Climate obligations and social norms

Abstract

Many governments are failing to act sufficiently strongly on climate change. Given this, what should motivated affluent individuals in high-consumption societies do? This paper argues that social norms are a particularly valuable target for individual climate action. Within norm-promotion, the paper makes the case for a focus on anti-fossil fuel norms specifically. Section 1 outlines gaps in the existing literature on individuals’ climate change obligations. Section 2 characterises social norms. Section 3 provides seven reasons why social norms are a particularly worthy target for individual climate actors. Section 4 asks which social norms individuals should emphasise, arguing that anti-fossil fuel norms have advantages over emissions reduction and offsetting norms. Section 5 outlines the pathways and mechanisms individuals might exploit to promote anti-fossil fuel norms.

Many governments are failing to act sufficiently strongly on climate change. Given this, what should motivated affluent individuals in high-consumption societies do? ¹ An individual lacks control over public policy; she cannot single-handedly be obliged to change that. Meanwhile, marginal individual differences (via private offsetting or reduced consumption) seem to ignore the large-scale structural nature of the problem. Navigating between public policy and private consumption, this paper discusses the value of acting upon social norms. Specifically, I argue that ‘anti-fossil fuel’ (AFF) norms are a fruitful target for affluent individuals’ climate efforts.

Section 1 begins by engaging with the philosophical literature on individuals’ climatic obligations. Although philosophers have argued that individuals have climatic obligations – and have identified social norms as one possible target of individual climate action – I explain that these discussions have left unaddressed the questions of (1) what social norms are (and, therefore, what it means to target them); (2) why social norms are a particularly worthy target for individuals and (3) exactly which social norms individuals should target. The rest of the paper addresses these questions.

Section 2 characterises social norms, bringing precision and action-guidance to the injunction to target them. Section 3 provides seven reasons why social norms are a particularly worthy target for individual climate actors. Section 4 asks which social norms individuals should emphasise, arguing that anti-fossil fuel (AFF) norms have advantages over emissions reduction and offsetting (ERO) norms. Section 5 gets concrete, outlining the pathways and mechanisms individuals might exploit to promote AFF norms.

Before I begin, one might wonder: given inaction from states, why focus on individuals’ obligations? Surely non-state collective agents – including corporations, charities, universities and civil society groups – also have climatic obligations to target social norms, and perhaps AFF norms specifically? ² Probably. However, the psychology, social dynamics and incentive structures of collective agents are sufficiently different from individual agents, that the arguments of Sections 2–5 will not neatly transpose onto collective agents. As we’ll see, humans engage with social norms – and, through social norms, with each other – via feelings of respect, approval, offence, guilt and so on. It's unclear whether collective agents are motivated by such feelings in quite the same ways (Björnsson and Hess 2017; Hindriks 2018; Collins 2023, ch. 5). I leave the precise actions demandable of collective agents to another day.

1. Individuals’ climatic obligations: existing accounts

1.1. Individuals have obligations

Given states’ insufficiency on climate change, which actions should individuals take? Some will think this question contains a false presupposition: the presupposition that individuals should take any action at all. After all, one might say, individuals don’t do climatic harm (Sinnott-Armstrong 2005, 297–298; Sandberg 2011, 231–232; Garvey 2011, 160; Cripps 2013, 119–124 MacLean 2019, 6). If that's correct, and if climatic obligations must be justified by an agent's contribution to climatic harm, then individuals are off the hook. What's more, if individuals don’t do climatic harm, then one might doubt whether individuals have the capacity to do climatic good. And if individuals cannot do climatic good, then we cannot justify their obligations via their capacity to do good.

Why think individuals do not do climatic harm and cannot do climatic good? Because most effects of climate change are not made different by any actions or omissions of a given individual: if we compare worlds in which one individual never flies again with the worlds in which that individual flies at every opportunity, we can expect the latter worlds to contain almost all the same climate harms as the former ones. And when we consider those climatic effects that are made different (whether positively or negatively) by a given individual's actions or omissions, we find that the difference the individual makes to these effects is so small as to preclude obligations. Or so the argument goes.

This ground has been traversed by others, so I will simply mention some plausible paths to individuals’ climatic obligations. First, Derek Parfit has proposed that ‘[e]ven if an act harms no one, this act may be wrong because it is one of a set of acts that together harm other people’ (1984, 69, emphasis original). Second, Julia Nefsky has argued that there are reasons against performing actions that non-superfluously ‘help’ to cause harm, even if those actions make no difference to that harm (2017, esp. 2753). By these principles, individuals’ contributions to climate harm are wrong, even if they make no (or insignificant) difference.

Analogous principles apply to individuals’ capacities to do good. An action might be morally required if it is part of a set of actions that would make a significant positive moral difference, even if others don’t perform their actions in the set (Woodard 2008, ch. 4). Elizabeth Cripps calls these ‘mimicking duties’ – duties to mimic one's part in a good-doing collective scheme (2013, 116). Or, an action might be morally required because it non-superfluously helps to produce some good, despite making no difference to whether that good occurs (Nefsky's (2017) ‘helping’ account encompasses both harms and benefits).

A different path to individuals’ climatic obligations points to the possibility that an individual's actions make a large difference. An individual might trigger climate ‘thresholds’, such as the permafrost melting or the Amazon rainforest dying out. Multiplying the badness of such thresholds by an individual's chance of triggering them, we find that many individual choices to fly, drive, eat meat, and so on do much more expected harm than expected good (Lawford-Smith 2016, 75–77; Broome 2019, 114–115). Via such ‘expected value’ reasoning, we can justify individuals’ climatic obligations using both the expected climatic harm of individuals’ past actions and the expected climatic benefits of their possible future actions or omissions.

In this paper, I assume such arguments produce individuals’ climatic obligations. I further assume these are obligations of justice: they correlate with claims or rights that the obligations be performed. Again, existing work supports this assumption. First, Darryl Moellendorf proposes the ‘principle of vulnerability reduction’, which he applies to climate change: ‘If a person is especially vulnerable to very bad things happening due [to] the actions and omissions of others, that person has a prima facie claim to have the vulnerability reduced’. (2015, 175) Second, Steve Vanderheiden argues that the unequal causation of climate change (it is produced predominantly by the advantaged), plus the unequal ill effects of climate change (which are experienced predominantly by the disadvantaged), render climate harms unjust by egalitarian standards (2018, 116–117). Finally, several authors argue that climate harms are violations of human rights (Hayward 2007; Caney 2010; Bell 2011; Shue 2011) – which, I assume, individuals have natural duties of justice to help protect.

This paper focuses on the following question: which actions do individuals’ climatic obligations demand? This differs from the question of how costly those obligations are. ³ Suppose I am required to spend one hour and $200 per month on my climatic obligations – suppose, then, that we know how costly my obligation is. Another question remains: what should I do with my hour and my $200? That is the question on which this paper is focused. I will consider actions of which one can do ‘more or less’, such that our answer to the action question is not dependent (within reason) on our answer to the costliness question. Of course, the more costly the obligation, the more action-types one can perform in the obligation's service. Still, regardless of our obligations’ costliness, it will be useful to know which actions are more valuable than others. That is the issue I address.

1.2. Norm-change as one possible action

In a way, the climate justice literature has answered our question: I must perform actions that contribute to mitigation, adaptation, and compensation. For example, I might contribute to mitigation by becoming vegan, to adaption by supporting organisations that build flood walls, and to compensation by supporting organisations that provide payments to disaster victims. Thus, Simon Caney has proposed the following principle: ‘For any agent, her course of action should be informed by the opportunities available for mitigating, adapting, or compensating climate change and opportunities for enabling/inducing others to mitigate, adapt, and compensate climate change’ (Caney 2016a, 17).

Caney's principle is highly plausible. It is also highly abstract. One might hope that climate ethicists could say which opportunities are promising, by and large, at least for particular groups of people (such as the group ‘affluent individuals in high-consumption societies’). Caney resists such conclusions, because individuals’ circumstances are highly variable (2016a, 18–19). However, he does mention anti-fossil fuel campaigning as one of many actions his principle might support (2016a, 17). Section 4 will provide grist to this particular mill.

Elsewhere, Caney alludes to the value of social norms more generally. He provides no fewer than six action-types that motivated agents might perform for their climatic obligations, given others’ noncompliance (Caney 2016b, 26–35). These are: reducing ambitions for responding to climate change; re-allocating noncompliers’ obligations amongst compliers; as compliers, imposing higher burdens on third parties; as victims, imposing burdens on noncompliers; violating non-justice ideals in pursuit of climate justice; and inducing non-compliers to comply by changing incentive structures. The sixth action-type is what Caney elsewhere calls a ‘second-order’ responsibility, that is, a responsibility of agents to ensure other agents alleviate climatic harms (2014, 134–5). Under this sixth action-type, Caney mentions the ‘fostering and maintaining of norms of environmental sustainability’ as one possible action (Caney 2016b, 34). When discussing second-order responsibilities, he also speaks of ‘creating norms that discourage high emissions lifestyles (or, alternatively, that foster a commitment to adaptation)’ (2014, 137).

Caney thus provides a thorough inventory of possible actions – an inventory that includes norm-targeting actions. However, he leaves unanswered the questions of (1) what social norms are (and, thus, what it means to ‘foster and maintain’ or ‘create’ them), (2) whether and why tackling social norms might be preferable to other potential climatic actions and (3) exactly which social norms should be promoted in the climate change context. (In Section 4, I will argue that anti-fossil fuel norms have some advantages over Caney's ‘norms that discourage high emissions lifestyles’ – the latter being ERO norms.)

Caney is not alone. Steve Vanderheiden also acknowledges social norms’ potential. He notes that the ‘key mediating role between individual action on climate change and the adoption of state policies to more effectively address it is played by social norms, which persons reinforce through their adherence but can challenge and disrupt through their public violation’. (2018, 125) Vanderheiden states that ‘norms allowing for excessive individual emissions like those resulting from joyguzzling have been both cause and effect of inadequate state mitigation actions’. (2018, 125). He briefly proposes norms in favour of walking or cycling (rather than driving), and powering homes with wind or solar (rather than coal or gas) (2018, 126).

Again, Vanderheiden's comments are plausible. However, Section 3 will provide a more thoroughgoing argument for why social norms are important in the climate change context. Further, Section 4 will argue that Vanderheiden's proposed norms – which are examples of ERO norms – face disadvantages compared to AFF norms. More generally, more action-guidance can be brought to Vanderheiden's comments. Sections 2–5 offer this. ⁴

1.3. Widespread social change

Other authors have argued that individuals should target widespread social change. For example, Anja Karnein writes that climate change requires us to ‘shift the focus away from our ill-equipped commonsense ethics [which doesn’t consider social context] and direct it toward an understanding of the institutional structures which embed the kind of mindset that would have to be changed’ (Karnein 2018, 950). Christian Baatz suggests that ‘[t]he core of the [climate] problem is that agents are constrained by the social and economic structures they rely on, and that collective action is required to change these structures’ (Baatz 2014, 15). And Elizabeth Cripps argues that ‘promotional duties’ are individuals’ ‘primary’ climatic obligations (2013, 140). Promotional duties require individuals to ‘attempt to bring about the necessary collective action’ (ibid.). Yet Cripps ‘leaves open innumerable ways in which an individual might fulfil promotional duties’. (2013, 143) Karnein and Baatz likewise do not provide specific guidance on how to enact widespread social change.

Targeting social norms is just one way to target widespread social change. Others include voting, petitioning, protesting, and resisting. Section 3 argues for social norms specifically. To funnel us there, this sub-section joins the above authors in advocating actions that are likely to promote widespread social change.

To see the importance of widespread social change, consider offsetting. Offsetting involves giving money to climate-preserving projects (such as tree-planting), to take greenhouse gases from the atmosphere equivalent to what one has put in, making it possible to offset so it's as though one had emitted nothing – even if one continues to engage in extremely luxurious emissions. This is made possible by offsetting's current low prices: on a recent estimate, the average US resident can offset their lifetime CO₂-equivalent emissions by giving $12,000 to offsetting charities (Broome 2021, 289). John Broome has previously argued that offsetting exhausts individuals’ climate-related obligations (Broome 2012, 14, 66–67). ⁵

However, as Kai Spiekermann explains, offsetting ‘fails as an institution’: if everyone did it, it wouldn’t work (2014, 927). That's because offsetting is affordable only if it focuses on cheap interventions (such as providing people with more efficient cooking stoves). The scope for cheap interventions would soon be exhausted, if everyone started offsetting. The remaining offsetting options would be increasingly expensive. Offsetting could not successfully mitigate the harmful effects of everyone's emissions, without being prohibitively expensive. This shows that we cannot advise each individual to mitigate by offsetting, without any regard for the possibility of widespread social change. The combination of all individuals acting on such advice would be highly inefficient.

Related points apply to actions of adaptation and compensation. Suppose an individual quietly donates to a small charity that helps farmers in a specific drought-hit area (adaptation) or that raises funds for people whose homes have been destroyed in local floods (compensation). Of course, such actions are morally worthy; in emergencies, they are even necessary. But if every individual made such actions the totality of their long-run approach to climate action, then we would miss out on more pro-active and effective collective schemes for adaptation and compensation (such as state-run schemes that can guarantee their benefits). Again, such collective schemes amount to society-level change, rather than piecemeal and reactive interventions.

Of course, one might contend that widespread social change is extremely unlikely. Most individuals are recalcitrant. For those few individuals who are motivated to act and who ask the question of what they should do, mightn’t offsetting (and other piecemeal and reactive approaches) be enough? In response: such actions might be part of an overall defensible approach for such individuals in the very short-term. But they fail to address the systemic nature of the problem, precisely because they would be inefficient at scale. As long as widespread social change remains possible, we should put more emphasis on actions that promote it than on those that don’t, because widespread social change is necessary to address the scale of the problem.

To summarise Section 1: individuals have climatic obligations of justice. These can be performed in numerous ways. Philosophers have mentioned social norms in this context, but without telling us what norms are, why norms are especially important, or why particular norms should be targeted. Other philosophers have argued that individuals should target widespread social change, but without comparing specific ways of doing so. The rest of the article addresses these issues.

2. What are social norms?

In International Relations, the classic definition of norms comes from Martha Finnemore and Kathryn Sikkink: norms are ‘standards of appropriate behavior for actors with a given identity’ (1998, 891). In James March and Johan Olsen's (1995, 30–31) terms, again designed for International Relations, norms obey a ‘logic of appropriateness’ rather than a ‘logic of consequences’. This distinction demarcates socially proper or fitting behaviour from behaviour that pursues an agent's one-off self-interest.

Recent work in the philosophy of social science has yielded more thoroughgoing frameworks. ⁶ Cristina Bicchieri contrasts norms with customs (2017, 15ff.). In Bicchieri's framework, ‘customs’ are reliable behavioural patterns within a group that are not undergirded by appropriateness. People might conform with a custom without caring whether others do, for example by taking umbrellas when it's raining. Or a custom might arise because people take others’ behaviour as evidence that the behaviour is worthwhile, as when certain foods come into fashion. Or a custom might arise because people want to be like other group members, without thinking this is what's expected of them – as when people follow fashions set by celebrities. Or a custom might arise because it's valuable that everyone does the same thing, without one thing being more appropriate than any other – as when shoppers keep to one side in supermarket aisles, without left or right being viewed as more appropriate (though such customs often quickly become norms). ⁷

We can demarcate a unified and interesting conceptual category by, following Bicchieri, stipulating that the crucial difference between customs and norms is that we follow a norm because of what others will think of us, or what others will do to us, if we don’t follow it: we follow norms to gain others’ respect and appreciation, or at least to avoid gaining their disapproval or offence (Bicchieri 2017, 38–39). And we often feel guilt when we violate a norm (Brennan et al. 2013, 20–21). This differentiates norms – or, at least, a unified and interesting category that I will label ‘norms’ – from customs. Examples of norms include wearing black at funerals, shaking hands on greeting, or taking wine to dinner parties. These practices encode respect when followed; they produce disapproval and offence when not followed (and, often, guilt for the violating party). These points will matter for Section 3's arguments: these reactive attitudes make norms motivationally powerful. And it is generally much more costly – socially speaking – for an individual to deviate from a norm than to deviate from a custom (Bicchieri 2017, 108). This implies that changing or abandoning existing norms is more socially costly than introducing a new norm into a domain in which there was previously nothing but custom. Again, I will return to this in Section 3.

Importantly, following social norms need not involve consciously thinking about respect, offence, guilt, and so on – or having recognisable and isolatable ‘beliefs’ or ‘preferences’ about these (Davidson and Kelly 2018, 196–198; Bicchieri's framework misleadingly implies otherwise). Social norms almost always operate covertly – in the fast, intuitive, automatic, and unconscious processes of our minds. And others’ sanctioning of our violations can be as subtle as a slight frown or momentary pause. Norms’ covert nature renders them resistant to change or abandonment (since we often don’t realise we’re enforcing them). Again, the difficulty of changing or abandoning existing norms will matter in the arguments to follow.

While remembering that norms’ adherents often lack isolatable ‘beliefs’ and ‘preferences’, I will tentatively adopt Bicchieri's framework, according to which:

A social norm is a rule of behaviour such that individuals prefer to [i.e., will] conform to it on condition that they believe [perhaps tacitly] that (a) most people in their reference network conform to it (empirical expectation), and (b) that most people in their reference network believe they ought to conform to it (normative expectation). (2017, 35)

For example, flying for international conferences is a norm within a particular group just if: individuals in that group will fly for international conferences given that (1) they believe most people in the group do in fact fly for international conferences and (2) they believe most people in the group believe those in the group ought to fly for international conferences. ⁸ If we have (1) without (2), then we have a mere custom. The ‘ought’ in (2) is not itself moral ought (though it might be backed up by moral oughts; see Section 4). The ‘ought’ is rather that of social appropriateness: the ought of what is socially apt, fitting, suitable or proper for people within the group. On this account of norms – which I adopt as a way of stipulatively picking out the phenomenon of interest – changing, abandoning and introducing norms is a matter of (1) changing people's ‘empirical’ expectations about what others in fact do, and/or (2) changing people's ‘normative’ expectations about what others think they ought (socially appropriately) to do.

With this sketch of norms in hand, why are norms a particularly worthy focus for individual climate action?

3. Why focus on norms?

Acting upon social norms (that is, acting to change others’ empirical and normative expectations) is just one way to promote the widespread social change necessary for meaningful climate change mitigation, adaptation, and compensation. Other such actions include voting, petitioning, protesting, and resisting. My claim is not that individuals must focus solely on targeting social norms, at the complete expense of these other actions. And these other actions might themselves be ways to act upon social norms, at least in some contexts. My claim is rather that norm-targeting is, in various ways, distinctively valuable. We must hedge our bets across action-types, but that doesn’t mean giving all action-types equal attention. In this section, I outline seven reasons to favour norm-targeting. The seven reasons are ordered from those least specific to norm-targeting (as against voting, petitioning, protesting, and resisting, without affecting norms), to those most specific to norm-targeting.

First, then, norm-targeting is supported by research on the attributes of the most climate-responsive individuals. Prominently, Whitmarsh et al. have developed the notion of an individual's ‘carbon capability’. This is ‘the ability to make informed judgments and to take effective decisions regarding the use and management of carbon, through both individual behaviour change and collective action’ (Whitmarsh et al. 2011, 59, emphasis added). They write: ‘fully carbon capable actors will be aware of, and seek to influence through collective and political mechanisms, the … policies, systems of provision, infrastructure, etc. … in order to overcome the structural barriers to low-carbon lifestyles and societies’ (ibid., emphasis added). That is: ‘carbon capable’ individuals act upon individual lifestyles via political outcomes. ⁹ Social norms sit exactly on the nexus of these two domains, as I’ll go on to explain (under reasons five and six, below). Thus, if motivated individuals should try to enhance their own carbon capability, one effective way for them to do so is to promote climate-friendly social norms.

Second, the injunction to target norms is action-guiding, while allowing contextually-specific implementation. Bicchieri's analysis of norms (as empirical and normative expectations) directs individuals’ attention and actions to others’ expectations, thus providing action-guidance. However, this guidance is contextually-specific. For example, quietly forgoing one flight will not contribute to a norm against flying, if no one sees the forbearance. ¹⁰ Even loudly forgoing one flight will not necessarily contribute to an anti-flying norm: it will not do so if the vast majority of one's network perceive the action as objectionable ‘virtue signalling’ or ‘moral grandstanding’ (Tosi and Warmke 2020, 37–40), to which they respond with sanctioning rather than changing their empirical or normative expectations (Monin et al. 2008, esp. 89). In general, what an individual should do to change others’ empirical and normative expectations will vary, with the relationship between the individuals, the context, and the behaviour in question. Thus, the obligation to target norms is appropriately flexible and context-specific. This differs from injunctions to vote green, petition, protest or resist. We could make the injunctions to perform these actions conditional on a lack of backfiring, but then the injunctions would become so conditional that they would lose their widespread applicability. Better, then, to follow the widely-applicable – yet flexible – injunction to target norms.

Third, norm-targeting is apt regardless of how we distribute the costs of climatic obligations amongst individuals. Three principles are often used to distribute costs: ‘polluter pays’, ‘ability to pay’, and ‘beneficiary pays’ (Caney 2014, 125–6). In Section 1.1, I briefly explained how individuals contribute to climate harm and can alleviate climate harm, in expectation. And all affluent individuals benefit from the resource use that causes climate harm. Thus, all three distributive principles can be used to place costs on affluent individuals. Further – and to the point – norms are a crucial mechanism through which these three distributive principles apply to individuals: norms encourage individuals to cause climate harm and reap climate benefits, and (as I’ll explain) targeting norms is a promising way for individuals to do climate good. So, norm-targeting directly acts upon the basis of climatic obligations, whichever basis one endorses (polluting, ability, or benefiting).

Aside from these three principles, there is another perspective on climatic obligations. As Caney puts it, “[t]his second perspective takes as its starting point the imperative to prevent climate change, and it works back from this to deduce who should do what.” (2014, 126) The focus is on responding to climate change as fully and effectively as possible. From within this second perspective, too, norm-targeting is apt: by targeting norms, individuals ultimately affect others’ behaviours, not just their own. The ‘ripple-effect’ of norm-change enables the individual to contribute to a fuller and more effective collective response – certainly when compared to actions like non-publicised offsetting, and arguably too when compared to voting, petitioning, protesting, and resisting – which can be politically divisive and alienating, as explained above. By definition, promoting climate-friendly norms has positive effects on others; if it does not, it is not really norm-promotion.

Fourth, social norms are often a better means to ongoing, long-term, mutually monitored behavioural change, as compared with economic incentives (Bicchieri 2017, 153–6). This is particularly so when the social norm is backed up by a moral norm (as is the case for climate-friendly norms). Simply, this is because – in many contexts – people are more motivated by ‘doing the appropriate thing’ or avoiding social sanction, than by money. ¹¹ One famous example is a study of childcare centres in Israel, who were facing problems with parents picking up their children late. When a fine was introduced for late pick-ups, the number of late pick-ups increased significantly – arguably, because what previously risked social disapprobation became a purchased service (Gneezy and Rustichini 2000; for further examples and discussion, Gneezy, Meier, and Rey-Biel 2011). Yet economic incentives and sanctions (broadly construed) are the end-game of voting, petitioning, protesting, and resisting – at least assuming these actions aim to induce governments to economically disincentivise climate-unfriendly behaviour.

Fifth, government-led climate action (which, again, is the focus of voting, petitioning, protesting and resisting) is more likely to be successful if it doesn’t stray too far from existing social norms (Bicchieri 2017, 144–147; Axelrod 1986, 1107; Posner 1998, 790; Muldoon 2017, 267). If, on the other hand, social norms don’t support government-led climate action, then legislators land in a ‘governance trap’:

[One] consistent finding in surveys of public opinion on climate change is that governments are seen as primarily responsible for taking action and exhibiting leadership on this issue; this is part of what has been termed the ‘governance trap’ of climate change between the public and politicians, whereby governments claim that they would be more ambitious on climate change if they had the electoral mandate, while the electorate looks to the government for leadership. (Corner et al. 2015)

In short, altering norms is one effective way to escape this governance trap, so as to produce socially-mandated government action. Consider the Regional Greenhouse Gas Initiative of the Northeastern US – where supporters of a cap-and-trade system had legal success largely via a ‘normative reframing’, in which the atmosphere was conceived as a public space that companies should pay to use (Raymond et al. 2014, 205). A new social norm underpinned government action.

Sixth and relatedly, if a policy outcome is supported by social norms, then that outcome is more stability-feasible than one not supported by social norms. To illustrate, suppose voting, petitioning, protesting, and resisting surprisingly achieved policy change without any corresponding social norms. The relevant policy would then be unstable. On Holly Lawford-Smith's analysis, there are two dimensions to any outcome's feasibility: access-feasibility and stability-feasibility (Lawford-Smith 2013, 252). Access-feasibility concerns the routes to the outcome: are they viable? Stability-feasibility concerns whether the outcome is stable, once achieved. Stability-feasibility is enhanced if the actions that sustain the outcome generate autonomous motivation – that is, if people perform those actions without being coercively legally mandated (Lawford-Smith 2013, 255). As explained in Section 2, norms often become internalised: they are performed unthinkingly with an internalised sense of social appropriateness. Thus, social norms that support a policy render that policy stability-feasible, compared to a policy that manages to be implemented in opposition to most people's sense of appropriateness.

That said, norms’ very internalisation makes them difficult to change or abandon, as I emphasised in Section 2. This raises the question: are climate-friendly norms access-feasible? And is norm-targeting more access-feasible than voting, petitioning, protesting or resisting, in the context of climate change? If so, that would be a seventh reason in favour of norm-targeting.

Let's take the second question first. Norms in many rich countries oppose protesting and revolution, making widespread social change by way of those actions relatively access-infeasible. (For example, in a relatively recent German survey, ‘taking part in environmental/climate campaigns’ was by far the least popular climate action (Stoll-Kleeman and O’Riordan 2020, 947). Monin et al. (2008, 76–77) explain how people dislike ‘moral rebels’.) This is not to say that all individuals find it highly difficult or costly to engage in contentious politics, or that individuals cannot have obligations to do what's difficult. It's to say that successful climate rebellion will require changes in norms, namely norms that oppose rebellion. By contrast, voting green or petitioning are perhaps less stigmatised, so may be more well-received by others, so may be a more access-feasible route to widespread social change. However, even here, the prospects for widespread change via voting or petitioning would be enhanced if social norms changed first. After all, voting and petitioning require large numbers in order to be successful. A change in social norms is an enabler for the access-feasibility of these other actions.

But is norm-change itself access-feasible? Here, the distinction between customs and norms becomes important. Recall that changing or abandoning existing norms is more socially costly than introducing a new norm into a domain in which there was previously nothing but custom. So, climate-friendly norms are more access-feasible if they are usurping existing customs, rather than upsurping existing norms, all else being equal. Fortunately, much current high-emissions behaviour is a custom, not a norm. To see this, consider fashion, transport, and food. High-emissions options in these domains are cheaper and more convenient than their sustainable alternatives: they follow a logic of consequences, not a logic of appropriateness. Of course, the norm of fashionable clothing (say) contributes to emissions, since much fashionable clothing is ‘fast fashion.’ Yet this contribution is contingent, since fashionable clothing could (for all the norm says) be slow fashion. ‘Dress fashionably’ is a norm. ‘Dress fashionably via high emissions’ is not.

The point can even be made about flying. While traditionally (and still) associated with high social status, flying also has a small (but growing) association with feelings of guilt and shame (Gössling et al. 2020, 1–2; Oswald and Ernst 2021, 81). This phenomenon even has a name: ‘flight shame’ (Gössling et al. 2020; Timperley 2019). If people are embarrassed or ashamed to do something, there's not norm in its favour – at least, not within those people's social network. Again, norms might mandate ‘visit family’, ‘attend international conferences’ or even ‘travel’ per se, as part of a socially desirable lifestyle (Oswald and Ernst 2021, 71). Flying may be the easiest means to following these norms. But it's doubtful that ‘flying’ in itself is a norm, particularly within the small but growing group who experience flight shame. Flying follows a logic of consequences, not a logic of appropriateness. Flying is not ‘apt’ – not like how shaking hands upon meeting, or wearing black at funerals, are ‘apt’. In this sense, there is no norm in favour of flying per se.

Another example is eating meat. In affluent countries such as the US, UK and Australia, people strongly report social norms in favour of eating meat, to which people with a strong national identity are likely to conform (Nguyen and Platow 2021, 7). ¹² Here, there does seem to be social ‘aptness’ to meat-eating. But, importantly, there are more and less emissions-intensive ways to produce meat. Is there a norm in favour of ‘consuming meat that has been produced in a high-emissions way’? It's doubtful: emissions-intensive meat is cheap and convenient, suggesting a logic of consequences rather than a logic of appropriateness. Again, high emissions are a means to following the norm. The norm mandates meat, not emissions.

If the link between norm-following and high emissions is contingent, then we can retain the norm while reducing the emissions, by severing the contingent link – for example, by maintaining the norm of visiting family while promoting a new norm of using trains to do so where possible, or retaining the norm of eating meat alongside a new norm of low-emissions meat. Of course, such changes cost time (for travel) and money (for meat). But they do not have an additional cost: the cost of opposing or altering an existing norm. This makes climate-friendly norms more access-feasible than one might think.

To summarise, the injunction to promote climate-friendly social norms is: (1) accordant with psychological ‘carbon capability’; (2) action-guiding yet flexible; (3) accordant with climatic obligations’ justifications; (4) aimed at more than economic incentives and sanctions; (5) enabling of a mandate for government action; (6) productive of stability-feasibility; and (7) relatively access-feasible because consistent with existing norms. While voting, petitioning, protesting and resisting share some of these benefits to some extent, the overall case here favours actions that target norms. After all, by targeting norms, we not only effect change directly – we also make these other actions more likely to succeed, thus multiplying the value of norm-targeting. But what are some examples of climate-friendly norms that might be promoted by affluent individuals in high-consumption societies? And how might they go about promoting these norms? I address these questions over the next two sections.

4. Which norms?

In Section 1, I mentioned the distinction between mitigation, adaptation and compensation. Targeting social norms is consistent with targeting norms related to mitigation, adaptation or compensation – or a combination. It is beyond the scope of this paper to compare all possible norms that individuals could promote. Instead, I limit myself to comparing two different types of norms: emissions reduction and offsetting (ERO) norms and anti-fossil fuel (AFF) norms, both of which I will characterise shortly. My aim is to argue that the latter have advantages over the former. This is significant, since ERO norms are the most commonly-mentioned norms in climate ethics – as suggested by Section 1.1's discussion of Caney and Vanderheiden, and Section 3's mention of ‘flight shaming’. My claim is not that we should entirely exclude ERO norms. We should hedge our bets. But, I suggest, we should hedge them in favour of AFF norms over ERO norms.

Both ERO and AFF norms target mitigation at the expense of adaptation and compensation. Is it defensible to focus just on mitigation-targeting norms? ¹³ I believe so. First, there are some climate harms – such as the uninhabitability of certain regions – which may be impossible to adapt to or compensate. For these, mitigation is necessary. Second, the promulgation of mitigation norms is likely to have the downstream effect of making climate harms more widely acknowledged as such. This would clear the way for the promotion of adaptation and compensation norms. So, it is worth considering which mitigation norms we should promote, without denying that adaptation and compensation norms will also be necessary. And it is worth focusing on ERO and AFF norms specifically, since these are two popular mitigation norms.

Let's begin by specifying the two norm-types. ERO norms include veganism, anti-driving, anti-flying, pro-cycling, pro-solar and general anti-consumption norms. Using Bicchieri's formulation, a group with ERO norms is one in which individuals engage in personal emissions reductions and offsetting, given their expectation that (a) most people in the group engage in personal emissions reduction and offsetting (empirical expectation), and (b) most people in the group believe those in the group ought to engage in personal emissions reduction and offsetting (normative expectation).

A group with AFF norms is one in which individuals challenge organisations that extract, burn, fund and invest in fossil fuels, given the individuals’ expectation that (a) most people in the group challenge organisations that extract, burn, fund and invest in fossil fuels (empirical expectation), and (b) most people in the group believe those in the group ought to challenge organisations that extract, burn, fund and invest in fossil fuels (normative expectation). That is, AFF norms hold amongst individuals, not amongst organisations. The norm demands individuals to challenge organisations (in ways I will discuss in Section 5). ‘Organisations’ here includes not just for-profit corporations, but also states, universities, pension funds, churches, charities and any other large-scale collective agent that engages in the relevant activities.

Which of these two norm-types should receive the lion's share of our time, money and energy? In answering this question, I will tentatively adopt Aaron Maltais’ three criteria for individuals’ climatic obligations:

our obligations under current conditions should entail only low costs because we have a realistic expectation that effective collective action will not occur … Personal obligations should also be implementable and potentially effective within the near term in response to the limited timeframe for an environmentally effective response. (2013, 602, emphasis added)

I adopt these criteria because they take seriously the limits on individuals’ time, money and energy – which is a presumption behind the need to balance our efforts between ERO and AFF norms. ¹⁴

First, the promotion of AFF norms has low willpower costs proportional to its benefits. ERO norms require lowering one's own emissions directly, and/or encouraging others to do likewise, alongside spreading the idea that this is what's socially expected of people. This is motivationally difficult and takes willpower – as demonstrated by the huge number of environmentally-concerned people who continue to fly in the face of flight shame, out of perceived necessity (Gössling et al. 2020, 1–2; Oswald and Ernst 2021, 69). Or consider driving. It takes extraordinary willpower to unilaterally stop driving – or to encourage others either to reduce their everyday transport emissions, or to think that others think they should – when the physical infrastructure is not set up for pedestrians and bicycles. By contrast, promoting AFF norms means making demands of organisations. To be sure, individuals might ultimately bear the costs of organisational reform (because organisations might ‘pass on’ the costs to shareholders or customers). But because these costs are not self-imposed, AFF norms do not require the exertion of willpower that ERO norms do. (On the importance of willpower for moral over-demandingess, see Chappell 2019, esp. 257.)

Second, the promotion of AFF norms is implementable. Individuals should refuse engagement with organisations that extract, burn, fund and invest in fossil fuels – insofar as the individual can, given the low-cost criterion mentioned above. And individuals should publicly condemn such organisations (where appropriate, given the possibility of backlash against such condemnation). Practicably, this means changing one's energy provider, financial services provider, pension fund and so on – as well as working for change within organisations where one can. I say more about implementation in Section 5. Notice that these actions do not necessarily – perhaps even do not usually – involve reducing one's emissions directly, in the form of flying less, driving less, eating less meat and so on; nor does this require convincing others to emit less.

Third and finally, AFF norms have the potential to be effective over the short-term, since they do not require anything from the slow-moving machinery of governments or legislatures. Indeed, any given act of AFF norm promotion has a higher chance of having at least some minor impact, than does any individual vote for a green candidate (Maltais’ preferred action (2013, 602)). This is because the effects of voting rely on the achievement of the threshold necessary to elect the relevant candidate – whereas pressure on organisations is a more scalar phenomenon, where the extent of achievement is closer to a scalar function of the amount of pressure imposed.

In addition to satisfying Maltais’ three criteria, AFF norms have other advantages. One is that AFF norms are underpinned by an existing moral norm with wide currency: the norm against doing large-scale harm. Due to moral norms’ relative resistance to compromise, the promotion of social norms is more effective when the promoted social norm appeals to an existing moral norm (Bicchieri 2017, 118). The specific moral norm against doing large-scale harm has more climate-related applicability to organisations than individuals, because organisations do more climatic harm than individuals. In this way, the promotion of AFF norms places proportionate costs on individuals and organisations: individuals must bear the (relatively) low costs involved in promoting the norm; organisations must bear the (relatively) high cost of reducing their engagement with fossil fuels. This mirrors the extent to which the two types of agent do harm.

By framing AFF norms within the moral norm against large-scale harm, norm-promoters can ‘piggyback’ on some virtues of the latter norm. For example, as Fergus Green has argued (2018, 104), AFF norms are attractively satiable: there is a measurable point at which one's conduct satisfies them. This makes AFF norms comparatively intuitive and easy to assess for violations, as compared with norms to simply ‘reduce one's personal emissions’. Green argues that this is one reason why AFF norms are already having some success amongst large-scale international actors (though he doesn’t consider the connection to individuals’ climatic obligations, which is my focus) (Green 2018, 113; see also Green and van Asselt forthcoming). As Neil Gunningham puts it (discussing the fossil fuel divestment movement specifically, which is one branch of the AFF norm movement): ‘[t]he global divestment movement … has engaged in an activist campaign involving a demand (divest), a promised reward if the target meets the demand (maintaining reputation, avoiding stranded assets as fossil fuel stocks rapidly devalue), and a threat of harm if the target rejects the demand (shaming and stigmatization)’. (2017, 310, emphasis original) The ‘shaming and stigmatization’ works because the new social norms piggyback on an already-existing moral norm.

Because AFF norms draw upon the moral norm against doing large-scale harm, AFF norms have the potential to overcome the psychological difficulties that make ERO norms comparatively access-infeasible. As one study summarised the literature, ‘[c]onverging evidence from the behavioural and brain sciences suggests that the human ethical judgement system is not well equipped to identify climate change – a complex, large-scale and unintentionally caused phenomenon – as an important ethical imperative’ (Markowitz and Shariff 2012, 243). That is, personal emissions can often seem disconnected from the harms of climate change. Fortunately for would-be advocates of AFF norms, organisations’ extraction and burning (and, to a lesser extent, funding and investment) of fossil fuels does fairly immediate and perceptible harm to local communities: more so than the emissions of a private individual. This makes AFF norms easier to promote, because more directly connected to identifiable harms.

Indeed, Julie Ayling explains that concrete and immediate climate harms have already been the focus of campaigns to prevent a large new coal mine in Australia: ‘the risk that coal mining poses to the Great Barrier Reef is a significant rallying point for the movement’ (Ayling 2017, 358). More generally, the Australian fossil fuel divestment movement focuses on isolatable harms by identifiable organisations (Ayling 2017, 363). By advocating for AFF in the context of specific large-scale harms that are relevant to particular local contexts, the seeming distance, diffuseness and unintentionality of climate harms are no roadblock to the promotion of pro-climate social norms. Yet ERO norms have difficulty overcoming that roadblock.

However, there's an important objection to the AFF norms proposal. Above, I presented the relative non-demandingness of AFF norm-promotion as a benefit of AFF norm-promotion over ERO norm-promotion. This might make one suspicious. Following Stephen Gardiner's discussion of moral corruption in the context of climate change (2011, ch. 9), the worry is: given our strong temptation to do nothing about climate change, ‘we are likely to be attracted to weak or deceptive arguments that appear on the surface to license such [selfish] behavior, and so to give such arguments less scrutiny than we ought’. (2011, 302) Given AFF norms’ relative low demands on individuals, is my argument for AFF norms an apology for relative inaction from individuals?

In response, my argument is an exercise in non-ideal theory. My argument addresses the world as we find it, in which even many highly environmentally conscious people consider flying a necessary evil in their lives. The argument addresses people who are motivated to act, but who are not moral saints and whose social networks contain many unmotivated people. To tell these people to focus on promoting ERO norms is to fail to engage with human motivation and interaction as we find it. ¹⁵

5. By what means?

As Fergus Green and Eric Brandstedt note, the recommendations of climate theorists are ‘typically issued to the citizenry (or human being, or states) in general, with little to no attention given to the role of particular agents of change and their epistemic standpoint, their position within power structures, or their unique constraints and opportunities’. (2020, 5, emphasis original) My argument so far has adopted this level of generality: I have discussed motivated affluent individuals in high-consumption countries, at large and in the abstract.

In this section, I aim to get more concrete. However, given this paper's purely theoretical remit, I do not report on engaged ethnographic fieldwork (as Green and Brandstedt advocate). Instead, I limit myself to outlining two broad pathways to the promotion of AFF norms: an ‘internal’ route and an ‘external’ route. Obviously, this two-way division subsumes many important distinctions. Further specification of these pathways, for different actors in different contexts, will be a task for future ethnographically engaged research.

In the ‘internal’ route, the individual conceptualises themselves as a constitutive part of an organisation that faces demands under AFF norms. If we consider fossil fuel investors such as states, universities, pension funds and churches, then this first route is open to individuals as citizens, employees, members and congregants. This route constrains the individual to using intra-organisationally authorised channels to induce change. This route has been seen prominently within universities, where faculty and students have used their roles in university committees, senates and advisory boards to push AFF norms amongst university stakeholders, ultimately pushing universities towards divestment from fossil fuels (Bratman et al. 2016).

The internal route has several advantages. First, it enables individuals to argue that the organisation's own values entail AFF norms. Such an argument will likely be most effective if made from the inside. If an internal argument can be made that organisational values support AFF norms, then this increases pressure on other members to also endorse AFF norms within the organisation, which can lead to a ‘norm cascade’ within the organisation, as I will explain below.

Second, if the organisation funds or invests in fossil fuels (rather than extracts and burns them), the organisation has ‘plausibly deniability’: the organisation is economically and physically distant from the harm, so it can claim that it simply didn’t see its own complicity. As disingenuous as such a claim may be, it enables the organisation to ‘save face’ while redirecting operations. Now, an agent's obligations are not usually conditional on that agent's ability to ‘save face’ while performing those obligations. ¹⁶ But I am asking about members’ obligations to effect change within organisations to which they belong. ¹⁷ While the content of my obligations are not constrained by my pride, the nature of my obligations to influence others may well be constrained by their pride. I cannot have obligations to influence someone (including an organisation) whose pride forecloses their being influenced at all. The internal route often enables a circumvention of damaged pride.

Third, if the norm promulgation and demand for change occurs entirely ‘in-house,’ then this saves the organisation bad publicity, making the change less costly for the organisation. Again, this increases the organisation's pliability, which makes the internal route an effective way for individuals to effect change. (Though this source of effectiveness must be balanced against the fact that ‘open’ change may be more effective at inducing other organisations to do likewise.)

The ‘external’ route involves individuals acting externally to organisations. Strategies here include consumer boycotts and ‘naming and shaming.’ The goal is to change other individuals’ beliefs about (a) whether most people challenge organisations that extract, burn, fund and invest in fossil fuels and (b) whether most people believe one ought to challenge organisations that extract, burn, fund and invest in fossil fuels. It is thus essential that the boycotts are publicised. This route has an advantage in being more widely accessible: anyone with an internet connection can shame a company on Twitter. This route also unifies the individuals who are promoting and following the norm, by creating a common pathway and a common adversary. The internal route lacks either of these advantages, insofar as different internal routes are particular to different individuals, and require collaboration with (rather than opposition to) organisations.

Cross-cutting the distinction between internal and external routes, there are three vehicles (mechanisms) for traversing each route (Brennan et al. 2013, 95–101). Although the vehicles are route-neutral in theory, some vehicles are more suited to the internal route, others the external route.

First, there is a two-step process in which the parties (that is, the individuals promoting AFF norms) ‘agree to agree’ on what they will demand from organisations, and then form a substantive agreement on what those demands will be (Bicchieri 2017, 115–118). This vehicle will often involve explicit persuasion, debate and discussion. This is most applicable to the internal route, as when a university's senate debates, and then agrees upon, the standards they will hold the university to regarding fossil fuel investments. But this vehicle also works on the external route: consider a union meeting or a boycott organising meeting, where members might have agreed to agree upon a precise set of norms for what (and how) they will act upon organisations.

The second vehicle is a free-flowing ‘norm cascade’ (Finnemore and Sikkink 1998, 902–904). Here, there is often a high cost for the initial trail-blazer, with the costs of compliance with the new norm gradually decreasing for each new adopter, until there comes to be a cost to non-compliance, and eventually there is a high cost to straggling on norm adoption (Schelling 1978, 91–110). The costs and incentives for occupying various temporal positions in the cascade are different for different individuals – some will find it more worthwhile to act earlier, while others will find it more worthwhile to act later. This vehicle applies well to both the internal and external routes. On the internal route, there will be costs (to one's intra-organisational reputation) in being either a trail-blazer or a straggler. Externally, the same is true – with trail-blazers running the risk of rejection and stragglers risking being labelled ‘on the wrong side of history’.

The third vehicle is a ‘follow-the-leader’ elite-initiated mechanism for norm-adoption. While the cascade vehicle relies on late adopters changing their behaviour because most others have, the follow-the-leader vehicle relies on late adopters changing their behaviour because high-status others have. This vehicle is less action-guiding for individuals promoting AFF norms, since most individuals are not elite. But all individuals have the capacity to either follow elites or not, making this vehicle an important one for individuals to consider: if one sees a celebrity promoting AFF norms, then there's good reason to follow them, since there's a good chance others are following them, too. After a follow-the-leader process has taken root, the costs of promoting the norm are lowered further, since people will soon follow the norm out of habit. Like the cascade vehicle, the follow-the-leader vehicle can run along both the internal route (where those with high seniority or status, and low stakes, within the organisation have stronger obligations to ‘lead’) and the external route (where those with high societal standing and influence have stronger obligations to ‘lead’).

6. Conclusion

Much more could be said about individuals’ climatic obligations. In particular, it would be worth investigating the extent to which some actions that my argument questioned – such as voting, petitioning, protesting, resisting and even perhaps offsetting – might contribute to the promotion of climate-friendly norms in general, and AFF norms specifically. Successful norm-promotion largely depends on how one's actions are perceived by others in the relevant group, such that those others change their empirical and normative expectations. For this reason, it's difficult to say anything general about the prospects of these actions for contributing to norm-promotion. There are also remaining questions about the relative merits of different climate-friendly norms, particularly norms that focus on adaptation or compensation rather than just mitigation.

These remaining questions notwithstanding, this article has aimed to advance the debate about individuals’ climatic obligations. I have provided an argument for why social norms are a particularly valuable target for individual climate action. This argument is an addition to others’ brief endorsements of norm-promotion or widespread social change more generally. Within norm-promotion, I have made the case for hedging our bets in favour of anti-fossil fuel norms. Perhaps most importantly, I hope to have brought prominence to a third way of thinking about individual climate action: rather than addressing policymakers (via voting, petitioning, protesting or resisting), and rather than addressing one's own ‘clean hands’ (via offsetting or private reductions in consumption), an individual's climate action can usefully target the informal tissue that binds policymakers and individuals: social norms.