Economic analysis of climate change

• 1. Introduction

  ›

  Economic analysis of climate change asks a question that sounds simple but turns out to be extraordinarily difficult: what does warming actually cost us, and how much should we spend to stop it? A 2026 study published in Nature put a specific dollar figure on this. Carbon dioxide emissions from the United States caused ten point two trillion dollars in cumulative damages between 1990 and 2020. China's emissions caused eight point seven trillion. The European Union's caused six point four two trillion. The researchers added one more finding that changes how you hear those numbers: future damages from past emissions are at least an order of magnitude larger than the historical damages from those same emissions. In other words, the bill has barely started to arrive.

  The core challenge is that climate damages are not a single number. They are market losses that show up in GDP, non-market losses that resist any dollar translation, and long-term risks that depend entirely on how you value the future against the present. The economists, modellers, and policy analysts who study this problem have built a range of tools to grapple with it, from statistical models that track how hot years hit poor countries harder, to massive integrated frameworks that weigh emissions cuts against projected damages across decades. What they find, and why their conclusions sometimes diverge dramatically, is what this documentary sets out to explain.

• 2. The Tools of the Trade

  ›

  Panel data from ground stations and gridded satellite observations sits at the foundation of one major modelling approach. Econometric studies using this kind of data have found, for instance, that hot years are consistently linked to lower income growth in poor countries, and that low rainfall is linked to reduced incomes in Africa. Negative effects of higher temperatures show up in agricultural output, in labour productivity, and in outdoor industries like mining and forestry. These are not projections about the future; they are patterns extracted from historical records.

  A second approach, structural economic modelling, looks at how climate effects ripple through the whole economy by tracing its inputs and outputs. Computable General Equilibrium models are one branch of this family, built to investigate how policies affect economic growth, trade, employment, and government revenues. A third approach, process-based modelling, simulates the physical, chemical, and biological changes under warming and then attaches economic consequences to those physical outcomes.

  Integrated Assessment Models, often called IAMs, sit across all of these by combining natural, social, and economic sciences into a single framework. The one most frequently cited in policy debates is the Dynamic Integrated Climate-Economy model, known as DICE, developed by William Nordhaus. DICE uses discount rates, uncertainty estimates, and risk assumptions to generate cost and benefit estimates for different climate policies. The model's outputs shift significantly depending on which assumptions are fed in, a sensitivity that turns out to matter enormously for what any analysis recommends.

• 3. Measuring the Damage

  ›

  The 2022 IPCC report landed on a striking range for what high warming means in practice. With warming around four degrees Celsius and low adaptation, annual global GDP could be reduced by between ten and twenty-three percent by 2100. With low warming, more adaptation, and different models, the same assessment found GDP reductions of one to eight percent. That is a wide spread, and it does not account for climate tipping points, nor for impacts on social well-being or distributional effects.

  A separate 2025 study tested what happens when you add trade linkages into the statistical model, since economies are deeply interconnected and a blow to one ripples outward. Adding global weather patterns rather than just national ones raised damage estimates from eleven percent to forty percent of global output by 2100 under a very high emissions scenario. Another study examined the last hundred and twenty years of data and found that climate change has already reduced welfare by twenty-nine percent, and projects that number rising to forty-seven percent with further warming. The temperature increase between 1960 and 2019 alone cut current GDP per capita by eighteen percent.

  Extreme weather events are also driving rising losses. Global disaster losses from weather, climate, and water events have increased sevenfold from the 1970s to the 2010s. Between 2015 and 2021, direct losses from disasters averaged above three hundred and thirty billion dollars per year. One study estimated attributable losses at an average of one hundred and forty-three billion dollars annually between 2000 and 2019, including ninety billion in statistical life value losses and fifty-three billion in economic damages. One 2020 study put the extra economic losses from current commitment levels, compared to limiting warming to one point five or two degrees Celsius, at between one hundred and twenty-seven and six hundred and sixteen trillion dollars through to 2100.

  Inflation is a less-discussed but measurable channel of damage. Climate change contributes between zero point three two and one point one eight percentage points of additional inflation per year, depending on the emissions pathway. A country running two percent inflation without climate change could see that figure rise to between two point three and three point two percent. Food inflation from climate impacts is estimated at between zero point nine two and three point two three percentage points per year.

• 4. Who Pays and By How Much

  ›

  A 2021 Swiss Re study put concrete numbers on the geographic gap in exposure. Wealthy countries like the United States would likely see their economies shrink by around seven percent under a severe warming scenario. Some developing nations face far worse outcomes, with projections of losses around twenty percent of economic output, and in some cases forty percent. A United States government report from November 2018 raised the possibility of US GDP falling ten percent due to warming, driven by shifts in geography, demographics, and technology.

  A 2019 modelling study found that climate change had already contributed to widening global economic inequality. Countries in colder, wealthier regions had felt little overall economic impact, and some may have seen marginal gains. Countries in hotter, poorer regions very likely grew less than they would have without warming. The underlying dynamic is not subtle: greenhouse gas emissions come mainly from high-income countries, while the impacts fall disproportionately on low-income countries that lack the resources to recover. Climate change can push more people into extreme poverty, particularly through sectors like agriculture and fisheries that are tightly coupled to weather conditions.

  Health costs cut across both wealthy and developing nations. The International Labour Organization published a report in 2019 projecting that even if warming is held to one point five degrees, climate change will cause productivity losses equivalent to eighty million full-time jobs per year by 2030, or around two thousand four hundred billion dollars annually. Agriculture is projected to account for sixty percent of that loss, with construction accounting for another nineteen percent. An estimated three point five million people die prematurely each year from air pollution tied to fossil fuel combustion. The health benefits of phasing out fossil fuels, measured using each country's economic value of life, are estimated to substantially exceed the cost of reaching the two-degree goal of the Paris Agreement.

• 5. The Cost of Acting

  ›

  A study published in 2024 estimated that keeping global warming below two degrees Celsius may cost about one percent of world GDP per year, but could prevent much larger losses of ten to twenty percent of GDP by mid-century. Economists estimate the incremental cost of climate change mitigation at less than one percent of GDP. The 2018 Paris Agreement target carries a potential upside: one 2018 study found that global economic gains from countries complying with the two-degree target could reach seventeen trillion dollars per year up to 2100, compared to a very high emissions scenario.

  Macroeconomic cost estimates for 2030 mitigation spanned a range. For a pathway consistent with atmospheric stabilization of greenhouse gases between four hundred and forty-five and seven hundred and ten parts per million of CO2-equivalent, estimates ran from a three percent decrease in global GDP to a small increase, relative to baseline. By 2050, stabilization costs in the same range spanned from a one percent GDP gain to a five point five percent decrease. These estimates assumed transparent markets, no transaction costs, and perfect implementation of cost-effective policy across all regions, conditions that rarely hold in practice.

  Adaptation costs are a separate line item. Across all developing countries, adaptation costs have been estimated at roughly two hundred and fifteen billion dollars per year up to 2030, with higher costs expected after that. Early and well-planned action on mitigation reduces overall costs; delayed action locks in higher expenses and forecloses future flexibility. An emissions stabilization target of four hundred and fifty parts per million, for instance, may become unreachable if near-term abatement is insufficient.

• 6. The Discount Rate Debate

  ›

  Nordhaus and Stern represent the two poles of one of the most consequential arguments in climate economics. Nordhaus, whose DICE model is the most widely used integrated assessment model, takes a descriptive approach to discounting: the rate should reflect actual market returns on capital, which he puts at around an average of four percent. Under this framework, a high discount rate makes future damages look smaller relative to present costs, reducing the urgency of immediate investment.

  Stern, whose 2006 review for the British government predicted world GDP would be reduced by several percent due to climate-related costs, applies a prescriptive approach. He argues that a zero pure rate of time preference is the ethical choice, since future generations have no voice in today's decisions and should be treated as moral equals to the current generation. Under the near-zero discount rate of zero point one percent used in the Stern Review, the social cost of carbon rises sharply, and the argument for aggressive near-term action becomes much stronger.

  The DICE-2013R Model generated estimates under seven distinct scenarios, ranging from a baseline case where climate policies have not changed since 2010, to an optimal case with fully implemented policies, to scenarios with near-zero and high discount rates. The spread of results across these scenarios illustrates how much the choice of rate shapes the answer. Markandya and colleagues estimated that in developed countries a discount rate of around four to six percent was probably justified, while in developing countries a rate of ten to twelve percent was cited, and private project discount rates could run between ten and twenty-five percent. There is still no consensus. The IPCC found no agreement on the appropriate long-run discount rate, and the prescriptive approach tends to produce long-term rates of two to three percent in real terms, while the descriptive approach produces rates of at least four percent after tax.

• 7. Sharing the Burden

  ›

  Direct global fossil fuel subsidies reached three hundred and nineteen billion dollars in 2017. When indirect costs such as air pollution are included, that figure rises to five point two trillion dollars. That number frames a set of arguments about who should bear the costs of transition, and those arguments draw on principles that go back centuries.

  The egalitarian position holds that each person has an equal right to the atmospheric commons. The polluter-pays principle draws on an Aristotelian idea that people should pay in proportion to the damages they cause. Historical responsibility assigns the greatest obligations to the countries that built up most of the stock of greenhouse gases now in the atmosphere, roughly two-thirds of which came from industrialized nations. A convergence approach allocates emissions allowances on a per-capita basis, which tends to benefit developing countries; by contrast, allocation schemes based on current emissions levels tend to produce welfare losses for those same countries.

  A 2019 modelling study added empirical weight to the historical responsibility argument by finding that wealthy countries in colder climates had often benefited or felt limited impact from warming, while poorer, hotter countries grew more slowly than they would have otherwise. The suggestion that developed countries could make side payments to developing countries addresses the efficiency objection to historical responsibility arguments, by shifting the ethical question rather than resolving it through the market. The Swiss Re assessment estimated that annual world economic output would be reduced by twenty-three trillion dollars per year unless greenhouse gas emissions are adequately mitigated, a figure that gives concrete scale to the stakes of how that burden is ultimately divided.