Guns do kill people: reasons to worry about moral hazard in geoengineering

Many environmentalists call geoengineering a false solution, ethicists often worry that it indicates a worrying hubris about human domination over nature, and some economists suggest that it would encourage decision makers to take on more climate risk. What ties these concerns together is a view that the (apparent) availability of geoengineering technology might in one way or another lead humanity collectively to make less effort to mitigate the causes of climate change than we would have done otherwise – that in some way it would deter emissions reduction. Many label this effect a ‘moral hazard’ – echoing the economists’ explanation of why ‘too-big to fail’ banks happily took excessive risks with other people’s money.

The arguments raised against such a concern by advocates for geoengineering research often include ones from three groups: first, largely semantic objections to the term ‘moral hazard’; second, arguments that taking on more climate risk would be the rational response; and third, claims that experience with the adaptation debate somehow disproves the effect. In this blog I plan to briefly rebut each of these arguments, and instead make the case that we must properly recognise the social implications of technology, and therefore should already be gravely worried about the moral hazard effects of geoengineering[1].

Historically the term moral hazard has been used most widely in the insurance industry, originally where insurers were concerned that people of weak moral character would take advantage of insurance by being careless of insured risks, or even to defraud insurers through deliberate acts such as arson. More recent definitions tend to focus on the tendency of people to adjust their behaviour to a certain level of risk – so, in one classic example, the introduction of compulsory seat-belt wearing led to otherwise more risky driving and an increase in other types of car-related accidental injury. This is distinct from any concern about poor character, and arguably can even be socially beneficial (in the example, perhaps improving average journey times).

Advocates of geoengineering who say we shouldn’t worry about moral hazard are often simultaneously defining it in this narrow way (as risk adjustment, rather than any concern about moral character), and arguing that such risk adjustment – resulting in less mitigation - is reasonable because geoengineering reduces the social risk of climate change in a different way. They may also be subconsciously resisting the implication that conceding a “moral hazard” somehow labels them as immoral – but that’s another story. Whatever the motivation, many economic modelling studies suggest reduced or delayed mitigation when geoengineering (whether carbon dioxide removal, or solar radiation management) is included in the model.

But there are two problems here. First, such risk adjustment, even if rational, is not necessarily morally neutral. Again in the example of seatbelts, much of the new risk was not borne by the drivers concerned, but by back seat passengers and other road users including pedestrians. Many economic definitions of moral hazard recognise the possible injustice of risks transferred to other parties as a serious problem. Paul Krugman, for example, defines moral hazard simply as “any situation in which one person makes the decision about how much risk to take, while someone else bears the cost if things go badly.” In the case of geoengineering, we see the same: those who might decide to deploy it and - ‘rationally’ - reduce mitigation, are quite likely to be in a different country and indeed a different generation than those bearing the brunt of the impacts if geoengineering proves either defective or ineffective[2]. So there is potential for moral hazard. In particular, arguments that mitigation can be slowed or delayed mean transferring climate risks (arising if geoengineering might not work as expected) to a group who cannot then act on them by mitigation, because it will be too slow[3].

Second, we should not be so quick to dismiss the effects of poor character. Political and corporate elites are – apparently – susceptible to the temptations and moral hazards of corruption, and - according to psychological and sociological research - more likely to cheat and act unethically than poorer groups in society. And in psychological experiments, people with ‘self-enhancing’ values - those that relate to the accrual of wealth, power or status to oneself -were found to be more vulnerable to the moral hazard in geoengineering. So I suggest we should very concerned about whether the relevant decision makers have the necessary moral character to take tough decisions on mitigation and adaptation, rather than adopting the hope of geoengineering as an excuse for not upsetting political allies, campaign funders or particular groups of voters. Such a ‘political’ moral hazard is, I suggest, far more of a worry than the idea that ordinary individuals will relax their efforts to cut emissions. Indeed, it is actually possible that many ordinary people, exposed to geoengineering ideas, might see them as so wacky and unpalatable that they increase their support for mitigation – the so-called ‘negative moral hazard’. Unfortunately, the ‘warts and all’ presentation of geoengineering risks and potentials that might stimulate such reactions is very different to the positive impression advocates are forced to present in order to win political support and corporate funding.

But the other argument mobilised against concerns of moral hazard appears more grounded. Critics say: but look what happened with adaptation … climate campaigners treated it as taboo for years, for fear that politicians would use it as an excuse to delay emissions cuts, and now we are struggling to catch up with desperate needs for widespread adaptation. In other words, in the face of poor progress on mitigation, let’s not be caught with our pants down again, and start talking about (and researching and testing) geoengineering while there is still time. It seems a powerful argument, especially when its advocates compare the ‘silence on adaptation’ with the attitude of southern Baptists on sex education – don’t teach it, because it will discourage children from abstinence.

The main problem with this argument is that there is no way to test it – we have no counterfactual of a world where we debated and researched adaptation more and earlier, where we could scientifically observe either more adaptation happening, or maybe (yes it is possible) less mitigation happening than in our reality. But the second problem is that it simply isn’t true that adaptation has been ignored. Adaptation has been a central part of UN and IPCC deliberations from the very beginning, gaining particular prominence in political negotiations in the mid 2000s (when adaptation could be described as ‘the most fashionable item on the climate policy agenda’). Moreover - while academic articles are not the only indicator of serious attention - the scholarly literature on climate adaptation massively outweighs that on mitigation (see graph 1). Between 1985 and 2014 there were more than twice as many articles on adaptation than on mitigation and the gap declined during that period, from over ten times as many in 1985, to just 30% more in 2013. This is the opposite trend to that implied by the ‘we didn’t talk about adaptation for fear of moral hazard’ argument, which would suggest the mitigation literature should be more prolific, at least until more recent years[4]. In fact, the pattern of academic publishing rather better supports the contrary view, that perhaps, very widespread academic discussion of adaptation may have supported and enabled climate deniers and vested interests to resist effective mitigation policy in the US and many other countries.