Powered by 
Climate policy debates often blur the line between what climate science establishes and what public discourse assumes it says. Drawing on the underlying scientific record, including model uncertainty, emissions scenarios, and the Intergovernmental Panel on Climate Change’s own economic assessments, this essay traces that gap and its costs.
The world has grown wealthier and safer even as it has warmed: extreme-weather mortality has fallen sharply, and the IPCC’s own central economic projections put the toll of continued warming this century at a few percent of global GDP.
Taking climate change seriously means engaging honestly with what the science actually shows, warts and all, rather than defaulting to the most alarming reading available. That distinction matters for how the United States and other free societies weigh emissions policy against the growing demand for reliable, affordable energy, and for what they owe a generation raised on forecasts of catastrophe unsupported by the science.
The question
This essay does not dispute that the climate is changing, that human activity is contributing to those changes, or that the changes matter. Those points are not in question.
The real question is how to arrive at a rational climate and energy policy, one that transparently weighs the risks and benefits of a changing climate against society’s growing need for reliable and affordable energy. That question matters because there is a significant gap between what climate science actually says and what the public has been told it says, and that gap has driven damaging policy decisions in many countries. The science supports concern, but not the panic that has shaped much of today’s climate and energy policy.
The context
The globe has warmed by about 1.3 degrees Celsius since 1900. That is real, measured, and unambiguous, and human activity has contributed to the warming.
But the same one hundred and twenty-five years also saw the global population quintuple and the economy grow roughly twenty-five fold. Life expectancy rose by fifty years. Meanwhile, deaths from extreme weather, including hurricanes, floods, droughts, and extreme heat, fell by around 95 percent.
That decline ran alongside the warming, driven in part by the prosperity and resilience that energy has enabled. It reflects what human ingenuity and economic development do: they make societies more resilient, more adaptable, and better able to cope with whatever nature delivers.
The IPCC, the UN body that synthesizes the world’s climate science, does not dispute any of this. Its most recent assessment is in fact considerably more hedged than public discourse implies: there are no clear global trends in hurricane frequency or intensity, and drought signals are mixed and highly regional. Economic losses from weather disasters, expressed as a share of global GDP, have if anything declined slightly over the past three decades, averaging about 0.2 percent a year. A wealthier world is a more resilient one. The headlines often overstate what the science supports.
In the United States, a similar pattern holds. CDC mortality data show that cold-related deaths are roughly double heat-related deaths in recent years. The mortality data cut against the common assumption that a warming country is more dangerous.
There is also a moral dimension to this picture that rarely enters the conversation. The 1.5 billion people living in the developed world enjoy abundant, affordable energy. The remaining 6.5 billion do not have enough. Americans consume roughly thirty times more energy per capita than Nigerians, and some three billion people use less electricity in a year than the average American refrigerator. Energy poverty also means cooking indoors with wood and dung, a practice whose smoke kills an estimated two million people annually.
Global energy demand is projected to grow by half again by midcentury as the developing world industrializes, and fossil fuels remain, as they long have been, the most reliable and convenient way to meet that demand.
When the developed world tells the rest of the planet that the science compels rapid decarbonization, the response from leaders in the developing world has been pointed: the Indian prime minister has objected that the path to development is being closed off to nations still climbing it, and Niger’s former president has said plainly that Africa is being punished by Western investment decisions and will exploit the fossil fuels it has.
The analysis
What, then, does the science say about the future? Answering that requires good models of how the climate will respond to growing human influence, and defensible assumptions about future emissions. Both are weaker foundations than commonly assumed.
Start with the models. The current generation, those that feed into IPCC assessments and from there into policy, have sensitivities to carbon dioxide (the warming response expected from doubling atmospheric CO2) that range from about 1.8 to 5.6 degrees, a factor of three. That gap reflects how much remains unknown about the climate system. The scientific community has worked to narrow that range by testing models against observations, and has found that many of the most sensitive models, the ones projecting the greatest warming, produce temperatures that do not match the observed record. They run too hot, and as a result have been de-weighted quietly, without the corrected headlines that should have followed.
Emissions scenarios present a similar problem. The most alarming projections, those underpinning the language of existential threat, are built on a scenario known as RCP8.5, which assumed global coal use would quintuple by the end of this century, a trajectory unrelated to any credible energy forecast. Even the scenario’s original authors have distanced themselves from treating it as a baseline. Yet it remains the scenario most used in studies that make news and most often invoked by politicians and commentators as “business as usual.”
When the models are constrained to those that perform well against observations, and the scenarios reflect plausible energy trajectories, the resulting warming by 2100 falls between 2 and 3 degrees, with a median of 2.2, of which 1.3 degrees has already been realized since 1900, a period during which humanity has flourished as never before. That additional warming warrants serious attention. It is not, however, the apocalypse, and saying so plainly matters.
The economist William Nordhaus, whose work on the economics of climate change won the Nobel Prize, found that there is an optimal pace at which to reduce emissions, and that moving faster than that pace causes its own kind of damage by forcing immature technologies into deployment before they are ready. His 2018 Nobel lecture noted that an economically optimal decarbonization path could allow global temperatures in 2100 to rise more than 3 degrees, twice the Paris Agreement’s 1.5-degree guardrail. That figure rests on assumptions that can be and have been challenged, but Nordhaus’s central point stands: the task is to reduce emissions gracefully, not to panic.
A figure rarely seen outside the technical literature comes from the IPCC’s own Working Group II report, which deals with impacts. It estimates that at 2.5 to 3 degrees of warming, the central effect on global GDP is a reduction of 1 to 5 percent, spread across eighty years of economic growth in a world that will, on current trajectories, be dramatically richer than today. The IPCC’s Fifth Assessment Report put the point directly: for most economic sectors, the impact of climate change will be small relative to the impacts of other drivers, such as population, income, technology, and governance. That is a real cost, but one equivalent to losing only one or two years of growth over eight decades.
Public debate rarely engages the central question this raises: how do the costs of climate policies compare to the climate impacts they are meant to avoid? In other words, is the trade worth making?
The gap between public understanding and the underlying science can be traced to a process of compression. An IPCC assessment runs thousands of pages, full of ranges, caveats, and genuine uncertainty. The Summary for Policymakers condenses that into something shorter and more definitive, negotiated line by line by government delegates, with uncertainties minimized and worst cases emphasized. Media and political communication compress it further still, until the remaining caveats disappear and uncertainty reads as confidence. A photocopy loses detail at random; this process does not. At each stage, the findings that survive are those that heighten alarm.
The implications
The costs of that distortion are concrete, and they arrive just as the country faces a genuine surge in demand for electricity.
US electricity prices rose nearly 7 percent in 2025, more than double the headline inflation rate, and Goldman Sachs projects households will absorb an additional 6 percent increase through 2027, driven substantially by AI data centers, which now account for roughly 40 percent of the nation’s electricity demand growth. Virginia, home to the largest concentration of data centers in the country, has already seen residential rates climb more than 13 percent in a single year, particularly in Loudoun County, and in Ashburn most of all, where the data centers cluster
Meeting that new demand affordably requires exactly the kind of reliable, dispatchable generation that emissions-reduction mandates have discouraged for two decades. Layering decarbonization costs onto a grid already straining to keep pace with AI-driven load growth is not a strategy; it is a self-imposed handicap.
Much of that cost traces to a basic fact about wind and solar power that public debate tends to skip past: they are cheap to build but cannot be depended upon. Solar panels do not produce at night, and wind comes and goes by the hour. Every renewables-heavy grid therefore needs a backup system, whether natural gas, nuclear, or storage, capable of covering the stretches, sometimes lasting up to two weeks, when neither wind nor sun contributes much. Such stretches, which the Germans term dunkelflauten, have been documented everywhere renewables have been deployed at scale, including Texas, California, the UK, and Germany.
That backup system has to be nearly as capable as the wind and solar fleet it supports, which means the most expensive part of a renewables-heavy grid is the backup needed to keep it reliable. Studies that model the US grid at something close to today’s reliability standard find that natural gas, with or without carbon capture, remains the cheapest way to deliver it, and that grids relying mainly on wind, solar, and storage cost two to three times more.
Wind and solar energy are diffuse and so demand far more land and far more material than the energy sources they are meant to replace. Generating the same electricity from wind takes roughly four times the land that gas requires, and thirty times what nuclear requires, along with about ten times as much concrete and steel as a nuclear plant of equivalent output.
The high-value materials involved in wind and solar technologies (copper, molybdenum, dysprosium, and the like) are concentrated in a small number of countries, several of them geopolitically inconvenient. The Democratic Republic of Congo produces three-quarters of the world’s cobalt. China dominates rare-earth and graphite extraction, and although it consumes well under half the world’s solar panels, it manufactures three-quarters of them and nearly all of the polysilicon and wafers that go into them, aided by cheap coal power, loose environmental standards, and, by the US government’s own determination, forced labor in parts of its supply chain. None of this is likely to get easier: copper demand is expected to double within the next decade even as new mines, which take upward of fifteen years to bring online, are expected to fall a quarter short of that demand.
There have been other, indirect costs incurred by attempts at rapid decarbonization that treat reliable, affordable energy as a problem to be solved, when it is in fact the foundation everything else depends on. Hundreds of billions of dollars committed to an energy transition with negligible measurable effect on the global emissions curve could have funded flood defenses, heat-resilient infrastructure, and energy access in the developing world. The industrial costs are also real: emissions and livelihoods alike have been exported, and the result called progress.
The climate panic has also imposed costs on science itself. Researchers who raised legitimate questions about model reliability, scenario plausibility, and the gap between what the science established and what was claimed in its name were not answered on the merits. They were labeled instead. The word “denier” became a career weapon, deployed against colleagues of genuine integrity. That is how science fails.
The hardest cost to accept, though, is what has been done to young people. A generation has been told, by schools, governments, and the institutions they were raised to trust, that they face an existential threat, that the planet might be uninhabitable in their lifetimes, that the future itself was cause for dread. The data do not support that message, nor does the IPCC, read carefully. An entire generation has grown up anxious about a future the evidence does not support, and that may be the most serious harm of all, because it is the hardest to undo.
The path forward
Treating climate change as a serious, ongoing challenge is different from treating it as license to abandon clear thinking in favor of alarm.
A more honest path forward starts with sustaining and improving climate science itself. Current knowledge of the climate system falls short of what policy needs: paleoclimate work shows how and why climate has shifted in the past, better observational coverage shows what the system is doing now, and improved models, tested against both, would narrow the uncertainties that are too large to usefully inform policy.
Alongside that technical work, public communication needs to change. It is possible to acknowledge that human influence on climate is real and growing while retiring the sound bites that treat every 0.1 degree of warming as an accelerant toward catastrophe; most people are perceptive enough to discount messaging that outruns the evidence, and alarmism that does so corrodes the credibility of the science.
Policy also needs to accept, openly, that reliability and affordability take precedence over emissions targets. Dependable, affordable power is the precondition for sustaining any emissions reductions. Europe’s recent energy troubles are largely self-inflicted: fossil investment and domestic production were curtailed in favor of import dependence and weather-dependent generation, a choice whose consequences were foreseeable and arrived on schedule. Decarbonization itself should proceed thoughtfully rather than by mandate, with sustained research, development, and demonstration of the technologies (small modular reactors, grid-scale storage, advanced batteries, carbon capture) that could eventually lower the cost of cutting emissions.
Energy systems are complex and slow to change for fundamental reasons, and pushing them faster than they can accommodate produces disruption out of proportion to any climate benefit gained.
Any serious policy also has to reckon honestly with the developing world’s energy needs, rather than waving them away. Wind and solar alone, without costly backup, cannot deliver the energy access that developing economies require, and few advocates of rapid global decarbonization offer an alternative that respects the technical, economic, and political realities. And alongside mitigation, adaptation deserves far more attention than it gets: it is local, effective, proportional to the risk actually faced, and something societies already know how to do.
Environmental stewardship is a genuine obligation, but it is not in conflict with abundant, reliable, affordable energy. Rather, it requires energy.
The countries best positioned to manage whatever climate future arrives will not be those that impoverished themselves in advance. They will be the ones that remained rich, innovative, and adaptable.
Science requires doubt, policy requires honest trade-offs, and democracy requires accurate information. All three have been ill served in the climate and energy debate, with the consequences visible, measurable, and still mounting. But none of this is beyond correction. The serious, informed conversation this issue demands can be had, and in growing pockets of policy and research, it already gives a reminder that clear thinking is still possible, and always necessary.
Steven E. Koonin is the Edward Teller Senior Fellow at the Hoover Institution, where he contributes to the Tennenbaum Program for Fact-Based Policy and the George P. Shultz Energy Policy Working Group. He is the author of Unsettled: What Climate Science Tells Us, What It Doesn’t, and Why It Matters (BenBella Books, 2024). This essay is based upon keynote addresses delivered to conferences in Sydney and London.