The political theory of technological change: Lessons from the liberalism-ecologism debate

Abstract

Little consideration has been given to the process of technological change in political theory. Given that ideas about this process play an important role in many strands of normative political thought, and are especially crucial to climate change politics, this is a remarkable oversight. It risks political theory being irrelevant to climate change mitigation. The implications of this oversight for political theory are explored here through an analysis of the liberalism-ecologism debate. The article argues that attempts to green liberalism – to move it beyond environmentalism – cannot succeed while liberalism is silent about technological change. More broadly, given that most political theory traditions make claims about technological change, claims crucial to their worldviews and normative goals, it argues that much more theorisation of the concept is necessary. Especially now that they shape how the world understands climate change mitigation, contests over the meaning of technological change are intensely political contests. Political theory needs to get much more involved.

Introduction

Technological change is a mute presence in political theory, important but mostly unnoticed. In an era of environmental crisis, when expectations about future technologies are driving the global response, the time is ripe to start paying attention. After Adam Smith, liberalism in particular has ignored technological change. Liberals consider neither the process itself, nor prevalent ideas about it, politically important. This is an error. Its absence thwarts their efforts to adapt liberalism to meet the threat of environmental disaster.

Other political theories should take note. Though not as deficient in this regard as liberalism, few possess detailed treatments of technological change. What are they missing? Technological change is the process by which, depending on your perspective, technologies and technological infrastructures transform, improve, develop, and/or grow more sophisticated and pervasive. As a concept, it is broad, taking in major transformations, like the spread of electric lighting in the late 19th century, as well as small developments, like improvements in LED light bulb efficiency in the early 21st century. Related concepts such as innovation, invention, technology, technics, and technique have all been given various meanings over the years (Ellul, 1964; Heidegger, 1977; Keary, 2016; Mumford, 1967). Contests over these meanings, as discussed in the final section of this article, are contests for power.

Still, serious consideration of technological change will raise two salient questions: what is it doing? And how does it work? The first of these questions motivated earlier critical discussions of technology (Ellul, 1964; Marcuse, 2002; Postman, 1992; Stiegler, 1998; Winner, 1977). These debates explored how technologies alter the human experience, but also considered technological trajectories and the impacts of change itself, such as feelings of dislocation. I ignore this aspect of the concept.

The second question – how does technological change work? – motivates this article. It has two streams. First, is it outside human control (exogenous)? Second, if it is controllable (endogenous), how controllable is it? In other words, can the pace and direction of technological change fall under the will of a political programme? The idea that it can is called ‘directed technological change’ (Acemoglu, 2015).

Some conservatives have treated technological change as under human control, others not, though they agree that there is little of merit to be controlled (Gray, 2007; Oakeshott, 2014). Other political theories tend to treat technological change as subject to human agency; indeed, this conceptualisation has sparked much utopian thinking (Bastani, 2019; Bookchin, 1971; Feenberg, 1990; Firestone, 2015).

The exception is liberalism. Adam Smith (1993: 8) integrated an endogenous model of technological change into Wealth of Nations. It played a crucial role in supporting the ‘universal opulence’ he saw as the inevitable result of his politico-economic programme. However, liberals since Smith have hardly touched the subject. Only Mill (1965) has written substantially on technological change and then only in his economics. This division of liberalism into economic and political domains is likely the reason why liberalism, alone of political theories, is silent about technological change. Considered economics, it has been walled off by political liberals who, again alone of political theorists, respect the artificial barrier between economics and politics.

The developing environmental crisis has knocked holes in this wall. Liberal political theorists have been forced to consider the economic dimension of the problem, and with it, the factor of technological change. This move in political theory is called ‘green liberalism’. Proposed by Wissenburg (1998), Bell (2002), Hailwood (2004) and others, green liberalism claims that liberal political theory can accommodate all that is of merit in ecologism. Scholars of ecologism counter that liberalism is incompatible with non-anthropocentrism, limits to growth, and other ideas crucial to a stable and equitable environmental politics (Dobson, 2007; Eckersley, 2004). The debate quietened around 2008, and though recently revived by Welburn (2017), one might ask: what more needs to be said?

A good deal about technological change, for a start. The liberal understanding of how it operates was never made clear in this debate. This matters because it undermines green liberalism’s pursuit of its environmental aims. As a mark of these aims, several green liberal texts reference this Dobson quote: ‘environmentalism and liberalism are compatible, but ecologism and liberalism are not’ (Dobson, 2007: 150; Stephens, 2016: 60; Welburn, 2017: 8). It is taken as a test: green liberalism is successful if it makes clear inroads into ecologistic territory. One signpost for ecologistic territory is the rejection of economic growth. Antipathy to growth is central to ecologism, whereas reconciling growth and the environment is the purpose of environmentalism. All green liberals claim that they jettison economic growth. However, when technological change is properly accounted for, liberalism must withdraw into environmentalism.

Inasmuch as political theory counts in the policy world, this debate matters. The world is presently choosing a gamble on technology over a commitment to environmental stability (Hickel and Kallis, 2019; Keary, 2016; Pielke et al., 2008; Victor, 2012a). To choose this gamble is to wager the future habitability of the earth on the prospect of rapid technological change reconciling economic growth with environmental stability. The 2018 Nobel Prize in Economics was awarded to this policy’s originators, signalling mainstream approval (Strauss, 2018). If political theorists oppose this gamble, it must be clear whether liberalism is a tenable standpoint from which to do so. This article will show that it is not. Only ecologism is clearly tenable; other political theories need first to develop their treatments of technological change. Finally, as technological change moves to the centre of global climate change policy, the world’s biggest challenge, political theory, which has long ignored technological change, ought to give thought to its political importance. A debate on the best form of green politics is an ideal place to begin.

Outline

This article has two main themes, one narrow and one broad. The narrow theme, which occupies most of the article, is the significance of technological change to the debate over the possibility of green liberalism. This feeds into the broader theme: the limited consideration of technological change within political theory and the consequences of this deficiency, particularly given the environmental crisis.

We begin with an introduction to the political theory analysis of Michael Freeden (1996). Given how hidden technological change is, it helps to employ a concept-focused framework like Freeden’s, which demonstrates the function of each idea within a political theory. The subsequent section introduces Rawls’ contribution to green political theory. Rawls has occupied both supporters and detractors of green liberalism, so his work is unavoidable. Then the contribution of Bell, who grounds his work in Rawls, is shown to offer only cursory environmental protection. We finish our analysis of Rawlsian work by considering recent efforts, especially Welburn’s (2017), to apply Rawls’ ideas to the political questions raised by climate change. We find, though, that this literature does not engage with the questions of mitigation policy generation that concern us. Wissenburg is discussed next. The best-reasoned contribution on either side, still Wissenburg’s green liberalism fails to sufficiently consider technological change. The Millian, perfectionist liberalism of Stephens is similarly deficient, despite being critical of the neutralism of the others. Ultimately, we see that liberalism’s internal dynamics require it to retain economic growth, and it thus fails by its own measure of success.

Finally, we reflect briefly on how differing contestations of technological change enable differing normative visions of society and how this can impact on the distribution of power in theory and in practice. If political theory desires greater relevance in the debate over the environmental crisis, it should pay much more attention to ideas about technological change.

Freeden and concept-level analysis

A scholar of ideology and political theory, Freeden’s (1996: 4) contribution of interest here is his identification of the concept as the key unit of analysis for political theory. He argues that the main concepts within the discipline – power, freedom, equality, and so on – are prime examples of ‘essentially contested concepts’ (Gallie, 1955). We need not discuss this notion at length; for our purposes, it means that scholars disagree about which elements make up these concepts and the relative value of each element.

For Freeden (2005: 121), the key to generating political theories is the ‘decontestation’ of concepts: allocating precise meanings to indeterminate concepts. For example, a political theory might decontest equality as referring to outcomes rather than opportunities. This affects not just the concept itself, but surrounding concepts because, as Freeden (1996: 4) recognises, much of the meaning of any concept in a theory comes from its ‘particular structural position within a configuration of other political concepts’. The relationships between the various concepts in a political theory contribute to the meaning of each. The decontestation of one concept impacts upon the decontestations of others.

Of course, within a political theory not all concepts are created equal. Freeden (1996: 77) identifies three types: core, adjacent and peripheral. Adjacent concepts pull the core concept towards a particular decontestation out of perhaps several realistic options: for example, liberty is at the core of all liberalisms, but when democracy is adjacent, liberty comes to mean self-determination (Freeden, 1996: 78). Add more concepts and each decontestation can change, so that a specific colour is given to the whole theory. As we shall see, ‘the environment’ moves into adjacent status within the version of liberalism known as green liberalism.

In the same theory, technological change is a peripheral concept, but of an important variety called ‘perimeter’ (Freeden, 1996: 80). Perimeter concepts are time- and place-specific concepts, which force themselves into political theories. Unlike more abstract core and adjacent concepts, they are concrete, and they alter the decontestations of the core and adjacent concepts, giving them the precision needed for real-world relevance. They are peripheral because times change and, while they are on the periphery (which is not fixed – movement from periphery to adjacency occurs), the ‘total ideological morphology’ can always move on without them (Freeden, 1996: 80). Finally, perimeter concepts can be ‘applications of more general concepts to specific practices’ (Freeden, 1996: 80). Technological change, as we will see, is ultimately decontested within green liberalism as technological optimism.

What Freeden offers is a strategy of political theory analysis. Examining political theories at the concept level allows for clearer comparison across otherwise very different formulations. It also allows one, as we see below, to identify previously unrecognised significance in the relationships between concepts.

Rawlsian environmentalism

Though Rawls wrote little about the environment, a literature has developed that applies his work to the environmental crisis. It mostly discusses how the abstract citizens of a Rawlsian state might endorse green public policy (Achterberg, 1993; Bell, 2002; Welburn, 2017). However, the full picture of decision-making in a Rawlsian liberal democracy has never been captured; important features have been missed, misinterpreted or accorded insufficient weight. So, the conclusion a Rawlsian polity might reach remains unclear. Below I determine the appropriate procedure for this debate and then show how it results in an environmentalist conclusion when technological change is considered.

This necessitates the proposal of a question. Let us ask: if both sustainable development (SD) (Intergovernmental Panel on Climate Change (IPCC), 2014: 115), and a degrowth alternative (Victor, 2012b), were put to Rawlsian citizens as solutions to the environmental crisis, which would they choose? Detailed definitions of these options are not necessary. We shall see why when we examine the decision-making. What matters are their differences on economic growth. The economic growth we have in mind here is the orthodox economics version, favoured by policymakers in capitalist states. It is a purely quantitative measure of total economic output, but it is argued to correlate with welfare gains (Nordhaus, 2007; Stiglitz, 2019). It certainly correlates strongly with increases in material usage, consumption, and greenhouse gas emissions (Kallis, 2011). Option one sees this growth as necessary and claims technological change will reconcile its extraction-heavy and polluting tendencies with environmental stability. The predominant approach, it is favoured by governments, international organisations, and the IPCC (Hickel and Kallis, 2019; Keary, 2016; Pielke et al., 2008). This is environmentalism (Dobson, 2007). Option two favours degrowth and advocates environmental stabilisation using existing technologies and alterations to our lifestyles and means of production (Victor, 2012b). This is ecologism. With these options, we can extrapolate how green is Rawlsianism.

Drawing primarily on Political Liberalism, the key lies in Rawls’ strictures regarding political decision-making, especially his emphasis on mutual respect and compromise, and his stipulations regarding public enquiry and argument. However, for the question to arise, we must make assumptions, all designed to help Rawlsianism be ecologistic. First, we assume it is 2023, where knowledge about the environmental crisis is considerable. We also assume that two similarly sized groups exist, one advocating SD, the other degrowth. This is generous; typically, the former is far stronger. Most people, though, are undecided. Also, in this ideal political community, there is no denialism and debate is free and fair. So, how does Rawls’ (2005: 448) ‘deliberative democracy’ discuss our question?

Two forms of discussion exist in the Rawlsian state: public reason and arguments from one’s own comprehensive doctrine. ¹ The former is for constitutional essentials and basic justice. The latter is for everything else. Let us unpack public reason by considering the purpose it serves. ² Public reason arises out of pluralism. How is a well-ordered society possible when its citizens hold mutually incompatible beliefs? Upon what basis can laws and institutions, to which everyone can give their adherence, be built? How does one citizen persuade another? Rawls argues that in all reasonable comprehensive doctrines, and in the political culture of democratic societies, there are values in which all citizens set some store. He calls these political values. Those affecting freedom and opportunity are described in general terms – Rawls’ constitutional essentials – so interpreting them requires acceptable justifications with which to settle disagreements. Issues of basic justice require this also. In Rawls’ (2005: 441) words, ‘citizens . . . need to consider what kind of reasons they may reasonably give one another when fundamental questions are at stake’.

These political values are the building blocks of ‘a political conception of justice’ (Rawls, 2005: 450). One political conception of justice is ‘justice as fairness’ – which Rawls (1973) espoused in A Theory of Justice – but others are possible. They differ in interpretation of basic freedoms and in their ordering of political values. For example, some value artistic liberty more than public decency, but others may not. The key is that the political values be complete. Any ordering of them must provide an answer to questions involving constitutional essentials or basic justice (Rawls, 2005: 454). The basis for disagreement in a political community thus founded will be the ordering and interpretation of these political values. Political values give the content to public reason and Rawls provides instructions about debating this content. We will show that together they lead inexorably to SD and its gamble on technological change.

Now we can examine how our question would be answered in a Rawlsian political community. It does not matter which political conception is active. The basis for debate ensures that all will reach the same conclusion. We may not get unanimity, but Rawls does not require it. If citizens follow the rules of public reason, disputes can legitimately end with a vote (Rawls, 2005: liii). For this question, the vote would turn on whether the SD option – the technological gamble – is low risk.

What political values are relevant to the environmental crisis? Rawls (2005: 245) provides a list: the general welfare of ourselves and future generations; human health (a diverse nature helps medical and biological research); our general recreation and the pleasure that comes from understanding the world. This list was not meant to be exhaustive, and we can add at least: equality of opportunity (which future storms and forest fires may curb), domestic tranquillity (which climate change jeopardises), free choice of lifestyle (which a degrowth option will likely infringe), and just treatment of other societies (who will suffer if we desert our mitigation responsibilities).

Does the precise ordering of the political values in play matter? Well, the political values of an explicitly anthropocentric political conception exclude most of the arguments that persuade people towards degrowth and environmental concern. These would be values belonging only to the comprehensive doctrines of dark green citizens: the moral standing of nature, human communion with nature, sustainable living, and so on. One could argue that if recreation and understanding of nature, and perhaps equality of opportunity, were voted top of the list, then the guarantee of climate stabilisation, which is the degrowth option’s strongest remaining card, might attract enough votes. Certainly, under a degrowth scenario, natural resource extraction and wasteful agricultural production would cease. At the least, much of the natural world would be spared ‘development’, hence becoming available for recreation. Comparably, a degrowth world, with fair redistribution of wealth, offers a guarantee of equality of opportunity that no technological change option can match. No mainstream scenario puts the probability of successful mitigation near 100% (Van Vuuren et al., 2008). We must, however, even in an idealised scenario, be realistic. No known society would put recreation at the top of the list recited above. And considering the other values in the balance, even if a particular society ranks them all below equality of opportunity, they would collectively be viewed as more important, especially when the loss of equality of opportunity under a climate chaos scenario is a risk, not a certainty.

I must pause here to consider Bell’s work on this matter. A fuller account comes below, but his arguments about public reason fit better here. Bell (2002: 708) maintains that a Rawlsian response to climate change, when drawn from public reason, would require neither growth nor future technological changes. His argument comes from Rawls’ ‘circumstances of justice’ idea: the uncontroversial notion that a just society can only be built once everyone has ‘a decent standard of life’ (In Bell, 2002: 708). To ensure that such circumstances prevail in the future, we must ensure that climate chaos does not preclude them. Furthermore, we must ignore ‘Promethean’ technological optimism, given Rawls’ stricture that controversial scientific work has no place in deliberations. Finally, Bell (2002: 709) argues that a Rawlsian response is not committed to economic growth since Rawls considers growth inessential to a well-ordered society. Let us consider these points as we continue, starting with growth.

For most of the political values discussed above, it can be argued that degrowth would have a harmful effect resembling that of climate chaos. For instance, though under a growth scenario species-loss will continue unabated, with negative consequences for medical research, growth advocates can claim that funding for medical research will plummet without economic growth. Similarly, if economic growth helps solve problems like weak health infrastructure, inequalities in education, inadequate housing, insufficient foreign aid and so on – or prevents them becoming problems – then the political value of the general welfare is captured by the SD option (indeed, this argument is at the heart of SD). So, while Bell is right that Rawls considered growth inessential, there are strong reasons, based on political values, for maintaining growth, arguments that economists have been making (Nordhaus, 2007; Stiglitz, 2019). Thus, if the SD option is low risk, the balance of political values will go against degrowth.

Whether it can be considered low-risk depends on how it is investigated. By Rawls’ strictures, however, the result is a foregone conclusion. At the heart of SD, as envisaged by the IPCC, is a gamble on future technological changes (Keary, 2016; Pielke et al., 2008). If these arise as predicted, both growth and climate change mitigation can be reconciled. This makes the plan sound rather controversial though and Rawls (2005: 139) insists that for the purposes of policy, public reason only accords scientific output weight when it is non-controversial. Perhaps both options are too controversial then? Much of the SD case rests on macroeconomic modelling, which for Rawls (2005: 229) exemplifies controversial science. Must a Rawlsian stalemate thus ensue?

No, the degrowth option is much more controversial. After all, the IPCC draws on a galaxy of prestigious economists and engineers and has been endorsed by nearly every state. Moreover, its technological change modelling has the approval of the Nobel Prize Committee (Strauss, 2018). Derivative policies, such as the Green New Deal or the EU’s European Green Deal, are endorsed by most politicians who are not active climate change deniers (European Commission, 2019; Hickel and Kallis, 2019). Bell, writing in 2002, could not know that Prometheanism was going mainstream. The degrowth option has substantial academic support (Daly, 2014; Kallis et al., 2018; Victor, 2012b), and much credibility among greens, but has far fewer supporters and much less prestige. It also makes policy-prescriptions that contradict the long-standing raison d’etre of governments: economic growth. Degrowth must be counter-intuitive to most people. The idea that growth is good probably meets Rawls’ (2005: 225) definition of common-sense: ‘plain truths now widely accepted, or available, to citizens generally’. This lends considerable weight to the SD option in Rawlsian deliberations.

Finally, stalemate is not an option. Not making a decision is worse than either policy and besides, any idealised political community that cannot decide how to meet the environmental crisis is surely not worthy of the name. Moreover, the cognitive powers Rawls (2005: 83) grants his citizens provide clear guidance: his citizens’ rationality requires them to choose ‘the more probable’ alternative. Given the controversy factor, there is little doubt which choice they will make.

Using our Freedenian analytic lens, then, within Rawls’ liberalism, the environment – decontested as a store of potential utility for humans, one increasingly in jeopardy – occupies only a peripheral position, and not the perimeter version discussed above, but a marginal position, meaning its importance to the core is ‘intellectually and emotionally insubstantial’ (Freeden, 1996: 78). Environmental problems are minor considerations for Rawls. The essential thing is how this case demonstrates the risk for political theorists of adopting concepts from the margin, especially when pressed by contemporary events: insufficiently considered decontestations can introduce additional conceptual baggage into models. Here, technological change, through its links with the environment-as-concept, enters the model undecontested. ³ Technological change must eventually be decontested alongside any decontestation of the environment because, if decontested as controllable, it causes the environment to be decontested as a fixable glitch in the political system, condemning it to marginal status. If technological change is decontested as, say, unpredictable, though, the environment must become adjacent, since it alters how the core concept is decontested: liberty, for one, must be circumscribed when lifestyles are environmentally destructive; equality, for another, must be less utopian.

Rawls’ failure to decontest technological change must result, as we have seen, in the adoption within the model of economic theory’s decontestation, the fulcrum of the IPCC’s projections. The pace and direction of technological change is controllable, and the environment thus moves to the margin. Many Rawlsian liberals will be content with this outcome. Environmentalism – the marrying of the market and environmental concerns – may seem reasonable. But for green liberals, like Bell, market environmentalism is what they want to get past. As a result of a failure to decontest technological change, they are dragged back to it. Let us now see how.

Bell

We begin with the remainder of Bell’s work. His public reason arguments we dealt with above, but Bell focuses on discussing what might be achieved beyond the limits of public reason. In Rawls’ framework, most of the humdrum legislation passed by parliaments does not require public reason. People can ‘vote their more comprehensive views’ (Rawls, 2005: 235). Any serious green legislation will touch basic justice and thus public reason. Still, let us assume Bell is right and see if this route can reconcile liberalism with environmental justice and ecologism. We shall find that Rawls’ liberalism stifles green political action.

First, as discussed, the claim that Rawls is not committed to economic growth is true, but unintentionally misleading. This is important because no variation of ecologism supports economic growth. Therefore, since degrowth would lose a public reason debate in all but the most affluent societies conceivable, liberalism can admit Bell’s green preferences only in principle.

The point about degrowth and mainstream economics deserves more attention. It is symptomatic of a general feature of Rawls’ liberalism: its immunity to radical policy. Even if Bell is correct, and substantial green reforms could be implemented without public reason, no well-functioning, developed Rawlsian political community could agree to them. Rawls (2005: 193) tells us that

It is surely impossible for the basic structure of a just constitutional regime not to have important effects and influences as to which comprehensive doctrines endure and gain adherents over time . . . We must accept the facts of common-sense political sociology.

This is an important point for Rawls. He insists that societies under political liberalism will grow increasingly tolerant and cohesive. The constitutional framework will come to be revered and trusted by citizens (Rawls, 2005: 86, 463). If we presume that Rawls’ basic sociology is correct, non-anthropocentric comprehensive doctrines will struggle to gain adherents. Over time, indeed, their numbers will likely dwindle, as attachment to the anthropocentric basic structure boosts anthropocentric ethical frameworks.

Furthermore, Rawls’ emphasis on reasonableness, compromise and unity makes decisive, radical action highly unlikely. It is difficult to see Rawlsian citizens throwing paint, occupying buildings, and staging mass demonstrations once the almost somnolent forces of Rawls’ state take hold. In a Rawlsian polity, the political conception ‘assumes a wide role as part of public culture’ (Rawls, 2005: 71). Rawls (2005: 71) imagines ‘a social world within which the ideal of citizenship can be learned and may elicit an effective desire to be that kind of person’, the kind of person ‘that [citizens] otherwise would most likely never be able to entertain’ being.

What kind of person is this? A person who is pleased that others hold different views from themselves, who is always tolerant and ‘ready to meet others halfway’ (Rawls, 2005: 157), possessed of the ‘spirit of compromise’ (Rawls, 2005: 163), dedicated to preserving ‘unity and stability’ (Rawls, 2005: 133) and in general ‘the most reasonable conception of the person that the general facts about human nature and society seem to allow’ (Rawls, 2005: 87). Not the kind of person likely to carry a radical and – with reference to the political culture thus fostered – widely regarded as unethical comprehensive doctrine from obscurity to general approval. Almost certainly the kind of person, when faced with approaches to the environmental crisis ranging from classical liberalism to ecologism, to vote for a compromise, especially one promising to satisfy both economic and environmental ends through a unifying expedient. SD, in other words. Hence, with or without public reason, the Rawlsian polity is not likely to choose policies compatible with ecologism.

Rawls and climate change

Let us consider what this analysis means for recent efforts to apply Rawls to climate change. While many scholars deny that Rawlsian ideas can help address the climate crisis (Brincat, 2015; Gardiner, 2011), more find them promising. The latter fall into two broad categories: those defending the fitness of Rawlsian ideal theory – suitably amended or expanded – for a world in climate crisis (Clements and Formosa, 2021; Kim, 2019; Nielsen and Hauge Helgestad, 2022; Welburn, 2017); and those suggesting Rawlsian ideas as frameworks for achieving international climate justice (Bernstein, 2015; Clements, 2015; Huseby, 2013; Kenehan, 2015).

Work in the latter category does not speak to my argument. Rather, it applies to the moment after optimal climate mitigation policy has been agreed, when the question of a just distribution of mitigation and adaptation burdens among peoples must be decided. For the most part, work in the former category also has different concerns. It focuses on making the case that Rawls is applicable to climate change at all, something I do not dispute.

The exception is Welburn, who revitalises the liberalism-ecologism debate by offering another way to read the constituent citizens of a Rawlsian polity. He argues that a conservative spirit of stewardship suffuses Rawls’s political community, reading the well-ordered society as an enduring search for justice over generations (Welburn, 2017). To ensure the continuance of this transgenerational project, the just institutions that underpin it must be preserved.

For Welburn, this preservation can be powered by a conservative spirit of citizen self-sacrifice: if a citizen must give up their own immediate good in the service of this end, they will do so. Welburn grounds this claim, which on the face of it is not very Rawlsian, in modern conservative readings of ‘human nature as it exists’ (125). As ‘stewards of a liberal polity’, then, these citizens will sacrifice their personal good in the interests of the environment because a stable environmental base is necessary for the continuance of the just institutions of the political community (131).

There is no need to dispute the compatibility of an archconservative like Roger Scruton with recent history’s most famous liberal. We can merely observe that Welburn’s reading does nothing to challenge the deliberative sequence outlined above. It is unclear just how self-sacrificial Welburn’s citizens are, but clearly there is a trade-off with self-interest, not the entire abandonment of it. That being the case, even if Welburn’s insights about their nature are correct, they will still choose the gamble of SD through technological change over the perceived gamble of degrowth. The political values remain the same and even if self-interest is a touch less compelling, the rational decision-making process Rawls describes remains unchanged.

Wissenburg

Still, as Rawls hinted, the transgenerational route that Welburn follows is perhaps the best way to a sustainable Rawlsian state. The idea is intuitive: if we agree that we have benefitted from our ancestry, then we might agree to curtail our avaricious instincts. Of the work taking this approach, the most substantial is that of Marcel Wissenburg.

Wissenburg’s (1998) Green Liberalism argues that liberalism can be as green as any political theory can reasonably be. The rub is twofold. One, this is simply what liberalism permits: Wissenburg (1998: 221–222) insists on non-interference in preferred lifestyles; and two, he gives us no reason to expect adherence from actual citizens of liberal democracies (Wissenburg, 1998: 225). However, he maintains that no political theory can go further without accepting authoritarianism (Wissenburg, 1998: 89). Wissenburg’s key contribution is his ‘restraint principle’ (RP), which looks, on the surface, to offer the environment considerable protection:

No goods shall be destroyed unless unavoidable and unless they are replaced by perfectly identical goods; if that is physically impossible, they should be replaced by equivalent goods resembling the original as closely as possible; and if that is also impossible, a proper compensation should be provided (Wissenburg, 1998: 123).

This is not as powerful as it seems. The RP is well satisfied by technological solutions to pollution problems (Wissenburg, 1998: 195). Wissenburg has no use for green shibboleths so his discussion of technology is unusually positive for a green. About technological change, however, he says much less, and his line is more orthodox green.

Let us remember our two questions from the introduction: what does technological change do and how does it work? Wissenburg focuses almost exclusively upon the first of these, which is not important for our concerns. His treatment of the second question is entirely cursory. Still, he states that technological change is ‘undependable (unpredictable) and questionable’ (Wissenburg, 1998: 82), which must mean that he considers it to be exogenous: that the pace and direction of it are not subject to human control.

Despite the limited discussion, then, Wissenburg’s model seems to produce a different outcome to Rawls’. If technological change does not respond in a reliable and timely manner to human need, then to avert the worst effects of climate change, a degrowth pathway, which Wissenburg (1998: 200) does not rule out, must be chosen under the RP. This conclusion is not warranted though. Wissenburg’s statement on technological change, made in 1998 when mitigation modelling was in its infancy, is an empirical, rather than a principled, statement. Given Wissenburg’s (2001: 39) defence of mainstream economics, it would be logical for him to revise his statement in line with that discipline. Certainly, there is nothing in his model that requires technological scepticism. Wissenburg is even enthusiastic about applying existing technologies.

Hence, we must assess Green Liberalism’s treatment of risk and uncertainty. We have noted that both sides are gambling: one on the dependability of technological change, the other on an untested economic system. Wissenburg insists that any gamble must accord with the RP. Since they attempt to avoid destruction, both clearly are. Beyond this, Wissenburg declines to lay out principles of risk/uncertainty. He argues (Wissenburg, 1998: 181–87), in line with liberal principles, that risk tolerance is a personal preference and thus liberalism can demand only a broad-minded cost–benefit analysis. As we have seen, in Rawls’ idealised liberal society the preferred gamble is a foregone conclusion. In Wissenburg’s world, however, not only are constitutional essentials and basic justice sacrosanct, but potentially any aspect of a person’s lifestyle. There is, therefore, even less doubt.

If, as Wissenburg claims, Green Liberalism protects nature as well as any other theory reasonably can, then ecologism is a theoretical dead-end. The latter stands for the subjectivity of the natural, the maintenance of environmental integrity, an end to economic growth, a novel and interdependent relationship between humanity and nature, and many other interrelated values that Green Liberalism neglects (Freeden, 1996: 526–54). While the exact effect of the RP within liberalism is hard to determine, the approach to climate change will be environmentalism and not ecologism. Using our Freeden framework, then, let us see how this happens.

Wissenburg, like Rawls, decontests the environment as a store of utility for humans. This is unsurprising given its dependent relationship with the core concept of liberty. However, the environment is a richer concept as Wissenburg decontests it. The material world it refers to is no longer a homogeneous mass, but rather a set of independently valued entities. The concept moves adjacent to liberty, to which it has a new significance: to guarantee liberty for others through the preservation of these natural entities, liberty, decontested as self-determination, is more qualified. As Freeden (1996: 78) observes, concepts ‘have their ineliminable cores . . . [but] are filled out in a distinctive way due to their mutual proximity’. Now, as written by Wissenburg, technological change occupies a marginal position, fitting Freeden’s description of a concept forced onto the agenda by contemporary thought but relating to the theory in question ‘only reluctantly and contingently’ (Freeden, 1996: 70). Wissenburg decontests it as potentially useful but unpredictable and this unpredictability makes it marginal. In other liberal theories, like Rawls’, this decontestation would have the opposite effect, making both technological change and the environment more significant. Wissenburg’s decontestation of the environment, however, means that technological change decontested this way has no further effect: the environment alone has a similar effect on liberty through the heterogeneity and significance of the material reality it represents. At most, this decontestation of technological change slightly loosens how Wissenburg understands the self-determination component within his concept of liberty. Still, the overall thrust of the theory is a constrained liberalism, albeit to an unclear degree.

The problem for Wissenburg is that, while the creator of a model can fix its value elements, the empirical elements are contingent. Wissenburg’s model imagines liberal citizens, and unless they can only be drawn from an idealised 1998, their empirical knowledge is automatically updated. It must be or the theory becomes irrelevant. The model’s 2023 green liberals have knowledge appropriate to that year and, by 2023, the IPCC’s technological change modelling is knowable. Wissenburg, unlike green political theorists from the ecologism tradition, takes no value position on technological change. Rather, he maintained in 1998, based on the facts available to him, that technological change is unreliable. As in Rawls’ theory, liberal citizens (and, as noted, Wissenburg himself) should now embrace technological change. The implications of the theory must therefore change.

To what effect? Technological change, now decontested as controllable, moves from the margin to the perimeter, the position occupied by peripheral concepts that provide real-world relevance to a theory (Freeden, 1996: 79). This vitiates the effect that the environment has on liberty by altering the decontestation of the former. Both the urgency and the severity of the threat to the material elements represented by the environment are greatly lessened, moving liberty closer to its familiar, less qualified decontestation. In other words, the self-determination component of its meaning becomes stronger, the other-regarding element becomes weaker. Overall, the new decontestation of technological change has a considerable de-greening effect on the whole theory, to the extent that it forces one to ask whether it is now green at all.

Stephens, Mill, and perfectionist liberalism

The liberalism discussed above is neutralist: it does not claim to know the good life. For Stephens (2001, 2016) this tradition is incompatible with ecologism because it cannot do more than permit its citizens to become green. Alternatively, there is a perfectionist tradition within liberalism, where the state holds a conception of the good and urges its citizens towards it. JS Mill favoured this approach and Stephens thinks Mill’s liberalism is thus more amenable to green aims (Brink, 2013; Gregory, 2014; Schramme, 2009; Stephens, 2016). Wissenburg (2001) disagrees, and their debate is layered; however, if I can show that Stephens’ green liberalism is also unable to avoid the gamble on technological change, then the remainder of their argument becomes irrelevant to this article.

So, might the hypothetical citizens of a Millian liberal state choose the degrowth option? Stephens (2001, 2011: 14) feels that, because Mill’s liberalism aspires for humans to develop beyond being mere consumers, the potential for citizens to embody ‘non-material values like concern for nature’ is greater. This is true, but it will prove insufficient to prevent them choosing SD.

Mill argues that a better class of happiness can be fashioned through a suspicion of orthodoxies and a relentless push for individuality. The state should facilitate this, helping to create citizens for whom flourishing means self-development, not mere self-enrichment (Mill, 2005: 165). For Stephens (Stephens, 2001: 19), Mill thus provides a route by which ‘the right kind of preferences’ for properly valuing nature can be generated.

There is no reason to believe that Mill’s developed individuals will generate them though. Mill claims all developed individuals recognise poetry as superior to pushpin (McClelland, 1996: 454). However, ecological values have not been embraced by today’s developed individuals; capitalist values still dominate. Ecologism is not more elite than capitalism, just different. Even if somewhat developed, the political climate would create as much dissent and heterodoxy as possible in opposition to ecological preferences. In Mill’s ideal, individuals drive relentlessly for heterodoxy, so much so that McPherson (1982: 255) refers to Mill’s writings a ‘diatribe against conformity’. Hence, we should see environmentalism thrive as a compromise, much as in Rawls’ state.

Also, while Stephens (2001: 19) chides Wissenburg for excluding government promotion of ecological values, the most Stephens (2001: 20) can propose is for citizens to experience more nature, perhaps through greener cities. It is difficult to believe this measure will broadly propagate ecological values and it is surely inadequate for the environmental crisis. Stephens can go no further though, since a Millian state cannot push specific opinions through, say, education. Mill’s conception of development is self-development; prescribing a set of values to be reached would defeat the point.

Moreover, the system’s predilection for individual autonomy will prejudice it in favour of an approach that does not limit action, which is certainly SD. Without any specific values that, say, caution against scientific hubris, or repel wagering the natural world, Mill’s emphasis on human flourishing and individual liberty makes the gamble on technological change inevitable. This is especially the case given how persuasive mainstream academic and policymaking support for environmentalism, which maintains that economic growth is a welfare necessity, would be in a Millian world where rationalism is revered (Freeden, 1996: 149). Mill’s oft-cited support for a stationary state will also be trumped by this fact (Mill, 1965: 756–757; see also Stephens, 2011).

So, the overall morphology of Mill’s liberalism is not substantially changed by Stephens’ addition of the environment. Initially, we have the Millian core concept structure, with individualism decontested as individuality, progress decontested qualitatively as progress in individual character, and thus liberty decontested as free self-development (Freeden, 1996). To this, Stephens adds the environment, which moves adjacent to progress.

Decontested as a source of experiences for human development, as well as resources – albeit threatened – the environment initially seems to affect the decontestation of the three core concepts. As the decontestation of progress shifts to accommodate the significance of human-nature interaction, so the decontestation of individuality is broadened slightly to allow for a less atomised conception of being, and the decontestation of liberty is constrained to ensure the survival of this threatened avenue for progress and individuality.

Yet this adds more urgency to the decontestation of technological change. Like Rawls, Stephens hardly mentions it, so it is decontested by the impetus of the decontestations in the rest of the model. As with Rawls and Wissenburg, this means it is decontested as controllable and its impact comes first on the decontestation of the environment, which becomes more resilient given the prospect of technological fixes. As a result, the constraints the environment places on liberty are much lessened, which is of course the key reason why the impetus to decontest technological change as controllable is so strong in the first place.

Thus, despite the differences between Wissenburg’s and Stephens’ green liberalisms, the inevitable decontestation of technological change as controllable leads to much the same environmental outcomes in both. Though for different reasons, this is the conclusion to which Wissenburg (2001) also comes.

Overall, we can say that in Stephens’ green liberalism, self-development does not equal ecologism-levels of environmental concern; even if it did, it would be a minority opinion and opposed by a range of others, with no clear prospect of becoming law; and even if environmental concern developed en masse, there is no reason to believe it would favour ecologism and degrowth, and every reason to believe it would favour the environmentalist gamble on technological change. Millian perfectionism, then, is not prescriptive enough to accomplish more than neutralism.

Technological change, power, and the environmental crisis

Finally, we should consider the relationship between decontestations of technological change and political power. Introducing technological change into a theory means introducing a fundamental indeterminacy into its picture of material reality. This is the ineliminable core meaning of the concept. Wherever the decontestation lies on the spectrum of human controllability, from a manifestation of human will, to a self-directing threat, it brings dynamism into theories that are, very often, otherwise static. The possibility it carries with it that crucial, inconvenient variables may be refashioned or eliminated, is what makes it so powerful.

For the early liberals, especially the Scots, this power was transformative. Hampered by the conservative charge that human nature was essentially violent and selfish, it was difficult to argue that more freedom would improve the lot of mankind. The breakthrough was the identification of scarcity as the as the root of human wickedness, since scarcity is theoretically subject to the fecundity of human ingenuity (McClelland, 1996: 433; Xenos, 1987). In Adam Smith (1993), scarcity is transformed into ‘opulence’ through the ingenuity of humanity unleashed by the division of labour. The material conditions for a well-ordered society, to use Rawls’ phrase, now theoretically established, liberals could detach this issue from questions about the form this newly permitted liberty ought to take.

The environmental crisis is essentially the reattachment of the problem of material conditions. To resolve it, despite efforts to be green, liberalism returns to the old solution of technological change. This was always likely to happen, given that economic and political liberalism never broke with each other; they merely pursued other interests. On this point, their interests again converge. Detaching technological optimism would require liberalism to develop either a new ethics of nature or a critical reading of economics that could underpin a non-optimistic decontestation of technological change. The first of these would likely be ecologism by another name. The second is hard to imagine, especially because it may threaten liberty. Any coherent liberal theory may thus be condemned to optimistic decontestations of technological change.

So, Dobson (2007: 150) is correct that environmentalism is the limit for liberalism. Genuine convergence with ecologism is beyond mainstream theories. For ecologism as a theory, technological scepticism is an essential value. It supports the green belief that the environment is too complex to be safely tampered with (Eckersley, 1992). Though contestable, as a value, it is not contingent, unlike an empirical claim. In ecologism, technological change is adjacent to the environment (now a core concept) and is decontested as a destructive, ideologically driven process of power-seeking and acquisitiveness. Though some ecologism scholars see technological change as endogenous, it has generally been decontested as at least partially exogenous, with stress placed on its unreliability (Barry, 2017; Keary, 2023). This morphology is stable because, as a critical theory, ecologism is suspicious of the products of mainstream policymaking. Its effort must be to broaden the acceptance of its decontestations and thereby empower its normative vision.

Feenberg (1990: 103) argues that ‘power depends on control of technology’. We might now reword this as: power depends on control of the idea of technological change. A major part of any political theory’s analysis of why power lies where it does, how it operates, and what might be accomplished with it, depends on its decontestation of technological change. Decontestations are political acts: they close off most possibilities, open a few.

The point is not that political theories should ‘predict’ technological change. Rather, political theories should have a clear conception of technological change, particularly in an era of environmental crisis. The idea of technological change is currently controlled, in government and policy, by technological change modellers advancing the notion of directed technological change (Keary, 2016; Van Vuuren et al., 2004). This idea empowers those claiming that growth and environmental sustainability are reconcilable: that is, supporters of capitalism. The contest over the meaning of technological change is thus an example supporting a broad conception of what politics is. It is just one that has received much less attention than its significance merits.

Conclusion

We began this article by outlining Freeden’s framework for understanding political theories. It allows one to abstract from the particularities of a theory and view it at the level of its concept configuration. We can thus explore how different decontestations of even a seemingly minor concept like technological change can affect the thrust of the theory. Applying it here, we saw that the theoretical infrastructure of Rawls’ liberalism allows only for light green or environmentalist politics. Technological change was shown to be the obstacle upon which green liberal pretentions foundered. Idealised liberal citizens are likely to choose technological optimism, at least until a liberal political theorist can introduce a theory of technological change that permits otherwise. None of Bell, Welburn, Wissenburg, or Stephens were able to overcome this predilection.

It was clear from this discussion that, despite rarely being appreciated, technological change has considerable significance to political theory. This significance is now greater given the advent of the environmental crisis. No political theory is spared the obligation to speak to this crisis and all will need to revisit their sparse treatments of technological change if they are to chart a coherent course on the issue. When doing so, they may well find that cherished doctrines are rendered problematic when sophisticated accounts of the process of technological change are introduced. Even if not, they may find that their existing theoretical commitments render genuinely green politics difficult for them to endorse.