The Paris Accords as 'climate insurance:' Unaffordable and unnecessary

By Steven E. Koonin and Mark P. Mills
Real Clear Wire

[The following is based on remarks delivered by Koonin and Mills at an MIT Free Speech Alliance debate.]

The climate change debate continues to rage. Though the science remains “unsettled,” what does seem settled is that President Trump will withdraw, again, from the now infamous Paris Climate Accords. Importantly, those accords are centered on pledges made to modify national energy policies.

A decision to exit the Paris Accords is no mere gesture. The central fact for citizens everywhere is that putative “climate solutions” would deploy trillions of dollars and implement mandates and diktats for the supply and use of energy in every aspect of society.

The stated rationale for proposals to alter completely how civilization is fueled is the need for an “insurance policy” against future climate catastrophes. In that framing, the climate-fearful argue that some possibility of consequential future harms warrants the “responsible” decision to “buy” insurance now. But this often-argued “insurance” construct assumes that we know enough to say that the consequences of future climate change justify paying for the insurance—and collaterally, that we know the “insurance” itself will be affordable.

It turns out that we do know quite a bit about both those domains. As we outline below, reality tells us that the climate-change consequences that we’re trying to avoid will be modest—and that the costs of the “insurance” are staggering.

What are we insuring against?

The proposition of paying for “climate insurance” requires that we first consider the “benefits” of 50-year decarbonization, a timescale that comes from the Paris goal of limiting global average temperature rise to 2oC. We can then turn to weighing those benefits against the cost of achieving so-called “net zero” greenhouse gas emissions. That comparison is complicated, not least because of the uncertainties on the impacts allegedly avoided by reducing human influences on the climate. There’s also the issue of “costs and benefits to whom,” as well as the question of whether there is in fact urgency to reduce emissions. 

There are three points to make: the timescale for emissions reduction is arbitrary; the climate “threat” is far from dire; and the cost/benefit calculus very much depends on who is doing the calculation.

Start with the Paris goal itself, which seeks to keep the rise in average global surface temperature to less than 2oC, which the climate modelers say would require net zero global emissions in the latter half of this century. Meanwhile, emissions are continuing to rise and will again reach an all-time high this year. The subtitles of the UN’s annual Emission Gap Report give a flavor of the lack of progress: in 2023 it was a “Broken record . . . Temperatures hit new highs, yet world fails to cut emissions (again)”, and this year it was “No more hot air, please.” But even that 2oC is not a hard limit. When Hans Schellnhuber, the so-called “father of the two-degree limit,” was once asked why he gave that number, he responded that it was about right, and it was an easy number for politicians to remember. There is no credible case to make that all manner of chaos will suddenly break out if the temperature rises two, or even three, degrees.

Next is the question of whether the climate threat is so dire that it requires precipitous and Promethean actions—transforming the entire world’s energy system in a few decades. The answer to that question is not as uncertain as the doomsayers claim. There is some guidance from recent history, since the globe has warmed 1.3oC in the past 120 years and about the same amount of warming is expected over the next century. Rather than catastrophe, humanity has seen unprecedented prosperity over that period: the global average lifespan has gone from 32 years to 72 years, per capita GDP has increased sevenfold, the literacy rate has soared, and the death rate from extreme weather events has decreased by a factor of 50! So, it’s hard to believe that a comparable warming over the next century will significantly derail such progress. In fact, the consensus of economic impact studies, as published last year by the Biden White House, is that there would be a few-percent decrement in the GDP for a few degrees of warming. That’s “in the noise,” as we physicists say. Of course, there will be differential impacts, there are uncertainties, and GDP isn’t the only measure of wellbeing. Nevertheless, predictions of catastrophe are not credible.

If you listen to the popular media, you might believe that we humans have already broken the climate. Yet even the Intergovernmental Panel on Climate Change (IPCC) can’t find any climatically significant trends in most climate impact drivers, let alone attribute them to human influences. Losses from extreme weather events are in fact declining as a percentage of GDP as the world becomes more resilient. And projections of the magnitude of future warming have decreased as the IPCC refines its models and the world emits somewhat less CO2 than had been expected because of both slower growth and a shift to carbon-light energy sources.

Finally, there’s the question of “worth it to whom.” While the 1.5 billion of us in the developed world have adequate energy, most of the world craves far more. The inequalities are astounding. Nigerian per capita energy consumption is 30 times smaller than that in the U.S., and some 3 billion people use less electricity each year than is consumed by an average U.S. refrigerator. Fossil fuels are the most effective way of providing the reliable and affordable energy those folks need to improve their lot, so any restraints on those fuels immorally hinders their development. In short, decarbonization is an unaffordable luxury for most people. They face many more immediate, tangible, and soluble problems than the risk of some future climate impacts, the latter best summarized as “we don’t know what, we don’t know when, and we don’t know how severe.”

Urging, cajoling, and requiring the developing world to forswear fossil fuels, as the World Bank and other financiers have been doing, is directly contrary to human flourishing. It’s like telling a starving person, “Don’t eat that steak because it might raise your cholesterol.”

One common objection to this argument is that decarbonization has other benefits—for example, reducing local air pollution. But consider the case of China, where life expectancy increased by 10 years from 1980 to 2020, even as fossil-fuel use increased by 700 percent. (Some of that owes to the reduction in indoor pollution due to cleaner cooking fuels like LPG, a fossil fuel.) Even the dirty Chinese coal plants had great benefits, since increased energy availability was much more important to most Chinese than cleaner air.

Finally, in accounting for global costs and benefits, one also must include the benefits of rising CO2 levels—as hard as that may be for some to believe. One benefit: deaths from extreme temperatures have decreased in recent decades, since roughly 10 times as many people die from extreme cold events (which are declining) than from heat waves (which are increasing modestly). Another benefit is that the earth has significantly “greened”—by one measure, the earth is 40 percent greener than it was 40 years ago. This trend has also helped agricultural productivity to soar, since plants “eat” CO2.

The bottom line is that most scientists know, and an increasing number are finally willing to (bravely) acknowledge publicly, that there is no climate emergency or climate crisis. Hence there is no need for the precipitous and universal decarbonization called for by the Paris Accords. That kind of energy transition will be (in fact, already is) disruptive and expensive. In fact, most in the emerging world are saying, understandably, “We won’t do it unless you pay us for it.” And we in the developed world don’t have the money required to do that.

What does the insurance cost?

The willingness of citizens and politicians to “buy” climate insurance boils down to a technological evaluation of the range of proposed energy systems and, critically, those that can deliver at societal scale. It is thus not as much about forecasting, as is the case with climate science, but rather about evaluating the cost of building and operating hardware based on various technology scenarios.

We have a tendency these days to be captivated by aspirational technologies, unproven systems, and, in social media terms, “clickbait” with breathless headlines about putative “breakthroughs.” The reality is that industrial-scale systems of any kind that can be built in the immediate future use technologies we already know how to build, that were invented years ago, and that are now mature, with viable supply chains. And for calculations (not forecasts) of costs, there is plenty of robust and reliable data about the hardware and systems we know how to build.

There are good reasons to invest in R&D to identify superior energy technologies. But that has no relevance to estimating the costs of the insurance policy now being contemplated because, again, what can be deployed at scale in the next decade or so is what we already know how to build, whether wind turbines or gas turbines.

We have evidence that illuminates real-world decarbonization costs, and over a similar period contemplated by climate activists. The year 2000 is the same distance in our past as the target date of 2050 is in our future. Since 2000, the U.S. and Europe have spent well over $10 trillion to avoid, replace, or minimize the use of hydrocarbons. Those efforts did succeed in lowering the hydrocarbon share of global energy, but only by about three percentage points, to today’s level of just over 81%. In absolute terms, the use of oil, natural gas, and coal all increased, collectively, by an amount equivalent to adding six Saudi Arabias’ worth of oil production. Similarly, a decade of subsidies directed at electric vehicles led to some 40 million EVs on the world’s roads. No doubt they are displacing oil that would otherwise be used. But the absolute consumption of gasoline still rose and now stands at a record high.

If spending $10 trillion didn’t cause any significant decarbonization, what would it take?

Based on that recent experience, and even assuming the favored technologies are, overnight, say 50% cheaper—which is not happening—reducing hydrocarbons’ share of energy to just below half of all 2050 demand pencils out to somewhere between $100 trillion and $300 trillion. That’s about five to fifteen times the capital that would be needed to meet demand using conventional energy. And even then, that lowered share of hydrocarbons in 2050 would still be, in absolute terms, about the same quantity used today because of greater energy demand in the future.

All of this assumes that future solar, wind, and battery costs will be radically lower, a claim unsupported by reality. Their rising costs aren’t a feature of supply chain disruptions from the Covid lockdowns but instead are anchored in an unavoidable fact: far more metals and minerals are needed to build so-called “green” energy machines than to build hydrocarbon machines. A seminal analysis from the International Energy Agency (IEA) found that partial decarbonization would require fantastical increases in global mining—ranging from a fourfold to a 40-fold increase over today, depending on the mineral. Other research finds bigger gaps: a recent Yale paper determined that global mining would need to increase 60- to 300-fold, depending on the mineral.

That points to a core problem: creating a new mine takes an average of 15 years. Relevant for “insurance” planners, the global mining industry is not now planning to mine such quantities. Even if one assumed that money and mandates could shorten the timeline for building new mines to a decade, there is still no arithmetical way to meet soaring mineral demands for building the decarbonization machinery.

The decarbonistas correctly respond that market forces will resolve this. That’s true, but not in the way that they imagine. The effect of demand stunningly greater than supply will be staggering inflationary price escalation—i.e., demand destruction. That will affect all markets because the same minerals are used everywhere. But for energy machines, material inputs constitute from 30% to 50% of the cost of fabricating solar modules, and 50% to 70% of the cost of an EV battery. In short, the costs to produce green machines will rise, not fall. This disconnect in materials reality is completely ignored in forecasts. It’s a gap that cannot be resolved by hand-waving about recycling, which at best can only slightly moderate the growth in net demand.

We also have Germany as another source of macro-economic evidence regarding the costs of real-world decarbonizing. Over the past two decades, Germany roughly doubled its total electric grid capacity, primarily building solar and wind, but necessarily kept about 80% of the original grid. (Most of the shrinkage came from an ill-considered shutdown of nuclear plants.) Meanwhile, Germany’s total electric demand grew less than 10%. That disconnect had an economic impact: Germany’s electricity rates have nearly tripled. It not only increased energy poverty in Germany but also made the nation energy-fragile, a prelude to the disastrous consequences of the Ukraine war’s loss of discounted Russian natural gas. If the solution to that problem were to build more wind and solar, Germany would have done it. Instead, it reversed course and built massive LNG import capacities. But that U-turn was too little, too late as Germany is now deindustrializing catastrophically, largely because of high-cost energy. Meanwhile, here in the U.S., we’ve seen a doubling in the wholesale prices for utility-scale solar and wind projects over the past half-dozen years. The real-world costs of “too cheap to meter” solar and wind are soaring.

At societal scale, experience has belied claims that solar and wind, especially when combined with utility-scale battery storage, are inherently cheaper on a so-called lifecycle cost basis. If that were true, the decarbonistas at data center companies would be cutting the utility cord entirely and building such solutions to meet the now-obvious, epic demands for electricity from the digital economy. They’re not. And buying and refurbishing old nuclear plants is a limited, one-time option.

These cost increases are separate from the inflationary impacts if the U.S. government spends the money appropriated and subsidized by the Inflation Reduction Act (IRA), which is, again, undisguised energy spending. The actual total costs of the IRA, if ultimately fully implemented, have been estimated to be from $2 trillion to $3 trillion. For context, that’s comparable to the $4 trillion (inflation-adjusted) the U.S. spent to prosecute World War II.

The IRA’s inflationary spending doesn’t include other energy spending underway and planned in about two dozen states pledged to follow California’s aggressive decarbonization plans. Nor does it count the fact that EV-only mandates will induce electric utilities to spend an additional $3 trillion to expand grid delivery infrastructure. Also not included are the costs of additional power plants to make the electricity in the first place. Taxpayers should be worried, not least because such rapid spending creates epic opportunities for waste, fraud, and corruption.

The scale of spending is perhaps better understood through a more specific lens: an analysis done by National Bureau of Economic Research (NBER). The NBER team dove into the IRA’s interstices and found that the EV subsidies alone total some $23,000 to $32,000 for each vehicle. This is truly China-level subsidization.

If the decarbonistas were serious about cost-effectiveness, they’d be far more focused on subsidies to induce purchase of, say, more efficient combustion engines. By the IEA’s own estimate, such a policy would reduce global oil use more than would a nearly seven-fold increase in the number of EVs in the world.

What should we really do about the climate?

A dispassionate look at trends in demographics, economic development, and energy technology shows that achieving global net zero by the end of the century would be extraordinarily challenging, if not impossible. At the same time, a dispassionate look at the consequences of missing the arbitrary Paris goal does not reveal catastrophe. That doesn’t mean that the world, or we in the U.S., should do nothing. 

Here’s what we should do.

First, we must sustain and improve climate science, for we have great gaps in our knowledge. Paleoclimate studies tell us how and why climate has changed in the past; current observations with improved coverage, precision, and continuity tell us what the climate system is doing today, and models give a sense of what might happen in the future. But we urgently need greater statistical rigor in the analyses and more focused modeling efforts to reduce uncertainties.

Second, we must improve public communications, for there is far too much “fake news” about the climate. We need to end the rhetoric about a “climate crisis” even as we acknowledge that human influences on the climate are real and that we should be thinking about what to do in the long-term and in an orderly fashion. The public must have an accurate view of both climate and energy and get beyond slogans like “We are on a highway to climate hell with our foot still on the accelerator.” Non-experts are savvy enough to dismiss hyperbolic scare stories; those engaging in such sensationalism contribute to the general erosion of scientific credibility.

Third, we must acknowledge that energy reliability and affordability take precedence over emissions reductions. A good start is the admission that oil and gas will be necessary for the foreseeable future. Europe’s current energy crisis is self-inflected; fossil-fuel investments and domestic production were abandoned in favor of unreliable imports and unreliable generation from wind and solar. It was easy to see that this would lead to trouble, and many predicted it, but decarbonization was nonetheless given primacy over reliability and affordability.

Fourth, governments must embark on thoughtful and graceful energy transition programs that incorporate technology, economics, regulation, and behavior, and that estimate costs, timescales, and actual impacts on the climate. To reduce the so-called green premium, one essential element of thoughtfulness is the need for more research and development leading to demonstrations, instead of premature deployment, of newer energy technologies. Small-scale fission, low-cost grid storage and management, non-carbon chemical fuels, and carbon capture and storage are all part of a reasonable list of promising ideas, but all are at very early, non-commercial stages today.

Energy is delivered at societal scale by complex systems that touch on—to borrow from a movie title—“everything everywhere all at once.” Those systems are best changed slowly. Precipitous actions to reshape the entire energy system is far more disruptive than any plausible impact of climate change. It’s scandalous that the U.S. is planning to spend trillions of dollars on deploying unreliable energy technologies when we have so many other tangible and solvable needs, including healthcare, infrastructure, and education.

Fifth, developed countries must acknowledge the inevitability, if not desirability, of meeting the developing world’s energy needs. Most of the world today is energy-starved, and fossil fuels are the only viable way of meeting that demand; they provide over 80% of the world’s energy now, as they have for many decades. Without costly backup systems, weather-dependent wind and solar generation cannot provide appropriate energy access for those people. Advocates of rapid global decarbonization engage in facile handwaving about how to meet the developing world’s energy needs.

Policymakers need a greater focus on alternative strategies for dealing with any hypothetical future consequences of a changing climate. Most important is adaptation. Adaptation is autonomous—it’s what humans do. It is effective, it is proportional, and it is inherently local and achievable.

What could we really do about changing the energy landscape?

When it comes to energy technologies and policies, we need to acknowledge three fundamental long-run trends, which some policymakers are trying to bend with money.

First: the efficiency metric. Engineers will always pursue improvements in efficiency; that is inherent to progress. Thus, we see that the popular measure of merit—energy consumption per unit of economic output—has continually improved. But that has not reduced overall energy consumption. The long-run reality of greater efficiency stimulating greater demand was first documented in the mid-19th century by British economist William Stanley Jevons; it’s now known as “Jevon’s Paradox.” Jevons himself wrote at that time that, for the casual observer, it would “seem a paradox,” but he noted explicitly that the outcome of greater efficiency was to lower costs and thus stimulate demand.  

The second long-run trend is for energy-starved societies to see continual increases in energy use per capita as their wealth grows—the latter an inevitable and desirable feature of technological progress. Robert Solow received the 1987 Nobel Prize in economics for his work showing that “technology remains the dominant engine of growth.” And growth itself is stimulated in significant measure by the availability of more energy efficiency because all technologies necessarily use energy. Technology progress thus, symbiotically, boosts both energy efficiency and energy demand.

And the third long-run trend is an unwavering one, with surprisingly little variation: a gradual, multi-century decarbonization of civilization’s overall primary energy supply. This trend, too, will continue on its own.

Such natural long-run cadences of civilization have what can only be called high inertia. In general, societies aren’t willing to spend, or likely prove capable of spending, the magnitudes of capital to bend such trends off their natural course.

Many hold the well-founded conviction that there must be better energy technologies than what we have today. The issue is not if, but when such technologies may emerge as practical, and at scale. We know from history that foundational shifts in science—along with revolutionary shifts in technology—do occur. But they have an inconvenient trait, one that Bill Gates has framed as lacking a “predictor function.”

For now and the next few decades, the bottom line is that if we want energy revolutions, a stable society, and economic growth, we should stop squandering precious capital on yesterday’s technologies—and, frankly, on kleptocrats. The kinds of energy-technology revolutions that we all think might be someday possible, even likely, require something in short supply in policy domains: patience. The promises of radically new fission reactors, even data-center-scale micro-reactors, and new, quasi-magical energy-enabling materials like graphene, are tantalizing. The elusive goal of practical fusion will one day happen. There will be new physics, too, someday. If we want more foundational magic, we’ll need the patience to focus on re-animating open-ended basic research.

Meanwhile, civilization needs enormous amounts of low-cost energy, and it needs it from the technologies and systems that we know how to build right now. Engineers, entrepreneurs, and businesses can meet that challenge—but mainly using hydrocarbons.

Steven E. Koonin is a Senior Fellow at the Hoover Institution, advisor to the National Center on Energy Analytics, and author of Unsettled: What Climate Science Tells Us, and What It Doesn’t, and Why It Matters. Mark P. Mills is Executive Director and founder of the National Center on Energy Analytics and author of The Cloud Revolution: How the Convergence of New Technologies Will Unleash the Next Economic Boom and a Roaring 2020s.