A Response to the Responses, Pt. II: Nudging, Paternalism, and Human Agency

Editor’s Note: Over the past couple weeks, we have hosted three responses to a thought-provoking essay by Charles Mathewes and Christina McRorie on governmental “nudging” and human freedom.  We are grateful to them for offering a careful and thorough response in two parts.

Our first post was focused on some immediate responses that the pieces individually provoked. But beyond the idiosyncrasies of each of the critics’ pieces, they collectively raised some issues that merit direct address on their own. We try to provide that here.

1. Nudges: Myth and Reality

Let’s start with a note of clarification regarding “nudges” themselves. What are these nudges that we have been discussing, and what do they do? Perhaps looking at examples that are already in play in our world—nudges developed by the Office of Information and Regulatory Affairs (OIRA)—may serve to bring a level of coherence, specificity, and clarity to an otherwise fraught, and likely to be mythologized, issue.

A “nudge”—a term brought to public attention by Cass Sunstein and Richard Thaler—generally refers, in the policy world, to a small modification to an already existing “choice architecture,” some context in which we make decisions; the modification is meant to promote certain decisions over others, in a context where some such promotion is inescapable. Many nudges work by simply structuring (we would say “improving,” but you decide) the disclosure of information. One example is the FDA’s new “plate” logo that has replaced the food pyramid, and more directly conveys the recommended proportion of food types. Another is the updated label for reporting the fuel economy of new cars. Like the old food pyramid, the old fuel economy labels worked, but not in the most efficient way possible; they didn’t give consumers a clear sense of what their fuel costs would be in coming years, or where a particular vehicle stood in comparison to others in its class. The new labels make this information more readily digestible, in the hopes that consumers will be more likely to purchase cars that are more fuel-efficient. As should be obvious, refusing to require such labels would not mean that anyone will be “freer,” as if individual’s minds are contingently epistemically constrained by the selection of information the government demands; rather, it would simply mean that their assessment of a given set of cars, say, will be shaped by all the other usual inputs, and without the information disclosure in question (the individual’s budget and sense of what she needs, the color of various models on the lot, the sales pitch of the salesman, the consumer’s preference for cultural images associated with various makes through marketing, etc.).

Other nudges work by building on the power of a default setting. It turns out that even when given a choice to ‘opt out,’ many people stick with whatever option is the ‘default’—whether this is the privacy setting Facebook selects (for its own profit-seeking reasons, please note), or enrollment in a cable service package. Corporations know this; ‘negative option marketing’ exploits this tendency toward inertia in an especially visible way by automatically enrolling those who accept a ‘free’ product in a program that carries a monthly fee, until the consumer explicitly opts out. Recognizing for-profit companies’ attention to lucrative defaults, in recent years the IRS and the Treasury Department have undertaken new initiatives to encourage employers to adopt automatic enrollment plans for their retirement savings programs, in the hopes that more Americans will have money in their retirement as a result—without the cost of either financial education or regulatory intervention. In this case, a government nudge is designed to ensure larger freedom from reliance on government late in life—again, hardly an example of government “control,” but rather quite clearly something like the opposite.

Nudges can also work by exploiting the power of “social norms,” and the fact that we are strongly influenced by what others do. This has been used to great effect by the United Kingdom’s tax agency, which has updated the wording of its correspondence with delinquent taxpayers to note the percentage of citizens in that town who had already paid their taxes. This simple tweak alone to the bills resulted in a measurable increase in compliance, and payment of late taxes.

Now, these nudges are actually rather milder than Vallier assumes when he says that a libertarian will see nudging “as a violation of the right of private property because it involves regulating the use of private property in business.” Violating the right of private property does sound bad! But wait: Unless you count a requirement to label a product as a violation of property rights, none of these nudges violate private property rights. There is a large difference between improving fuel economy labels (in the hopes that consumers will choose more efficient vehicles) and mandating new fuel economy standards (to force producers to make more efficient vehicles). The latter move is just plain old regulation, and not, strictly speaking, a part of the conversation on nudges.

Public policy nudges generally indicate fairly sensible and rather ordinary tweaks to the parts of our lives impacted by government—like a general upgrade in the quality of public service announcements, regulations, and service delivery, based on empirical awareness of the human psyche.

 

2. Paternalism

Of course, clarifying how nudges work may not resolve the underlying philosophical debate about political paternalism and the nature of human freedom tout court. We’d like to turn to that discussion, then, in the hopes that continued disagreement can at least be clear about its terms. Starting with nudging: just what sort of paternalism does it involve?

By nature of being nudges, rather than “pushes,” nudges express what Sunstein and Thaler call “soft paternalism.” (Hard paternalism, such as plain old regulation, “pushes”—it does not preserve freedom of choice.) Furthermore, according to Sunstein, the nudging used by OIRA engages in a “means paternalism,” rather than “ends paternalism”—OIRA is tasked with improving upon regulation and public information campaigns originally built by other agencies, who themselves are designing laws (promoting ends) based on mandates from Congress.

The line between soft and hard paternalism, of course, is a bit fuzzy, and as of yet no exhaustive list of policy options divided clearly into “nudge” and “push” exists. In his recent book Simpler, which recounts his time as the director of OIRA, Sunstein is open about some of these: for example, should financial incentives count as nudges, or pushes? Financial disincentives (vice taxes, for example, or fines) have obviously crossed the line into push territory—but what does that mean for incentives such as tax breaks?

Ambiguities aside, nudges appear to have already streamlined a good amount of policy, enabling regulation that is more efficient, with less oversight. (And less oversight is, we trust our colleagues will agree, a prima facie good, both on philosophical grounds, and for the purely pecuniary reason that oversight and enforcement is expensive.)

A reasonable libertarian skeptic might still object on philosophical grounds that paternalism of any sort—soft or hard, choice-preserving or choice-limiting—is inappropriate for government.[1] Why can’t we just keep government action as neutral as possible, one might ask, so that we know what we’re getting?

Here, we begin to arrive at one of the misconceptions undergirding Vallier’s response. What the concept of choice architecture used in behavioral science reveals is that there is no option to not have choice architecture. Or, in other words, government action is never neutral—because no action is. If a re-worded tax bill is paternalistic, so is the pre-re-worded tax bill; they each offer a choice architecture. One of them is x% likely to induce you to pay, and one of them is x+y% likely to induce you to pay. But there is no non-inducing structure, no utopia of unconstrained agency.

Recognizing this goes a long way toward addressing some of Vallier’s concerns over whether we ought to “give someone the moral authority to alter our preferences”, and whether “handing over power” is appropriate. He asks, “where do the nudgers get that sort of authority over individuals, authority sufficient to deliberately alter their preferences?” What this question still relies on is a misconception of nudging—or, perhaps more accurately, a misconception of human agency itself. Tweaking policy with behavioral science insights won’t necessarily alter our preferences—or, if it does, it just does so in a different way than non-tweaked policy did before.

 

3. Agency

The question of an “un-tweaked” life—and the fear of having one’s preferences altered—brings us to a third topic in need of clarification, and to our original philosophical claim. It has to do with a fundamental misconception about the conditions of agency today, and especially the conditions set by life within a market-based society.

The more libertarian respondents seem to assume that while nudging is inappropriate for governments on the basis that one cannot escape government nudging, it is appropriate in markets, given that markets are disciplined by consumer desire in ways governments are not (meaning, theoretically, that unwanted nudging will go unrewarded in markets, and die out). In short, while they suspect government nudging for different reasons (subsidiarity, no guarantee that administrators have special epistemic insight regarding the good, bureaucratic corruption, government unresponsiveness, etc.), market nudging is less problematic because, as Vallier claims, markets are “input-responsive social systems.”

This account of our agency in markets sounds pretty. And on a certain, superficial level, this is true. We can choose which cereal to buy, and those of us who have recently read something on behavioral economics may remind ourselves to think twice about grabbing the cereal at eye level, and outsmart Kellogg’s after all, slotting fees or no.[2]

On a deeper level, however, the market itself is not an input responsive social system. The irony in Baker and Watson’s comment that the influence of government can only be avoided by immigration is that, while a move to Canada is at least thinkable, a move away from the market is not: you can’t immigrate away from the choice architectures set by the market. We may only interact with government choice architecture a few times a week or day, but we are never not interacting with markets—they are the waters in which we swim. Advertising for commodities is only the most visible form in which this is so. (And the more you know about the explosion the advertising industry over the past century—it is now the second largest industry in the world, behind weapons—the more alarming that thought is.) But we are also shaped in nondiscursive ways, ways that make it harder to maintain the illusion that consumer choice is simply a pure and unmodified exercise of “full human agency”.  That is, living in a market system is the biggest “nudge” there is, and the libertarian account of agency is symptomatic of that nudge.

Consider, for example, the fact of our suburbanized world. While it is not fair to blame post-WW II suburbanization entirely on the market (government-underwritten loans had a lot to do with incentivizing the development market), we certainly can blame a lot of it on the market.[3] We may now have an unprecedented level of choice between home styles in various suburban lifestyle enclaves—but is such a racially and socioeconomically segregated and vehicle-dependent configuration of life the choice architecture that we, upon considered reflection, might have chosen? It turns out, no; judging from the renewed popularity of urban life, many Americans now would not prefer the suburbanization of our cities (although wide scale returns to them have required substantial government support, in the form of increased civic infrastructure). This is a simple, if invisible, market choice architecture that faces many of us, and, whether we live in suburbs or not, its shaping impact on our lives extends beyond mere consumption into our very politics and culture.[4] This returns us to our original conversation about varying accounts of freedom, and the age-old distinction between what can be called “freedom from” and “freedom for.” (As should be clear by now, we’re more convinced by Amartya Sen’s account of freedom than we are Robert Nozick’s.) While the market ensures us freedom from, the case of suburbia is just one instance in which this negative freedom may not, at the end of the day, be enough to let many of us achieve the ends that we reflectively hold to be good.

The example of suburbanization also nicely illustrates the stakes of the disagreement Vallier helpfully raises, over whether the “spontaneous order process” of the market is intrinsically valuable, because it preserves procedural choice. (The scare quotes indicate that, if you agree that preexisting choice architecture conditions agency, then the orders emerging in markets may not deserve the term ‘spontaneous.’) While more minimal arguments on behalf of spontaneous order simply argue that it is more effective than planned action, stronger cases see in it a normative ideal, based on normative respect for choice. Vallier plainly evidences an even more robust commitment to this ideal with his remarkable claim, “God’s will is best expressed through spontaneous order processes rather than the cruder and more corruptible top-down hand of government.”

This is certainly a bedrock area of philosophical and theological disagreement, and it is good to have it out in the open. This blanket endorsement of spontaneous order—which can be read as code for procedural freedom, freedom from—often has the effect of forcing us to baptize the results of any spontaneous order as just, no matter how undesirable—or worse, tricks us into simply exempting them from moral scrutiny altogether.

Disagreement with Vallier over the intrinsic good of spontaneous order, however, does not require inverting his boldly drawn market-good, government-bad equation. Indeed, it is a mistake to read this response, or our original essay, as either a naïve and undiscriminating endorsement of government activity or a censure of market activity. That said, we are rather less sanguine than is Vallier about laissez faire outcomes. (You could say that we’re somewhere in the middle, with Adam Smith, who, despite his apparent libertarian reputation, devoted almost an entire book of his Wealth of Nations (Book V) to enumerating various market failures, and outlining government interventions to correct for these failures.)

In fact, we are as ambivalent about market processes themselves as we are about government processes, and for reasons that are at least as theological as practical. That Vallier’s response assumes a theological account of market activity as uncomplicatedly expressing the divine will strikes us as extraordinary. And equally extraordinarily, Vallier simply assumes, rather than explains, why the outcome of the aggregated pursuit of profit should result in the accomplishment of God’s will. In the wake of the 2008 financial crisis (as we mentioned in the last post, a crisis in no small part precipitated by overconfidence in market outcomes in the absence of regulation), such staunch optimism seems as practically indefensible as it is theologically opaque. Almost as importantly, this sort of claim cloaks in quasi-theological language a move to outsource responsibility for the common good to the market—of baptizing the market as theologically sanctified—and accordingly encourages a humble obeisance before any and all market outcomes.

The only way such a perspective makes sense is if rational choice (in which “rational” is understood very minimally, as simply any choice) is the highest good that can be aimed at in public deliberations—a view that brings us full circle to the libertarian philosophy undergirding William’s original claims. This perspective refuses to arbitrate between the ends that people appear to have rationally chosen—even when they express discontent with these ends—on the grounds of an extreme epistemic reserve with regard to the good. Do many people prefer the long-term effects of smoking, as judged by the fact that many people smoke? the logic goes. Then it is not the government’s job to engage in anti-smoking campaigns.

Although we also value choice, as scholars of religion, we feel no need to be quite so agnostic about the good, are hardly likely to be satisfied with such a hollow account of rationality, and, as noted, are skeptical of claims that equate the divine will with consumer choice. Moreover, as we have been trying to argue, the very premise of this line of reasoning misconstrues some of what is at stake: consumer choice does not intrinsically maximize freedom (as freedom for), given that the market is not a spontaneous order process—in each moment, consumer agency is conditioned by the choice architecture of previous moments, and often for (what we are not afraid to call) the worse.

 

To begin to sum this up: our interlocutors point out that government accountability matters, and we agree: public input over the ends toward which we are being nudged by government regulation, policy, and public education is absolutely necessary for democracy to work well. This was true before the advent of behavioral economics, and it is true now. But we should not exempt markets from the ethical scrutiny to which we subject government action, especially where such exemption is based on indefensible assumptions about how markets work, or about how powerful and unconditioned consumer agency really is. Simplistic narratives setting “market freedom” against “government tyranny” reduce our ability to understand the real questions of our time, and doom our conversations to an insular sort of philosophical circularity.

The question before us now is not, “Should we engage in nudging on behalf of the public?” In light of the fact our lives are constantly being nudged—both by government and the very shape of the markets in which we swim every day—the question is instead, “How ought we to use the tools we have at hand to reflectively order our lives together so as to best promote the common good?” In this way, discussions over nudging and the practical impact of our public policies can bring to the fore fundamental questions about the nature of human freedom, and our common life together.