Climate Change by the Numbers: What the Data Says, What It Doesn't, and Why Both Sides Are Uncomfortable

The global mean temperature in 2025 reached approximately 1.44°C above the 1850–1900 pre-industrial baseline, making it the third-warmest year in the instrumental record [1]. The Copernicus Climate Change Service put the figure at 1.47°C [2]. The UK Met Office projects 2026 will reach a central estimate of 1.46°C, marking a fourth consecutive year above 1.4°C [3].

The three-year average from 2023 to 2025 exceeded 1.5°C for the first time — the threshold the Paris Agreement designated as the boundary of acceptable warming [1]. That does not mean the long-term warming trend has permanently crossed 1.5°C; year-to-year variability from El Niño and La Niña cycles affects individual readings. But the margin is shrinking. At the current rate, the long-term average is expected to breach 1.5°C by the end of this decade [4].

How Fast Is It Warming?

The rate of warming has accelerated. From 1900 to 1980, global temperatures rose roughly 0.1°C per decade. Since the early 2000s, that rate has roughly doubled [5]. Climate models from the 1990s — including James Hansen's 1988 projections and the first IPCC assessment — have proven broadly accurate on global temperature trends, generally projecting warming within the range that has actually occurred [6][7]. A 2020 NASA-funded study confirmed that 14 of 17 climate model projections from 1970 onward accurately predicted subsequent warming when fed actual emissions data [6].

Where the models have been less reliable is in regional predictions, precipitation patterns, the pace of ice sheet loss, and the behavior of carbon cycle feedbacks like permafrost thaw and methane release from wetlands [7]. The range of uncertainty for 2100 projections remains wide: between 1.1°C and 5.4°C of additional warming above today's levels, depending on emissions trajectories and climate sensitivity — a parameter that even the latest models cannot pin down more precisely than a range of 2.5°C to 4°C [8].

This uncertainty cuts both ways. It means outcomes could be less severe than central estimates suggest. It also means they could be considerably worse.

Who Emits, and Who Has Emitted

The United States has produced roughly 400 billion tonnes of CO2 since 1751 — about 25% of all historical emissions. The EU-28 accounts for another 22%. China, despite being the largest current annual emitter, is responsible for approximately 13% of cumulative emissions. Until 1882, more than half of the world's cumulative CO2 came from the United Kingdom alone [9].

Current annual emissions tell a different story. China produces roughly 30% of global CO2 emissions. The United States accounts for about 14%, the EU for 7%, and India for 7% [9]. But per capita figures reframe the picture again: U.S. per capita emissions are approximately double China's and eight times India's [10].

Source: U.S. Energy Information Administration

Data as of Mar 20, 2026CSV

This disparity sits at the core of international climate negotiations. Developing nations argue — with historical data on their side — that wealthy countries built their economies on fossil fuels, accumulated atmospheric CO2 for over a century, and now ask poorer nations to decarbonize before they have achieved comparable living standards. The counterargument, advanced most pointedly by the economist Bjørn Lomborg, is that no country in history has industrialized without fossil fuels, and that denying cheap energy to the world's poorest 4 billion people imposes immediate, measurable suffering to prevent distant, uncertain harms [11].

The Cost of Action vs. Inaction

The economic calculus has grown more precise, though the estimates still span a wide range. A 2024 study from the Potsdam Institute for Climate Impact Research found that even if emissions were cut drastically starting today, the world economy is already committed to a 19% income reduction by 2050 due to locked-in climate damages — six times larger than the mitigation costs needed to limit warming to 2°C [12].

The Swiss Re Institute projected GDP impacts by 2050 ranging from 4% loss if Paris targets are met, to 11% with partial mitigation, to 18% under no action [13]. The 2022 IPCC report estimated annual GDP reductions of 10–23% by 2100 under high-warming scenarios (~4°C) with low adaptation [14].

These figures consistently exceed the estimated $3–5 trillion in annual global investment needed for a net-zero transition [15]. The Potsdam study's finding — that damages cost six times more than prevention — is perhaps the single most important data point in climate economics [12]. But it carries caveats: it relies on economic models that struggle to account for tipping points, assumes damages are smoothly distributed (they are not), and cannot fully capture the compounding effects of simultaneous regional disruptions.

Lomborg and others in the cost-benefit tradition argue that the damage estimates rely on discount rates and modeling assumptions that can dramatically change the result, and that money spent on aggressive near-term decarbonization yields lower returns than investment in adaptation, R&D, and addressing immediate causes of death in the developing world — indoor air pollution, malaria, malnutrition [11]. The Grantham Research Institute at LSE has challenged Lomborg's specific numbers, arguing he understates climate damages and overstates mitigation costs, but the underlying question — how to allocate scarce resources — remains legitimate [16].

The Climate Finance Gap

At COP29, wealthy nations agreed to increase climate finance for developing countries to $300 billion annually by 2035, with a broader target of mobilizing $1.3 trillion per year [17]. In 2022, developed countries provided $115.9 billion — exceeding the $100 billion annual goal for the first time, but two years late [18].

The gap between commitments and needs is vast. The IPCC estimates developing countries need $2–3 trillion per year for climate action. The UNFCCC's own analysis puts the figure at $6 trillion by 2030 just to meet half of existing national commitments [17]. The $300 billion pledge amounts to roughly 1.4% of developed countries' GDP — meaningful but nowhere close to what recipient countries say they need [18].

This raises the hardest equity question in climate policy. Developing nations where most population growth is occurring account for a small fraction of cumulative emissions. Asking them to spend 5–10% of GDP on rapid decarbonization — versus the 1–3% that wealthy nations spend — would require forgoing the fossil-fuel-based development path that lifted hundreds of millions out of poverty in China, India, and across East Asia over the past four decades. Whether renewable energy can now provide a comparable development pathway at comparable cost is a bet, not a certainty.

The Paris Agreement: Does It Matter?

The emissions gap between current nationally determined contributions (NDCs) and the 1.5°C pathway is large. The UNEP Emissions Gap Report 2025 found that meeting 1.5°C requires emissions reductions of 55% below 2019 levels by 2035. Current NDCs achieve less than 14% of the additional reductions needed [19]. If all current commitments were fully implemented — itself an optimistic assumption — warming would reach 2.3–2.5°C by 2100 [20].

The U.S. withdrawal from the Paris Agreement cancels roughly 0.1°C of projected improvement [19]. This is small in absolute terms but significant given how narrow the remaining window is. Whether the Agreement "matters" depends on what one expects from it. It has never been binding. No enforcement mechanism exists. But it has functioned as a coordination mechanism, a ratchet for ambition, and a framework for accountability. Without it, the alternative is not a better agreement — it is no agreement.

The Energy Transition: Costs, Tradeoffs, and Paradoxes

U.S. electricity generation data from the EIA reveals a structural shift underway.

Source: U.S. Energy Information Administration

Data as of Mar 20, 2026CSV

Coal generation has fallen from 1.2 million GWh in 2018 to 729,000 GWh in 2025. Natural gas remains dominant at 1.8 million GWh. Solar has quadrupled from 64,000 GWh in 2018 to 296,000 GWh in 2025. Wind has grown from 273,000 to 464,000 GWh. Nuclear has remained stable near 780,000 GWh — the largest single source of zero-carbon electricity in the United States [21].

Lazard's 2025 LCOE+ report confirms that unsubsidized utility-scale solar ($27–118/MWh depending on region) and onshore wind ($25–70/MWh) are the cheapest sources of new generation in most markets [22]. This is without tax subsidies. However, the LCOE of onshore and offshore wind increased 55% and 23% respectively compared to 2024 due to supply chain pressures [22].

Nuclear: The Uncomfortable Clean Energy Source

Nuclear power provides roughly 20% of U.S. electricity and remains the largest source of zero-carbon power. Yet the EIA's 2025 Annual Energy Outlook projects no new nuclear capacity in its reference case, reflecting overnight capital costs of $7,821 per kilowatt — roughly four to five times the cost of solar or wind per unit of capacity [23]. New nuclear projects in the West have a dismal track record: Vogtle Units 3 and 4 in Georgia came in at more than double the original budget and years behind schedule.

Germany's Energiewende offers an instructive case. After shutting down its remaining nuclear plants by April 2023 — a decision accelerated by post-Fukushima politics — Germany replaced much of that capacity with coal and imported Russian natural gas, increasing both emissions and energy dependence [24]. PwC estimated that if nuclear had remained operational, Germany's emission-free power generation could have reached 94% in 2024 [24]. By 2024, Germany's coal and gas consumption had declined from 2022 peaks, but electricity prices remained among the highest in Europe [25].

The lesson is not that nuclear is always the answer. It is that removing a proven zero-carbon source without adequate replacement is costly in both emissions and money.

China's Paradox

China simultaneously leads the world in renewable deployment and emissions growth. In 2025, China's combined solar and wind capacity exceeded its coal-fired thermal capacity for the first time [26]. It installed 360 GW of wind and solar in 2024 alone — more than half of global additions [27]. And yet China commissioned more than 50 large coal units in 2025, up from fewer than 20 per year over the previous decade [28].

Both facts are true. China's emissions from energy and industry dropped 0.3% in 2025 even as energy consumption rose 3.5% — meaning renewables are beginning to displace fossil generation at the margin [29]. But the coal buildout continues because China treats coal plants as backup capacity for intermittent renewables, essentially retrofitting them as peaker plants [28]. Whether this strategy results in net emissions reductions depends on utilization rates: if coal plants run at low capacity factors as backup, emissions fall; if energy demand surges and they ramp up, they don't.

Grid Reliability: The Intermittency Problem

The criticism that wind and solar cannot provide reliable baseload power is not a political talking point — it is a physics constraint. Wind and solar generate electricity only when conditions allow. Grid-scale battery storage is growing rapidly: 49.4 GW / 136.5 GWh came online in the first nine months of 2025, a 36% increase over the same period in 2024 [30]. Batteries can inject or absorb power in milliseconds, providing frequency regulation faster than any thermal plant [30].

But current battery technology provides 2–4 hours of storage at scale. Long-duration storage (12–100 hours) remains in early stages [30]. The question is whether storage technology will advance fast enough to match the pace of renewable deployment. NERC's 2025 summer reliability assessment flagged battery storage and solar as contributing over 35 GW of on-peak capacity, but stressed that extended periods of low wind and cloud cover during extreme weather — precisely when electricity demand peaks — remain a vulnerability [31].

Source: Swiss Re Institute / Potsdam Institute for Climate Impact Research

Data as of Mar 20, 2026CSV

Climate Migration: The Human Cost

The World Bank projects that 216 million people could be forced to migrate internally by 2050 under a high-emissions scenario, driven by water scarcity, crop failure, and sea-level rise. Aggressive mitigation and adaptation could reduce that figure to 44 million [32]. Sub-Saharan Africa faces the highest displacement risk, with up to 86 million people affected by water stress alone [32]. South Asia could see 40 million internal climate migrants, with a third concentrated in Bangladesh [33].

These projections should be compared to current displacement figures: UNHCR counted 117.3 million forcibly displaced people worldwide in 2023 [34]. Climate migration on the projected scale would roughly double that number — but it would unfold over decades rather than as a sudden crisis, making it simultaneously more manageable and easier to ignore.

The populations most affected — subsistence farmers in the Sahel, coastal communities in Bangladesh and the Pacific Islands, urban poor in regions facing extreme heat — have contributed the least to cumulative emissions.

The Mitigation Menu: What Works?

The cost-effectiveness of different mitigation strategies varies enormously. Utility-scale solar and wind, at current LCOE figures, avoid CO2 at roughly $20–50 per tonne when displacing coal, making them the highest-return mitigation investments available [22]. Energy efficiency improvements often pay for themselves. Reforestation costs $5–50 per tonne of CO2 sequestered but faces permanence and verification challenges [35].

Carbon capture and storage ranges from $15–25 per tonne for concentrated industrial streams to $40–120 for dilute streams like power plant flue gas [36]. Direct air capture remains at $280–580 per tonne — an order of magnitude more expensive than preventing the emission in the first place [36]. As a backstop technology it has value; as a primary strategy it is prohibitively expensive at current costs.

Adaptation investments — seawalls, drought-resistant agriculture, early warning systems, urban heat mitigation — are harder to compare directly with mitigation. But in the near term, they often deliver higher returns per dollar in developing countries, where climate impacts are most acute and the cheapest mitigation opportunities have already been captured by wealthier nations.

The Political Divide

One faction treats climate change as an existential emergency requiring immediate, economy-wide mobilization regardless of cost. James Hansen, the NASA scientist who brought climate change to Congressional attention in 1988, has argued that current policies amount to intergenerational injustice. Climate scientist Michael Mann has warned that the window for action is closing and that delay is the new denial [37].

The opposing faction, exemplified by Alex Epstein's "Fossil Future" thesis and Vaclav Smil's work on energy transitions, argues that fossil fuels remain indispensable to human flourishing, that energy transitions historically take 50–70 years regardless of policy, and that premature abandonment of fossil fuels would cause more suffering — through energy poverty, economic disruption, and food insecurity — than the climate change it aims to prevent [38][39].

Neither side fully reckons with the tradeoffs. Climate urgency advocates often understate the costs of rapid transition, the grid reliability challenges of high-renewable systems, and the regressive impact of higher energy prices on the poor. Fossil fuel defenders often minimize the tail risks of continued warming, the declining cost competitiveness of renewables, and the asymmetric vulnerability of the world's poorest populations to climate impacts they did nothing to cause.

The data supports several things simultaneously: climate change is real, accelerating, and caused primarily by fossil fuel combustion. The economic case for mitigation is stronger than the case for inaction. Renewable energy is now cost-competitive without subsidies in most markets. And yet: the transition will take decades, not years. Fossil fuels currently provide 80% of global energy. No country has decarbonized its economy at the speed climate scientists say is necessary. The 1.5°C target is almost certainly out of reach.

These facts are not contradictory. They are the actual situation. Policy that ignores any of them — the urgency of the problem, the difficulty of the solution, or the suffering caused by both action and inaction — will fail.