Fallacies, Paternalism, and The Need for Nudges - Part II

The following is a continuation of my series on nudges. We are picking up directly from the 401k plan experiment that was discussed at the end of the first article, but I will also re-state it below. Make sure to read that! Let's continue with our exploration of liberal paternalism.

3. The Grounds for Liberal Paternalism:

Let us consider an employer who wants to increase the savings rates of their employees. Option 1 is to set a default of enrolling all employees in a 401k plan unless the employee requests not to be enrolled. Option 2 is to provide employees with a choice to agree to increase their contributions to savings with every raise. Both work towards achieving the employer’s goals – however, Option 1 sets the default setting to encourage savings while Option 2 sets the default setting such that all earnings will only contribute to current income. In this scenario, an automatic opt-in design can be considered purely paternalistic since it steers employees towards saving earlier. It even retains high enrollment rates by playing into the employee’s psychology to stick to the status quo and avoid extra effort for a retrospectively irrational choice. However, this is clearly non-libertarian paternalism since it presents the choices on an unbalanced scale to the employee in the first place. On the other hand, the latter option can be said to pass the libertarian test since the default design features are not constructed such that one choice is heavily pushed over the other; it meets “libertarian muster” [3] in the sense that no choice is any less available or accessible than the other. 

The savings plan example given above highlights an important aspect of human decision-making in the real world – it is highly context-dependent and not always optimal. Mental factors can be played with by tweaking default settings – for instance, Option 1 of the ‘savings plan’ example leverages the individual’s psychological inertia [16] (i.e., the tendency to maintain the status quo) to retain their enrollment in the 401k plan. Similarly, there are a few other commonplace fallacies stated by Thaler and Sunstein that contribute to an individual’s proneness towards irrationality:

A. Framing effects: The sentence “Of those who undergo this procedure, 90% are still alive after five years” is more likely to convince an ill person to undergo treatment than “Of those who undergo this procedure, 10% are dead after five years.”, even if they are logically equivalent. [4]

B. Bayes’ Rule: In a city with overwhelmingly more non-terrorists than terrorists (take 100 terrorists and 1 million non-terrorists), a facial-recognition system with a 1% false positive rate will actually ring 99% of the time at non-terrorists according to Bayes’ Rule. However, many would falsely say that the system would be accurate 99% of the time instead of 1%.

C. Preference Reversals: People say they prefer A to B as well as B to A in the same context on separate occasions. [3]

The supposition that interference is unnecessary for a functional society must presume that humans are unambiguous, consistent, and well-informed decision makers. However, Thaler and Sunstein’s concern is that this is not always guaranteed due to the cognitive errors mentioned above, which gives them reason to push liberal paternalism. They draw their evidence from studies in cognitive science and behavioral economics which show that individuals can sometime be prone to making “…inferior decisions in terms of their own welfare – decisions that they would change if they had complete information, unlimited cognitive abilities, and no lack of self-control.” [3]

Now that we have a sufficient understanding of the motivation and aim of liberal paternalism, we can delve into the concept of “nudging”. All the considerations that differentiate liberal paternalism from hard paternalism can essentially be classified as a nudge – as Thaler defines it, they are “liberty-preserving approaches that steer people in particular directions, but also allow them to go their own way.” [5] We will look at how it is the hallmark of a liberal paternalistic government and the ethical debate surrounding it.

4. Nudges and Their Effects:

Nudges have been used by governments are other institutions for years and may lie outside of the parameters set by Thaler and Sunstein (henceforth T&S). For instance, a nudge may just as easily steer someone towards a bad decision than a good one, and it may constitute coercive behavior and put constraints on choices as well. On the other hand, the act of, say, changing the order of food in a cafeteria line to promote healthy eating by placing the cake at the very end is a classic example of a nudge in libertarian paternalism. [1] Thus, nudges can be made compatible with liberal paternalism, and these non-coercive variants are pushed as the main weapon to implement T&S’ political ideology.

Simply leaving the choice set unaltered, however, does not mean that nudges have achieved the potency of libertarianism just yet. They may not constrain options, but is that an extensive definition of the spirit of free will? Surely, that entails a broader sense of freedom, and this is the concept of “autonomy” – defined by Hausman & Welch as “the control an individual has over his or her own evaluations and choices.” [6] Nudges under liberal paternalism may justifiably fail the autonomy test since the individual may still be influenced non-trivially, such as by merely changing the arrangement of choices in the cafeteria example. The compromise of autonomy has been a long-standing point for nudge critics and has been addressed in several different ways.

As Saghai points out [7], T&S establish two main arguments to defend that nudges preserve free will: “…either nudges preserve freedom of choice simply because the nudger does not foreclose any option, or nudges preserve freedom of choice in a more substantive sense.” In the former argument, the definition of liberty is simply taken to mean an equal accessibility of all options, thus allowing a wider variation and intensity of nudges to be included in the toolbox of the liberal paternalist. The second considers conventional autonomy as the bar for liberty, making it closer holistically to the spirit of libertarianism but “… more humble because fewer influences will qualify as nudges” [7] in comparison.

The ideal solution would maintain the wider nudge spectrum of the former argument while expanding the definition of liberty towards that of the latter argument. Dealing with this comes down to the question of whether nudges respect individual autonomy or not.

Now we are ready to look at the pros and cons of nudges. In the next article, I will present a couple of supporting and opposing justifications for nudges, so that we can get a well-balanced view of the moral landscape. See you next time!

REFERENCES:

[1] Thaler, R. H., & Sunstein, C. R. (2008). Nudge: Improving decisions about health, wealth, and happiness. Yale University Press.

[3] Thaler, R. H., & Sunstein, C. R. (2003). Libertarian Paternalism. The American Economic Review, 93(2), 175–179. http://www.jstor.org/stable/3132220

[4] Sunstein, C. R., & Thaler, R. H. (2003). Libertarian Paternalism Is Not an Oxymoron. The University of Chicago Law Review, 70(4), 1159–1202. https://doi.org/10.2307/1600573

[6] Hausman, Daniel & Welch, Brynn. (2009). Debate: To Nudge or Not to Nudge*. Journal of Political Philosophy. 18. 123 - 136. 10.1111/j.1467-9760.2009.00351.x.

[7] Saghai, Y (2013). Salvaging the concept of nudge. Journal of Medical Ethics. 39:487-493.

[16] Jung. D. (2019). Nudge action: Overcoming decision inertia in financial planning tools. behavioraleconomics.com.