Personal actions, ranked by impact
Approximate effect of high-impact vs low-impact choices. Notice the order — it's not what most campaigns emphasise.
Stylised synthesis from Wynes & Nicholas (2017), Poore & Nemecek (2018) via Our World in Data. Bars are indicative, not additive.

But the deeper claim of the book is that individual virtue was never the main event. Personal choices matter and set norms, yet the gigatonne-scale change comes from systems: the price of solar, national grids, fishing law, supply-chain rules, agricultural yields. The most important thing most readers can do, Ritchie argues, isn't a perfect lifestyle — it's pushing those systems, through how we vote, work, invest, build and influence others. Effective beats pure.

Hold both halves of sustainability

Meet present needs and protect the future. We're the first generation that could do both — no past era ever did.

Doom and denial both end in inaction

Urgent optimism — "things can be better if we act" — is the only stance that motivates change. Apocalyptic predictions have a perfect record of being wrong because humans respond.

Many trend lines already bent

Peak per-capita CO₂, peak deforestation, peak farmland, collapsing solar prices, recovering ozone and acid rain, rebuilt fisheries. Progress is real and underreported.

Food is the master lever

Less beef and lamb, less waste, higher yields. It eases climate, deforestation, biodiversity and water all at once.

Distinguish the lever from the gesture

What you eat ≫ food miles. Waste infrastructure ≫ home recycling guilt. Energy systems ≫ unplugging chargers. Follow the magnitudes.

Systems over purity

Vote, build, invest and pressure supply chains. The biggest wins are structural, not personal — and that's empowering, not deflating.

? Test your grok

Did it stick?

Six questions on the counter-intuitive core of the book. Answer each to reveal why the data says what it says.

0/6

The first generation

This is not the end of the world

Ritchie's title is a double meaning. It rejects apocalypse — and it insists that now is not the end of our agency. We have, for the first time, the technology, wealth and knowledge to meet human needs without wrecking the planet. The data shows the turn has already begun; the question is only how fast we finish it.

Her closing reframe is the whole point of the book. Every previous generation faced a brutal trade-off between human welfare and the environment, and most chose — or were forced into — terrible outcomes on one side or the other. We are the first generation rich enough, clever enough and aware enough to refuse the trade-off. That makes this a uniquely hopeful moment, not despite the problems but because, for once, we can actually solve them.

The thesis, in one line

We could be the first generation in human history to achieve real sustainability — to leave the world genuinely better than we found it. Not guaranteed. Not automatic. But, for the first time ever, possible. That possibility is the reason to act, and the reason it is not the end of the world.

"Being a good ancestor means leaving the world in a better state than we found it. For the first time, that's within reach."