What this covers
Matt Ridley and Russ Roberts conduct a wide-ranging conversation about climate science, policy, and the sociology of scientific debate. Ridley, a science writer, presents the case for what he calls a lukewarmer position: humans warm the climate through CO2 emissions, but the warming is modest and poses less danger than mainstream climate models suggest. The discussion moves between technical climate questions—what climate sensitivity really is, how feedback mechanisms work, and what observational data actually show—and broader arguments about how scientific consensus forms, how incentives shape research, and whether expensive mitigation policies serve the poor or harm them.
Ridley challenges several pillars of alarm-driven climate policy. He argues that climate sensitivity is likely 1 to 2 degrees rather than the 3 degrees IPCC models assume, that the water-vapor feedback amplifying warming lacks strong empirical support, and that the fat tail of catastrophic scenarios—long cited by economists like Martin Weitzman—has shrunk as data improves. He disputes the famous 97% consensus figure, shows how confirmation bias can make intelligent people read identical evidence opposite ways, and contends that rising CO2 has boosted plant growth and brought measurable environmental benefits. He argues further that the real cost falls on today's poor through energy rationing and subsidy schemes, while the beneficiaries are speculative future generations who will likely be far wealthier. Roberts and Ridley also examine how the IPCC process, by concentrating authority rather than dispersing it across rival groups, may weaken the adversarial self-correction that normally keeps science honest.
Ridley argues for a 'lukewarmer' position: humans do warm the climate via CO2, but climate sensitivity is far lower than IPCC models assume, catastrophic outcomes are improbable, modest warming and higher CO2 are often beneficial, and costly mitigation policies harm today's poor more than they protect future generations.
- Climate sensitivity is likely 1-2 degrees, not 3, so observed slow warming is explained without catastrophe
- The water-vapor amplification feedback that triples warming in models is poorly supported; clouds appear to dampen warming
- Expensive mitigation (renewables subsidies, denying fossil fuels) imposes real, regressive costs on poor people for a thin-tailed risk
Climate science is corrupted by institutional consensus and confirmation bias, preventing rival challenge and treating skepticism as heresy rather than science.
- Climate science developed a 'psychology of taboo' where legitimate scientific challenge (the kind that earns gold stars in a PhD seminar) was reframed as heresy, with dissenters labeled 'deniers' (a term deliberately echoing Holocaust denial) and accused of being funded by fossil fuels, frightening journalists out of applying to climate the skepticism they apply to economic or foreign policy.
“it became, 'You've joined the dark side; you're a heretic; you're funded by fossil fuels. You're a denier'--which is an extraordinary phrase that began to be used for anybody who didn't sign up to alarmism and was deliberately intended to echo the Holocaust denial nonsense.”
- Science stays honest because it is normally a geographically dispersed enterprise where rival groups in different universities challenge each other; the IPCC process is dangerous because it makes a single 'church' out of climate science, removing the rival-group checking that keeps a field self-correcting.
“science is often a very geographically enterprise with people in different universities all over the country and all over the world challenging each other... The fact that there are rival groups prepared to say Professor Jones has got this wrong is what has kept science honest all these years.”
- Science is not a majority-rule process; its very purpose, as Feynman said, is to prove experts wrong, illustrated by the reversal of the longstanding consensus that dietary cholesterol causes heart disease, which was based on dodgy 1950s science enforced by ruthless consensus-building.
“as Richard Feynman famously said, 'Science is in the business of proving that experts are wrong.' And you know, until very recently, 97% of medics agreed that cholesterol was the cause of heart disease. Now, that's gone.”
- Confirmation bias is unavoidable: researchers naturally seek evidence confirming their own theories rather than acting as their own worst enemy, so the realistic safeguard is not self-skepticism but ensuring that a rival (Professor Smith) is incentivized to be the theorist's (Professor Jones's) worst enemy.
“if I come up with a theory, then I'm not going to go out and look for ways in which this theory is wrong. I'm going to go out and look for ways in which this theory is right... what I do think we should expect is that Professor Smith should be Professor Jones's worst enemy. That's what's kept science honest over the years.”
- Because of confirmation bias, well-informed honest people can find ample evidence to be alarmed or not alarmed about warming; the alarm-versus-lukewarm divide functions less like a scientific question and more like a values dispute (akin to social justice versus wealth creation, or left versus right), where each side reads the same evidence to confirm its prior.
“if you look hard enough, you can find lots of evidence to be alarmed by global warming, while being intellectually honest. And if you look hard enough you can find lots of evidence not to be alarmed while being intellectually honest.”
Climate sensitivity is much lower than consensus claims, with observational data supporting 1-2 degrees rather than 3.
- It is widely agreed, including by the IPCC, that doubling the level of CO2 in the atmosphere produces only about 1 degree of warming directly; the higher 3-degree sensitivity figure comes entirely from a modeled amplifying feedback, mainly water vapor, and the curved relationship means each doubling (300 to 600, 800 to 1600) yields the same ~1 degree.
“it's widely agreed by the IPCC and everybody else, that if you double carbon dioxide, you only get 1 degree of warming: 1.2, 1.1, that sort of zone of warming.”
- Fourteen peer-reviewed papers based on observational data, available to the IPCC for its 2013-2014 reports, all pointed to much lower climate sensitivity than previously assumed; the IPCC narrowed its range and declined to give a best estimate (previously 3 degrees), yet impact models worldwide still assume 3 degrees, while most scientists working on it now accept sensitivity is in the 1-2 degree range.
“they had before them 14 different peer-reviewed papers of this climate sensitivity issue, all of which pointed to a much lower climate sensitivity than they had been assuming in the past... pretty well every scientist who is working on this is now accepting that climate sensitivity is in the 1-2 degree range.”
- The fat tail in the climate sensitivity distribution (the small-but-significant chance of very large warming) that underlies Weitzman's precautionary argument has gotten much thinner with updated data, partly because a distorted graph in the 2007 IPCC AR4 report (identified by Nick Lewis) exaggerated the tail; correcting it and incorporating the latest observational papers yields much thinner tails, so a claimed 10% chance of 6 degrees becomes much less than 1%.
“When you correct that, the number gets smaller... they turn out to have much thinner tails than was portrayed in the 2007... So the 10% chance of 6 degrees of warming in 100 years becomes much less than 1% if you look at these charts now.”
- The model assumption that 1 degree of CO2 warming triggers a trebling of warming through water vapor and cloud feedbacks is poorly supported by evidence; observed water vapor changes in the relevant atmospheric layers and cloud cover changes do not justify the positive feedback, and most evidence suggests clouds actually dampen warming rather than amplify it.
“Some clouds amplify warming; some clouds do the opposite--they would actually dampen warming. And most of the evidence would seem to suggest, to date, that clouds are actually having a dampening effect on warming.”
Observed warming has been substantially slower than models predicted, with measurement problems inflating surface temperature trends.
- The observed rate of warming over the past 20-50 years (a little more than 0.1 degree per decade) has been much slower than IPCC climate models predicted; the IPCC attributes only about a third of a degree over 50 years to human activity, which is extremely hard to measure accurately, and surface temperatures show faster warming than satellites, suggesting urban heat island contamination.
“we have seen much less warming over the last 20 years, 30 years, 40 years, 50 years, than those models have predicted. We've seen a little more than 0.1 degree per decade.”
- Despite both theory and data supporting lukewarming rather than alarming warming, alarmism continues because there is now a huge vested interest in the alarming-future narrative.
“Why is everyone going on about the future being alarming? Because there's a huge vested interest in that now.”
- A 'lukewarmer' accepts that CO2 levels are rising, that CO2 is a greenhouse gas, and that warming has occurred and will continue, but challenges the idea that there is a strong likelihood of dangerous climate change, and indeed thinks benefits are likely.
“What I mean by a lukewarmer is somebody who is not challenging the idea that carbon dioxide levels are increasing or that carbon dioxide is a greenhouse gas or that we have seen warming in recent years or are likely to see warming to continue. But is challenging the idea that there is a strong likelihood that this will turn dangerous at some point in the future.”
Past apocalyptic predictions have repeatedly failed, justifying skepticism toward climate catastrophism; renewable subsidies regress poor people; adaptation and innovation offer better paths forward.
- Across the last three decades, repeated apocalyptic predictions (population explosion and inevitable famine, oil running out, nuclear winter, advancing deserts, rainforest destruction, SARS/bird flu/swine flu, collapsing sperm counts, acid rain, ozone layer) all failed to materialize, while lifespan, child mortality, health, wealth, happiness, IQ, and violence all improved; this track record justifies approaching new apocalyptic predictions like dangerous climate change with great skepticism.
“I've always, always, always come to the conclusion that the most alarming predictions are the ones that get the most coverage and the ones that are most raw... None of those things happened. In fact, there were extraordinary improvements in lifespan, child mortality, health, wealth, happiness, IQ, violence.”
- Renewable energy subsidies in the UK are funded by adding charges to consumers' energy bills (a stealth tax) rather than transparent taxation; because energy bills constitute a larger share of poor people's budgets, this is a regressive transfer from poor people to rich landowners and developers.
“energy bills are a more important part of the budget for poor people than they are for rich people, so this is a regressive tax, going from poor people to rich people.”
- Rather than rolling out inefficient, expensive renewable technologies, society should emphasize adaptation (as the latest IPCC report recommends) and intensive research into energy, including nuclear options like fusion, thorium, and molten salt reactors; currently about 100 times more money goes to subsidizing renewable energy production than to renewable R&D, which is the wrong allocation.
“at the moment we are putting about 100 times as much money into subsidizing the production of energy from renewables as we are into R&D in renewables. And nuclear. Let's get fusion. Let's get thorium. Let's get molten salt reactors.”
- Human living standards are likely to keep improving this century because the mechanism of innovation, the exchanging of ideas, has become far easier in the last 10-15 years through the Internet; even the IPCC's own charts project GDP per capita 3 to 16 times higher by 2100, meaning future generations facing 1-2 degrees of warming will be wealthy.
“the mechanism by which we innovation, i.e., by exchanging ideas, got a heck of a lot easier in the last 10 or 15 years. The Internet. So I think it's very likely we will produce innovations that improve people's living standards during the current century.”
- The vast majority of people with strong views on climate have only scratched the surface of the science and have mistakenly equated the fact that climate is changing with the conclusion that it is dangerous; the 'therefore' linking observations like earlier-arriving birds to a dangerous future is a gigantic, unjustified leap, though this is an honest illusion rather than dishonesty.
“they've failed to appreciate that there's a difference between climate changing and climate changing dangerously.”
- The famous '97% of scientists' figure refers to agreement that humans influence climate and that CO2 is a greenhouse gas (a view nearly all skeptics share), not agreement that climate change is dangerous; no study has shown 97% think man-made climate change is dangerous, and a larger American Meteorological Society survey found only 52% considered it likely dangerous.
“whenever you hear that 97% number, it's not referring to a consensus about dangerous climate change. It's referring to a consensus about humans' ability to affect the climate.”
- The UK's Bowland shale holds about 1400 trillion cubic feet of gas, larger than Pennsylvania's Marcellus and among the biggest reserves found worldwide, and US extraction techniques should work there, but extensive environmental requirements (12-month aquifer baselines, tremor and contamination rules, local profit distribution) risk making it uneconomic and slow to develop.
“the Bowland shale, which seems to have about 1400 trillion cubic feet of gas, which is huge. Bigger than the Marcellus in Pennsylvania.”
- Invasive species are decimating native species on islands worldwide and constitute a more serious environmental challenge than climate change, but the problem has been neglected because attention is consumed by climate change.
“Top on my list is invasive species, which are helping to decimate native species on particular islands all over the world, still... And I think those issues have been neglected. Because we are spending all our time talking about climate change.”
Rising CO2 has net environmental benefits, including massive vegetation increase and reduced drought, while fossil fuels prevented forest destruction.
- Rising atmospheric CO2 from fossil fuel emissions is in many ways improving the environment, evidenced by a roughly 14% increase in green vegetation over 30 years (largely attributable to extra CO2 enabling plants to grow more vigorously, especially in arid areas), plus decreasing drought and more rainfall benefiting crops.
“we've seen a 14% increase in the amount of green vegetation in the planet in the last 30 years. Almost certainly largely because of the extra CO2, which is enabling plants in all ecosystems to grow more vigorously.”
- Switching to fossil fuels (coal, oil, gas) instead of burning wood enabled humanity to stop destroying forests on a massive scale, and forests are now recovering large areas of land rapidly, particularly in the Western world.
“if we burn coal oil and gas, we don't burn wood. And so we don't chop down forests. There is no doubt that switching to fossil fuels enabled us to stop destroying forests on a massive scale.”
Precautionary fat-tail logic is religiously motivated and inappropriate once catastrophic tail risks become thin enough relative to other small-probability harms.
- Weitzman's precautionary fat-tail approach is a new version of Pascal's Wager and contains an element of religiousness; once the catastrophic tail is thin enough, it deserves no more attention than other small-probability catastrophes like asteroid strikes or volcanic explosions, and one must at some point judge certain possibilities too small to spend a fortune on.
“I would argue that it's a new version of Pascal's Wager... it's got to the point where it's so thin that, in my view, it deserves no more attention than the small possibility that we will be hit by an asteroid.”
- Weitzman's precautionary approach is unattractive because mitigation is not costless: damaging poor people today (through energy policy, biofuels, denying affordable electricity) in order to protect potentially wealthier great-grandchildren is not altruistic but selfish; it is like taking chemotherapy for a cold.
“to damage poor people today because you are worried about the fate of your great grandchildren doesn't seem to me to be altruistic. Quite the reverse: it feels selfish to me... I just think we are taking chemotherapy for a cold.”
The hockey stick reconstruction was methodologically flawed, relying on tree rings unrelated to temperature and statistical techniques that exaggerated recent warming.
- The hockey-stick temperature reconstruction was misleading on two counts: it relied heavily on bristlecone pine tree rings whose width was known to reflect overgrazing-induced bark-wound growth rather than temperature, and a 'short-centering' statistical method exaggerated any 20th-century-uptick series over a hundredfold; removing the bristlecone pines and one Gaspe Peninsula series makes the hockey stick disappear, as the scientists themselves found in a file labeled 'Censored'.
“the statistical method was--and this was beautifully demonstrated by Ross McKitrick and Steve Macintyre... was fishing out any data with a hockey stick shape and giving it undue emphasis. So, what happens if you leave the bristlecone pines out, and one other paper from the Gaspe Peninsula in Canada? And the answer is if you do that the hockey stick disappears altogether.”
- Attempts to rescue the hockey stick using large trees from Siberia turned out to depend on a single incredibly influential larch tree, whose unusual 20th-century growth could have been caused by local factors (such as surrounding trees falling and exposing it to sunlight) unrelated to global temperature.
“There were further attempts to sort of rescue the hockey stick which involved large trees from Siberia. And again, when you drilled down, it turned out that there was one incredibly influential larch tree in the sample. One tree.”