The rhetoric of reaction, extended

Sunstein's (2022) stimulating article highlights the now classic book by Albert Hirschman, The Rhetoric of Reaction: Perversity, Futility, Jeopardy (1991) taxonomizing three rhetorical strategies – Perversity, Futility and Jeopardy – that can be used by powerful interests, and their supporters, in opposing reforms to the political, social or economic system. Suppose, for example, a fossil fuel company is attempting to undermine regulations enforcing a rapid transition to electric vehicles. The rhetoric of Perversity claims that such regulation will have the opposite of the intended consequence – for example, suggesting that when the “embedded carbon” involved in battery manufacture, or the fossil fuels consumed by the power network are taken into account, switching to electric vehicles will actually worsen, rather than remediating, greenhouse gas emissions. The rhetoric of Futility, in contrast, suggests that the proposed reform will make no meaningful difference. One version of this approach suggests that if one nation burns fewer fossil fuels, those same fossil fuels will be burned instead by some less scrupulous nation. Finally, the rhetoric of Jeopardy highlights the danger of unintended consequences – perhaps, the switch to (currently expensive) electric vehicles will impose unreasonable costs on the poorer sections of society, hamper economic growth and/or reduce international competitiveness.

Sunstein uses a discussion of Hirschman's ideas as a springboard for critiquing a recent paper of ours (Chater & Loewenstein, in press), in which we question the promotion of nudges as a key public policy tool. In this response, we take the baton from Sunstein by beginning with Hirschman and showing how the same critiques can equally be levelled in the opposite direction. We then identify the widespread use of a fourth rhetorical strategy that is inherently reactionary, and which leads to very different conclusions than those advanced by Sunstein.

Sunstein accurately describes the three strategies identified by Hirschman as arguments used by reactionaries opposed to progress, and much more briefly mentions the three historical examples that Hirschman carries through the book to illustrate those strategies (Hirschman wryly notes his own attraction to tripartite distinctions): (1) the assertion of equality before the law and of civil rights in general (which Hirschman associates with the French Revolution), (2) the enactment of universal suffrage and (3) the implementation of the diverse reforms that have collectively come to be known as the “welfare state.” These are hugely consequential systemic changes that fundamentally changed the rules by which society operates – and were, Hirschman argues, vigorously and persistently resisted by powerful vested interests who gained from the status quo, using the rhetoric of Perversity, Futility and Jeopardy.

In what is itself a rather striking rhetorical move, Sunstein suggests that these very same reactionary arguments are now being used, not just by opponents of transformational changes, but also by those skeptical of nudges as a public policy tool.¹ Indeed, he suggests that our paper exemplifies the rhetoric of reaction. But in the context of the changes Hirschman studied, it would seem that our approach is the opposite of reactionary. Given the severity of problems facing the USA, the UK and other parts of the world, we argue, in our paper, for the need for just the types of far-reaching, substantive, transformative, policies that Hirschman sees as essential for human progress, and as impeded by the forces of reaction. In contrast, nudges, aptly described by Daniel Kahneman as “achieving medium-sized gains by nano-sized investments,”² are almost invariably incremental, typically have at best modest impacts (as Sunstein acknowledges) and are often ineffective and sometimes counterproductive (here the concerns of Futility and Perversity are not grounded in mere rhetoric but rather come directly from meta-analyses of empirical trials, for example, DellaVigna & Linos, 2022; Maier et al., 2022; Mertens et al., 2022).

The rhetoric of reaction can, it is true, be marshalled by those on either side of the political spectrum. Hirschman's examples of the rhetoric of reaction are almost all cases in which right-wing commentators attack more left-wing proposals. But as both Hirschman and Sunstein acknowledge, the same rhetorical strategies can be deployed to many different ends.³ Indeed, a point made by Thaler in our personal debates with him about these issues is that politics in the USA is so broken that there is no realistic chance of substantive systemic reform. Nudges, therefore, are one of the only tools we have; more ambitious policy proposals are “pie-in-the-sky” and risk undermining the good that nudges can do. So, rhetorical charges of Futility and Perversity can potentially be directed just as much at systemic reform as at nudges. From one viewpoint, systemic changes are painted as politically impractical and standing in the way of more feasible nudge interventions; from the other, nudge interventions are viewed as an ineffective sticking-plaster, the promise of which distracts attention from the need for substantive reform.

Are critiques of nudges reactionary, or are nudges themselves? A hint may be offered by the titles (and approaches) of the two papers that ushered in what could be called the “nudge revolution”: One (Thaler & Sunstein, 2003) was titled “Libertarian Paternalism”; the other (Camerer et al., 2003) “Regulation for Conservatives.” Thus, the initial appeal of nudging came from the promise of avoiding conventional forms of regulatory reform – thus, aligning with a conservative agenda – but nonetheless hoping to achieve welfare-improving objectives through light-touch, behaviorally-inspired methods.

It is easy, then, to see how the nudge movement might align with what would traditionally be seen as a reactionary agenda – to the extent that nudges are seen as a potential substitute for (rather than a complement to) more substantial systemic reform. Indeed, in the context of Hirschman's rhetoric of reaction, the main thrust of our paper (Chater & Loewenstein, in press) could be seen as the identification of what amounts to a fourth rhetorical strategy that could be called Blaming the Individual (BTI). Unlike Hirschman's three rhetorical strategies, the BTI strategy does not focus on the proposed systemic reform and its putative consequences. It aims to shift the focus away from systemic reform of any kind, by suggesting that the real source of the social problem at issue lies with failings in individual behavior, and that overcoming these individual failings is the appropriate route to a solution. Although BTI, unlike the three rhetorical strategies identified by Hirschman, is inherently reactionary, a particularly elegant aspect of the BTI strategy is that it does not, superficially, look like a mode of reaction at all. Indeed, the audience for the rhetorical strategy is encouraged to believe that the reactionary interest group (typically a large corporation) is actually trying to help solve the problem.

A large proportion of our paper is devoted to illustrating the prevalence of the BTI strategy, in practice, across a wide range of topics: from climate change, to the obesity crisis, to plastic pollution, the problem of long-term savings, health-care costs, gun control and many more. Take, for example, British Petroleum's promotion of the idea of personal carbon footprints (Mann, 2021), backed by a huge advertising budget, including the creation of a personal carbon calculator, and injunctions to personal behavior change (“it is time to consider a low-carbon diet,” Learmonth, 2020). Or consider the long-running Keep America Beautiful campaign, created and funded (for more than half a century) primarily by the packaging and soft drinks industry. These campaigns do not attack particular systemic reforms directly, by suggesting that they have perverse effects, no effects or unwanted side-effects – instead, they promote the idea that the source of the problem, and any likely solution, lies with individual behavior: and the solution must, it is therefore presumed, lie with helping individuals to overcome their own limitations, rather than focusing on system-level change. People need to be better informed to make more climate friendly consumer choices and need to be educated about the importance of not dropping litter. With the blame for the problem firmly pinned on individual behavior, the role of oil companies and the packaging industry fades into the background.

To put the point in the terminology of cognitive psychology and behavioral economics, the BTI strategy is an exercise in framing. Specifically, the strategy aims to promote what we call the i-frame (that the problem at hand, and any solution, arises from individual behavior), rather than the s-frame (that the problem arises primarily through the operations of the current system, rather than weaknesses of its individual members, and hence that solutions may likely require conventional policies, such as regulation and taxation). Why is this type of framing effect potentially so powerful? We do not have space to give a full outline here (though see Chater & Loewenstein, in press), but reasons include the fundamentally serial nature of thought, which means that adopting one framing of a problem tends to block out other frames; the generally “competitive” nature of blame attribution (if factor A is responsible, then factor B must be less so), leading to what Weber (1997) refers to a “single-action bias,” and what appears to be a default psychological tendency to focus on individuals rather than situational and systemic factors (known in social psychology of the “fundamental attribution error” or “correspondence bias”).

Industry and the PR and marketing agencies that represent it are certainly firm believers in the power of the BTI rhetorical strategy, backing this belief with billions of dollars of spending over many decades. Indeed, throwing responsibility back to the individual is so ubiquitous that we scarcely notice it: drinkers are helpfully enjoined to “please drink responsibly” on packaging and ad campaigns in the UK; similarly, UK gamblers are repeatedly told “when the fun stops, stop.” The message is pretty clear: the devastating consequences of alcohol or gambling addiction are the responsibility of the individual. And the industry vigorously opposes systemic changes that have a good evidential basis, such as minimum alcohol unit pricing, or restricting the stakes on so-called fixed-odds betting terminals. Of course, industry fully realizes that systemic interventions are where the real leverage lies: the focus on individual-level behavior as the cause, and key to the solution, of the problem is largely a smokescreen, rather than a serious proposal.

And here we come to the problem, which is at the core of our original paper, prompting Sunstein's critique. The problem is that promotion of the i-frame is not merely the domain of powerful reactionary interest groups; it has also received a boost from many well-intentioned policy makers and academics (including ourselves). The essence of the nudge approach is to identify individual cognitive and decisional limitations as the cause of problems such as obesity, high healthcare costs and low levels of financial preparedness for retirement. Indeed, the first several chapters of the “final edition” of Nudge adhere to the first edition's focus on individual biases and mistakes. Subsequent chapters propose how to attack problems in the policy domain with nudges that help individuals to overcome those limitations. These include attempts to improve financial education and help people make better food choices through calorie and nutritional labeling; mandatory disclosures of conflicts of interest, especially regarding financial or medical advice; having to “sign up” to terms and conditions concerning how one's on-line data will be used, and much more. The idea is that people can be helped to make better environmental, health, pensions or other decisions concerning their own lives, or their impact on others, by fixing faults in their decision making. People can be opted in to a pension, defaulted into a green energy tariff, given information about their neighbors’ energy consumption, and so on, making the “right” choice easier (where “right” is usually thought of primarily, though not exclusively, in terms of individual welfare). Yet by focusing on interventions that are supposed to change individual behavior, and treating the system as fixed, we have inadvertently been promoting the idea that large-scale social problems have individualistic solutions, just as the forces of reaction would like us to do. Our worry is that large parts of behavioral science research have inadvertently aligned with what Hirschman would have seen as the forces of reaction – not only through the three rhetorical strategies that Hirschman highlights but also by adopting an approach to public policy that targets, and blames, the individual and thus reduces the focus on the systemic nature of most problems in public policy.

We entirely agree with Sunstein that the debates over competing approaches to public policy can be usefully informed by thinking about the rhetoric of reaction. As we have argued, we see the promotion of an individualistic perspective as forming a fourth, and particularly insidious, rhetorical strategy, by which powerful interests hoping to maintain the status quo appear to be helping to address deep problems, while actually trying to frustrate meaningful reform.

To take a particularly extreme case, consider the NRA's famous slogan “Guns don't kill people, people do.” The problem of gun violence is, the slogan tells us, a matter of individual behavior – and the solution, it is implied, must be to focus on individuals (hence the emphasis on “mental health” as a supposed line of intervention, although one which has no credible evidential basis (Lu & Temple, 2019)). Few behaviorally oriented public policy researchers have actively engaged with the problem of gun violence from an individual-level perspective. But were they do to so, we suspect they would be at pains to make clear that the real levers in reducing gun violence are almost certainly radically restricting firearms, as is clear from cross national comparisons and historical trends. We as a field should be concerned about giving the impression that dramatic regulatory changes can be avoided if we are to significantly reduce gun violence in the USA (see Kang et al., 2020).

Unfortunately, we have not, as a field, always been so cautious. Enthusiastic claims have been made for the efficacy of green defaults (Liebe et al., 2021), even in direct comparison with traditional substantive measures to reduce carbon emissions (Sunstein, 2021). Researchers have genuinely hoped that the crumbling of defined benefit pension schemes might be repaired by cleverly designed combinations of auto enrolment and auto-escalation. There have been well-intentioned and serious attempts to help address the obesity crisis, which has grown out of radical changes in the food system with the prevalence and pricing of ultra-processed high energy density food over the past few decades, with behavioral interventions to encourage gym membership (Charness & Gneezy, 2009) or reduce portion sizes (Downs & Loewenstein, 2011). And there have been misguided hopes that problems caused by the byzantine health insurance system in the USA could be at least partially mitigated merely by giving people better advice about how to choose the right insurance policy (Johnson et al., 2013), rather than entirely restructuring the insurance market to eliminate a decision that people dread and are not competent to make (Bhargava & Loewenstein, 2015)

Indeed, within the behavioral policy community, there has sometimes been the sense, at least in informal discussions with regulators, governments and other academics, that in these domains, a behaviorally oriented researcher proposing traditional policy measures, including incentives, taxation and regulation, is simply revealing their lack of ingenuity. Such heavy-handed and old-fashioned measures should surely not be necessary, one can be led to feel, when the full repertoire of behaviorally inspired nudges is at our disposal. This is, of course, precisely what the forces of reaction want to see: policymakers and academics becoming wary of using the powerful levers for systemic reform which are typically the real hope for substantial change.

The rhetoric of reaction through the looking glass

As we have indicated, Sunstein's perspective on Hirschman's (1991) framework is very different from our own. While we see our arguments as complementing and extending Hirschman's discussion of the rhetorical strategies deployed by powerful interests to maintain the status quo, Sunstein, to our minds paradoxically, views our critique as itself reactionary.

From Sunstein's point of view, advocates of nudges are identified as the forces of progress; and opponents of nudges are, by inevitable extension, the forces of reaction. So, when such critics (e.g., Bubb & Pildes, 2014; Osman et al., 2020), including ourselves, suggest that nudges may sometimes backfire, be ineffective or have unfortunate side-effects, those critics can be identified with, and perhaps dismissed as, exemplifying the rhetoric of reaction.

Opposition to any viewpoint, of course, can be viewed as reacting against it, and, to that extent, being reactionary. But in the context of public policy, and in the light of the huge social, political and economic challenges that we collectively face, it seems far more natural, and appropriate, to see the reactionary forces as those opposed to serious policy changes.

Moreover, the need for, and direction of, policy change is often relatively uncontroversial, particularly in the light of historical and international comparison. In many rich nations, individuals have become dramatically heavier, and substantially less well prepared for retirement, within a little more than a generation. But surely their individual decision-making is no more or less rational – and, for example, their tendency to favor present rewards (whether in eating or spending) over future gains is surely unchanged. What has changed is the system within which they operate – and to fix such fundamental problems, we surely need the system to change further (perhaps reversing some of the changes that generated the problem in the first place). Looking across nations, for example, at pensions in the USA vs Australia (where dramatic pension reform has been enacted), we see that large-scale systemic reforms can and do work. On the other hand, individual-level changes have not had meaningful impacts on the problems of savings or obesity – or indeed any other social problem.

If, as we claim, at least the general direction of required systemic policy change is often widely agreed upon, why are such policies not enacted? Part of the explanation, we suggest, comes from the rhetoric of reaction generated, and heavily funded, by powerful lobbies who benefit greatly from those changes not occurring. Most researchers in behavioral public policy would like to see themselves as on the side of the individual citizen, rather than the powerful interest groups seeking to maintain the status quo. But many of us, ourselves included, may have inadvertently been playing for the wrong team.

Why behavioral science still matters for public policy

Our paper does not call for an end to research on, or implementation of, nudges. We simply highlight findings that both question their general effectiveness (DellaVigna & Linos, 2022) and their potential to crowd out support for more substantive policies (Werfel, 2017; Hagmann et al., 2019).⁴ To the extent that nudges are effective, and do not reduce (or can even increase) support for the kinds of far-reaching policy changes that are so sorely needed, we enthusiastically embrace their continuing evaluation and implementation. And while we lament what we see as a tendency to see nudges as the only applications of behavioral economics to public policy (see Loewenstein & Chater, 2017), and question whether the enormous human and financial resources being applied to researching nudges are being spent as effectively as they could be, we do not question the importance of behaviorally informed public policy. Indeed, more than a quarter of our paper is devoted to outlining an expanded vision of how behavioral science can inform public policy. Behavioral scientists can play an enormously important role in designing, testing and implementing effective public policy – as long as we do not allow ourselves to be distracted by the forces of reaction.