Michael Shellenberger’s recent book Apocalypse Never: Why Environmental Alarmism Hurts Us All has received enormous media attention in the relevant circles. There has been much dispute over the book’s factual claims. But should Apocalypse Never be considered an ecomodernist work, as has been widely assumed?
In 2003, Shellenberger co-founded the ecomodernist think-tank The Breakthrough Institute, whose professed mission is to solve environmental and human development problems by technological means. Shellenberger also co-authored the 2015 “Ecomodernist Manifesto,” which fleshes out the institute’s vision, and gives it the label ecomodernism. Shellenberger left the Breakthrough Institute later that year, but he has never distanced himself from ecomodernism.
Ecomodernism does not have a precise or universally agreed upon definition. Like many other isms, it is an amalgam of interrelated ideas, susceptible to a variety of interpretations. I am the co-founder and current chair of a German ecomodernist society. Ecomodernism, as I understand it, includes the following three core tenets: the idea of a great anthropocene, an ideal of pragmatism and a flexible conception of nature. Shellenberger’s thought in Apocalypse Never is in conflict with all three.
- Great Anthropocene: We are currently at the onset of a new geological era that has been crucially shaped by humans: the anthropocene. The Ecomodernist Manifesto claims that we can shape the anthropocene in highly beneficial ways, through “knowledge and technology, applied with wisdom.” For the anthropocene to be truly great, all humans will have to soon enjoy living standards as high as those of people in industrialized countries today.
- Pragmatism: Ecomodernists recognize that trade-offs between human development and protection of the environment will typically be resolved in favour of the former, especially in developing countries where poverty is still rampant—so the ecomodernist’s goal is to eliminate the need for trade-offs. According to the Ecomodernist Manifesto, “any conflict between climate mitigation and the continuing development process through which billions of people around the world are achieving modern living standards will continue to be resolved resoundingly in favour of the latter.” The task at hand is to let living standards rise in the most environmentally benign ways that are achievable, not to solve environmental problems by preventing living standards from rising.
- Flexible Conception of Nature: The Ecomodernist Manifesto is written with “deep love and emotional connection to the natural world,” but it emphasizes that the concept of nature is flexible and that “there is no single baseline prior to human modification to which nature might be returned.” “Environments will be shaped by diﬀerent local, historical, and cultural preferences”—and that is totally fine.
Some of Shellenberger’s claims in Apocalypse Never are in perfect agreement with these aspects of ecomodernism. He relentlessly stresses the need for rapid economic development in currently poor countries, even when this conflicts with nature conservation. Shellenberger provides a captivating account of how people living close to Virunga National Park in the Democratic Republic of Congo, struggling to fulfil even their most basic needs, take firewood from the park to make charcoal. He paints a dramatic picture of the drastic and often brutal actions the authorities take to protect the park and its wildlife. Shellenberger’s compelling message is that locals urgently need economic development and richly deserve it. The message here is in line with both pragmatism and the belief in the possibility of a great anthropocene: these people’s dignity and autonomy must be respected; they must be enabled to be freed from poverty as quickly as possible; and they have the right to resolve trade-offs in accordance with their own preferences, even if that ultimately means things like oil drilling in Virunga National Park.
Wind and Solar Power
But Shellenberger also defends positions that are extremely unpragmatic—most notably in his categorical rejection of wind and solar power. He complains that these energy sources have low power densities, which means that they require large amounts of land, that they are material- and mining-intensive, and that they make electricity expensive (and thereby make it harder to mitigate climate change) because they require expensive dispatchable backup generation and/or vast amounts of energy storage. Shellenberger instead champions nuclear energy as an alternative clean energy source: its power density is two to three orders of magnitude higher than those of solar and wind power, it requires fewer mined materials and much less, if any, dispatchable backup generation.
Shellenberger’s claim that the environmental footprint of nuclear energy tends to be lower than that of solar and wind energy is defensible. So is his view that decarbonization may become significantly cheaper if nuclear energy enjoys a revival and is rolled out on a large scale. But, in many countries, the cost of new electricity generation from wind and solar power is currently so much lower than that of obtaining electricity from new sources of nuclear power, that expanding wind and solar plants will often be a far cheaper way of reducing emissions, at least in the short term, than building new nuclear power stations. Solar and wind power are also more popular and encounter less public resistance. Pragmatism, at least prima facie, suggests that paying the greater environmental price for the deployment of these sources will very often be worth it.
Shellenberger points to the experiences of countries like Germany and Denmark and of some US states that have massively expanded their use of solar and wind power and experienced an increase in electricity prices. But much of the increase in wind and solar power in those countries happened during a time when the costs of those energy sources were much higher than they are today, and, in fact, those country’s investments in those power sources seem to have helped bring down those costs. Comparisons of electricity prices in such countries with those in nuclear-heavy France, where the bulk of investments in nuclear power plants were made decades ago, are not very informative. Energy system modelling, even when it assigns a central role to “firm low-carbon sources,” such as nuclear energy, also assigns a central role to wind and solar power in carbon-constrained electricity systems, at least if costs remain at or near current levels and/or match current trends.
Reducing emissions from electricity at minimal cost today is only the first step towards a maximally economical zero-carbon energy system. Shellenberger may well be right that the cheapest achievable system of that kind will include a large amount of nuclear energy at the global level. As the most scalable, dispatchable, low-carbon source that we have, nuclear power may indeed be a key technology in establishing zero-carbon energy systems that can outcompete fossil fuel-heavy ones. Getting there will require shifts in public opinion—Shellenberger himself, to his credit, has contributed significantly to such shifts—and it will also require significant cost reductions and speed-ups in reactor construction, which in turn will require practice and the acceptance of first-mover disadvantages. There is a real danger that focusing on the comparatively cheap short-term emission reductions that wind and solar power offer today may prevent us from making the investments in nuclear energy that will enable affordable and effective deep decarbonization over the next few decades.
But taking this danger seriously does not mean that one has to abandon pragmatism about wind and solar power. In many cases, they offer by far the cheapest short-term emission reductions, and, depending on future developments in energy generation and storage technology, may continue to do so in the long term and perhaps permanently. Shellenberger cites a study that claims that solar and wind power are unable to support a modern industrialized society as a matter of physical principle (because their energy return on energy invested is allegedly too small). But this study has been persuasively critiqued, and more recent studies (notably here), not cited by Shellenberger, indicate that there are, in principle, no physical impediments to energy systems that rely entirely on “renewables.”
Shellenberg’s primary criticism of wind and solar energy is that they require too much land. He cites calculations by energy expert Vaclav Smil, according to which 25–50% of land in the US would have to be used for energy generation in order to sustain energy production at close to current levels using only “renewable” resources (see pp. 245–9 here). But, while Shellenberger only mentions solar and wind energy in this context, Smil’s results rely on a very large hypothetical contribution from biofuels, which require at least an order of magnitude more land than solar energy and significantly more than wind energy. As long as the goal is “merely” to significantly reduce emissions from electricity sources, the expansion of solar and wind remains an option and that will require far less land, especially since they can both be combined with other land uses.
Some countries could deploy offshore wind power plants to lessen the land required. Offshore wind has a significantly higher power density than onshore wind, does not interfere with land use at all, and has a less variable production profile than onshore wind. Under the UK’s current financing models, electricity from new offshore wind farms will be sold at less than half the price of electricity from its first new nuclear plant, Hinkley Point C.
The example of offshore wind highlights the tension between Shellenberger’s blatant dismissal of wind and solar power and his supposed pragmatism. As Shellenberger emphasizes, wind farms can pose risks to bird populations, and a 2016 Swedish investigation has confirmed that this also applies to offshore wind, especially in the case of wind farms in areas favoured by seabirds as feeding and overwintering grounds. But the same research also found that offshore wind farms had very positive effects on fish and marine mammal populations—probably because the turbines function as artificial islands that create additional habitats. This positive effect was found to be similar or perhaps even larger than that produced by designated marine protection areas. A recent report by Dutch consultancy Witteveen on Dutch plans to scale up offshore wind capacity to 11 gigawatts by 2030 concludes: “11 gigawatts of wind farms in the Dutch North Sea … can interfere with the proper functioning of the ecosystem. However, the wind farms can help to actively enhance the ecosystem. Smart adaptations to the design of an offshore infrastructure can help to kickstart rehabilitation of degraded North Sea habitat.” The report catalogues measures that could maximize these opportunities.
Ecomodernism embraces a flexible conception of nature. If we value marine biodiversity, why not let offshore wind farms contribute to it, since they also have the desirable side effect of clean energy production? The idea that only a “pristine” sea, “unspoiled” by wind farms, can be viewed as “nature” is incompatible with ecomodernism. This is not a defence of every conceivable offshore wind farm project, nor an argument for trying to cover all our energy needs using offshore wind. But there is no good, principled reason for ecomodernists to deny that wind and solar power can contribute significantly to our energy systems. Shellenberger’s categorical rejection of these sources is incompatible with both pragmatism and with a flexible conception of nature.
Shellenberger’s opposition to solar and wind energy is particularly perplexing in view of his confidence that alarmism about climate change is misplaced. Fellow ecomodernist Mark Lynas, author of the recent book Our Final Warning, believes that current global developments could realistically result in global heating of six degrees or even more, and potentially even cause runaway warming leading to a hothouse earth. Shellenberger, by contrast, holds that “global temperatures today appear much more likely to peak at between two to three degrees centigrade over preindustrial levels, not four, where the risks, including from tipping points, are significantly lower.” He bases this estimate on the International Energy Agency’s 2019 World Energy Outlook in conjunction with work by climate scientists Hausfather and Ritchie. The IEA’s future emission trajectories, however, rely significantly on the assumption that additional electricity generation capacity will in large part come from wind and solar power. According to the IEA authors: “Cost reductions in renewables and advances in digital technologies are opening huge opportunities for energy transitions. Wind and solar PV provide more than half of the additional electricity generation to 2040 in the Stated Policies Scenario and almost all the growth in the Sustainable Development Scenario.”
If, like Shellenberger, you do not believe that wind and solar power are capable of providing a large portion of global electricity generation and could significantly contribute to a reduction in emissions, it does not make sense to regard the IEA report as a reason to be any less pessimistic about future warming than you would otherwise be. If only nuclear energy can provide deep decarbonization, as Shellenberger claims, confidence that temperatures will peak at two to three centigrade over preindustrial levels is also misplaced.
Risks to a Great Anthropocene
Shellenberger is correct to criticize the overblown claims of some activists, according to whom humanity is headed for extinction over the course of the next one to two decades. But the attitude towards large-scale climate-induced risks expressed in Apocalypse Never is too dismissive to be compatible with a commitment to the possibility of a great anthropocene. Shellenberger focuses on discussions of extreme warming scenarios, in which vicious cycles involving tipping points, such as the melting of the icecaps or permafrost, result in unstoppable rises in temperature that could ultimately make Earth inhospitable to human life. Any such scenario, if it occurred, would cut short the geological anthropocene before it has really begun.
Shellenberger is dismissive of these extreme scenarios, which he views as “unscientific” due to the “high level of uncertainty” regarding tipping points and the complexity of the possible interplay of relevant factors. But, as Gleick points out, this is misleading: that there is scientific uncertainty concerning tipping points does not make it unscientific to take the possibility that a tragic combination of such events might occur into consideration. Quite the contrary.
Shellenberger argues that “there is no scientific evidence that [a catastrophic tipping point scenario] would be more probable or catastrophic than other potentially catastrophic scenarios, including an asteroid impact, super-volcanoes, or an unusually deadly influenza virus.” Let us grant this assumption for a moment: that the existential risk to humanity from climate change is about as large as those from asteroids, super-volcanoes and “naturally occurring” pandemics. The most systematic recent assessment of existential risks, undertaken by Toby Ord, assigns a probability of 1/10000 to extinction due to one of the two latter causes over the next century (and a much lower probability of 1/1 million to extinction from asteroid or comet impact). This may be too pessimistic, but we should bear in mind that human extinction would only be the most extreme—not the most likely—outcome of any of these calamities. The rational probability that climate change may provoke a global catastrophe causing hundreds of millions of casualties is at least one—and perhaps two or three—orders of magnitude greater than the probability that climate change will cause human extinction. So, if Shellenberger’s comparisons are apt, it may be rational to assign a likelihood of as high as 1/1000–1/10 to the possibility that climate change may cause the loss of hundreds of millions of lives over the next couple of centuries. According to the doctrine of the great anthropocene, the future can be “great” on geological time scales, so the value of what is at stake here is enormous. In the light of contemporary climate science, denouncing worries that climate change may pose a catastrophic or even extinction risk as alarmist is therefore incompatible with ecomodernism.
There are geological records of asteroid impacts, super-volcano eruptions and species extinctions due to pandemics that enable us to make somewhat informed guesses about our extinction risks from these causes. But there is no similar record of the anthropogenic risks that have only arisen recently. Climate change is one such risk, anthropogenic pandemics another. It is true that there were epochs in Earth’s history when CO2 levels were far higher than they are today, but the current rate of increase in those levels is unprecedented. As Mark Lynas writes:
the combined efforts of human beings to dig up and burn fossil fuels to power our global industrialised economy is taking place at least ten times faster than the catastrophic carbon release that drove the world’s worst-ever mass extinction … there is no episode in the planet’s geological history that truly mirrors what we are doing now in terms of the speed and volume of our greenhouse gas emissions.
For all we know, it may be rational to treat the risks from catastrophic tipping point scenarios as considerably higher than those from asteroids, super-volcanoes or natural pandemics. Shellenberger’s record as a climate activist may be perfectly consistent with such an attitude. His focus and rhetoric in Apocalypse Never are not.
Similar considerations apply to Shellenberger’s relaxed approach to the spread of nuclear weapons. He attributes the global decline in deaths from armed conflict since 1950 to the deterrent effect of nuclear weapons, argues that the risks from such recent and potential future nuclear member states such as North Korea and Iran are minimal, and is therefore not worried about weapons proliferation due to the expansion of civilian nuclear energy. A believer in the possibility of a great anthropocene could only credibly take this stance if it were supported by a detailed weighing up of the comparative risks involved.
The nuclear winter that would follow large-scale nuclear war could do long-term damage to human civilization. Some also regard it as an existential risk to humanity that could cut the anthropocene short when it has barely begun. Like climate change, under certain circumstances—however improbable—nuclear weapons could lead to hundreds of millions of casualties. For a believer in the possibility of a great anthropocene, that is an extraordinary waste of human potential. Even if the probability of nuclear war is small, an ecomodernist would not dismiss it without a compelling argument that it is so small that it is worth disregarding, even if it puts the great anthropocene at risk. If Shellenberger is correct, deterrence may dramatically reduce the probability of nuclear weapons being deployed per unit of time. But an increase in the number of countries with nuclear weapons may increase that probability. Only someone who believes that the future of humanity is unlikely to be very valuable (and may well be short in any case) could think it rational to accept the increasing proliferation of nuclear weapons unquestioningly. Such a person would not be a believer in the possibility of a great anthropocene.
Animals in the Anthropocene
Shellenberger touts the health benefits of eating meat and the environmental benefits of industrial over free-range animal husbandry. The comparatively modest land use requirements of industrial beef and the short lives of industrially kept cows make this the environmentally virtuous option compared with grass-fed beef, according to the author, who quips that we can “have our steak and eat it, too.”
However, as Shellenberger is aware, industrial animal husbandry is widely believed to cause suffering on an immeasurable scale. As Yuval Noah Harari points out, the conditions in which domesticated animals are kept on industrial farms are not aligned with their evolved psychological needs and there is robust evidence of their suffering, for example when calves are separated from their mothers immediately after birth. If the anthropocene is to be great, humans will care about the psychological well-being of other sentient beings.
Converting everyone on the planet to vegetarianism or even veganism is neither pragmatic nor necessary. But Shellenberger does not mention plant-based or cell-based substitutes for animal products, even though these offer the kind of technological solutions to the ethical and environmental problems of animal farming that ecomodernists have called for. Since the land use requirements of plant-based ingredients tend to be about one order of magnitude less than those of industrially raised animal-based food, this is a curious omission—especially since Shellenberger places so much emphasis on land use when it comes to the choice between nuclear and renewables. Clearly plant-based meat replacements are the nuclear option when it comes to energy sources that can serve our protein needs.
To his credit, Shellenberger provides one valid counterargument to the ethical objection against industrial farming on a global scale: namely, that there may be greater animal suffering in the wild than on industrial farms. Wild animals can, indeed, undergo gruesome suffering, as examples collected by animal activist and effective altruist Jacy Reese highlight: “Gulls peck out and eat the eyes of baby seals, leaving the blinded pups to die so they can feast on their remains. A shrew will paralyze his prey with venom so he can eat the helpless animal alive, bit by bit, for days.” Leaving wild nature alone is often associated with ecomodernism, so interfering with these instances of suffering may not seem to be an option. Surely we cannot complain about animal suffering in industrial agriculture if we tolerate the suffering of wild animals on a global scale?
In light of ecomodernism’s flexible conception of nature, there is no principled reason to think that we cannot or should not. We may not yet have the knowledge, skills or resources to implement large-scale animal vaccinations, birth control or other measures to decrease suffering in wild populations. But thinking in terms of a potentially great anthropocene means thinking in geological time scales. Hopefully, we will not succumb to any lurking existential risks and will have plenty of time to develop such measures.
Despite its flaws, Apocalypse Never is a passionate call for “nature and prosperity for all.” But as far as pragmatism, global risk minimization and human and animal welfare maximization are concerned, it is neither bold nor ambitious enough. At least not by the standards of this ecomodernist.