By Diana van Hout

---

This blog is based on: D. van Hout, Gedragsbeïnvloeding in het belastingrecht: Are you ‘nudge’. Tijdschrift voor Fiscaal Recht, 2018 [549-550], p. 928-936. The title of this blogpost is a translation of the title of aforementioned publication.

1.       Introduction

At the beginning of January 2020, new documents were revealed regarding Cambridge Analytica. In the spring of 2018, it was discovered that Cambridge Analytica used data of 87 million Facebook users, which may have swayed voters’ behaviour. The new documents have come from whistle-blower Brittany Kaiser and, over the following months, more than 100,000 documents would lay bare new information regarding these allegations. Nowadays, detailed personal information of civilians is more and more readily available due to social media, the Internet, mobile phones, surveillance camera’s et cetera. Therefore, it became much easier to influence human behaviour. In tax law, the latter development is also recognizable. The aim of influencing taxpayers’ behaviour is mostly to enhance voluntary compliance. These behaviour interventions are mostly defined as ‘nudges’. In, e.g., Canada, the United Kingdom, the Netherlands and Belgium, tax administrations already use nudges in tax law. A nudge is, simply put, a way to encourage people to make better decisions. Thaler, the 2017 Nobel laureate in economics, calls it ‘choice architecture’[1] because it assists persons to make the right choice. But what are the right choices in tax law? Is it to pay more taxes? Is it to pay taxes in time? Is it to refrain from tax avoidance? Who decides what the best choice is and on what grounds? Lastly, how do we distinguish the line between nudging and manipulation? This Blogpost analyses the boundaries and risks of nudging in taxation. It is meant to answer the question if and how legal protection of taxpayers can be provided against nudging by tax administrations. Dutch tax law will be the starting point for the analysis.

2.       Nudging

In tax law, nudging is usually meant to encourage taxpayers to meet their legal obligations. In other words, nudging is used to improve compliance and to stimulate taxpayers to pay their taxes in time. Thaler and Sunstein’s general definition, however, is that a nudge is ‘any aspect of the choice architecture that alters people’s behavior in a predictable way without forbidding any options or significantly changing economic incentives’[2]. In other words, nudging does not affect personal autonomy, and merely jogs a person towards a decision they already wanted to make. This last aspect harbours an element of paternalism or, as Thaler and Sunstein put it, ‘libertarian paternalism’. According to them, the term ‘nudges’ is an acronym of ‘iNcentives, Understanding mappings, Defaults, Give feedback, Expect error and Structure complex choices”[3]. Simply said, it can be defined as follows:

• iNcentives: make persons aware of the benefits of certain behaviour;
• Understanding mapping: structure options to simplify comparisons;
• Defaults: make the desirable option the default;
• Give feedback: show what a person is doing well or warn them otherwise;
• Expect error: anticipate on certain mistakes;
• Structure complex choices: structure options in such a way to influence strategy.

In addition to the nudges mentioned above, Thaler and Sunstein point out many other distinguishable types of nudges. Therefore, the term ‘nudges’ is used to define more types of behavioural interventions than those in the acronym. Over time, many other researchers have started to use the term ‘nudges’ to define all kinds of behavioural insights with a paternalistic motive.

3.       Applications of nudges in tax law

Tax administrations appears to have different initiatives regarding nudging although literature about it is rare. The following sections cover some examples of nudging used by the Dutch and Belgian Tax Administration, but most examples became known through incidental reports in newspapers and journals.

A.      Communication

Nudging is often associated with the way in which the tax administration approaches taxpayers. The British Behavioural Insights Team (BIT team) – informally referred to as the nudge unit – has used insights of behavioural science to improve communication with its taxpayers. Their conclusions have been an inspiration to their colleagues in Europe. Belgium, for example, has used the BIT’s conclusions to revise their letters to defaulted taxpayers. They removed legal provisions, addressed the taxpayer by name and pointed out the potential consequences of their defaults. As a result, the speed at which taxpayers paid their debts accelerated significantly. The idea of this form of nudging is that simpler, more specific and personal letters lead to improvements in compliance. This idea has also been implemented in the Netherlands. The Dutch tax administration uses guidelines instructing agents to, for example, cut back on legal terminology.

B.      Pop-ups

The Dutch tax administration uses nudges in pop-up notifications in their digital forms. Taxpayers are, for example, reminded of the benefits of using the government’s official digital e-mail service (as a replacement of physical mail), and of the fact that it only takes five minutes to activate that service. The pop-ups can be closed, but many taxpayers activated the digital mail service because of the pop-ups.

C.      Preventing errors

The Dutch tax administration has used data analysis and discovered that taxpayers who have just been through a divorce are more likely to make mistakes on their tax returns than other taxpayers are. Therefore, the tax administration proactively approached these taxpayers through letters pointing out this statistic and offering assistance. Research has shown that this proactive approach has reduced the number of mistakes by 30 to 70%.

4.       Risks of nudges in tax law

The idea of nudging is benign: taxpayers are stimulated in a friendly way to make better choices. Proponents of nudging argue that it is a much less invasive method of achieving compliance compared to, for example, fines. Nevertheless, even Thaler and Sunstein have pointed out several dangers as a consequence of nudging[4]. In this blog, I solely mention two dangers. Firstly, the boundaries of nudging are unclear so nudging can easily turn into manipulation. Secondly, nudging is not always in the best interest of the target.

The risks of nudging are already apparent when trying to identify them because it is unclear which measure is a nudge and which is not? The slippery slope risk exists because the boundaries are not clearly delineated, and it is unclear up to what point nudging is acceptable. As far as nudging is used to communicate with the taxpayer in a better and friendlier way, this only seems desirable. The same goes for nudges that make taxpayer compliance easier. However, it has been argued that nudging in the context of taxation will always affect the taxpayer’s autonomy[5]. The aforementioned pop-up about the official digital mail service might have been effective, but the Dutch National Ombudsman concluded that some taxpayers felt pressured and sometimes subconsciously activated the service. It is unclear if the pressure was a direct result of the pop-up, but this example raises the question whether nudging leads to the desired effects and for whom those effects are desirable. Some taxpayers never actually used the digital mail service because they felt like they consented to something they did not really want. A default option can therefore have adverse effects.

5.       Should taxpayers be protected against behavioural interventions?

No one will be opposed to the idea that everyone should file their tax returns truthfully, pay their taxes on time or that the communication between tax administration and taxpayer is clear. These goals are not controversial. However, in my opinion, the current Dutch law (including the principles of good governance) offers too little legal protection against the use of behavioural insights by the tax administration. After all, the aforementioned nudge to activate the digital mail service is completely in accordance with the law, and it was not a violation of any principle of good governance. Nevertheless, these ostensibly benign measures have caused taxpayers to feel pressured. To make sure these interventions do not violate taxpayers’ autonomy too much, it is important to have clear boundaries, even when the intervention is merely used to make sure a legal obligation is complied with. Therefore, I would like to mention three tests that can be used to find the boundaries of acceptable nudging in tax law.[6] However, this does not mean that passing these tests should be the only requirement. The tests are merely a guideline to determine whether the use of behavioural insights violates a taxpayer’s autonomy. When the interventions become more pervasive, a specific legal standard is necessary to safeguard the legal protection of taxpayers.

A.      Publicity test

Thaler and Sunstein (pp.224-246) mention publicity as an important aspect in the determination of the boundaries of nudging. In short, the principle means that the government must not use a nudge that it is not capable or willing to defend in public[7]. For example, a nudge to pay one’s taxes in time is acceptable but a nudge to pay more taxes is not. At the same time, tax administrations should be transparent about the goals of (intended) nudges. After all, if taxpayers are aware of the kind of nudges they are being subjected to, they can more easily make a different choice. Whenever nudges or other behavioural interventions are used, that use must be verifiable – especially given the rise of available information for taxpayers. The tax administration ought to be transparent about these interventions, even if it affects the effectiveness of said intervention.

B.      Taxpayer test

After the publicity test, it is important to test whether the nudge leads to an outcome that is desirable from the perspective of the taxpayer. After all, nudging can be used in the context of taxation to help and support taxpayers in making decisions that are best for them. The standard for this determination is an analogy to New Year’s resolutions[8]. Such a test allows you to determine for whom a certain choice is desirable. As a New Year’s resolution, most people do not decide to start smoking or to drink more alcohol. Likewise, taxpayers plan to pay their taxes in time to avoid penalties but do not plan to start using the government’s official digital mail service in the coming year. For some taxpayers, a digital mail service is indeed a good choice, but this does not mean that it is the right choice for all taxpayers.

C.      Autonomy test

As mentioned before, Thaler and Sunstein argue that nudges must not violate the freedom of choice – the autonomy must be left intact (see section 2 supra). This means that there must be a choice for the taxpayer and the alternate choice must not be too difficult. Additionally, a made decision should be easily reversible. This part of the definition of nudging offers, in my opinion, a sufficient protection against behavioural manipulation.

6.       Conclusion

The idea that the government uses insights from behavioural science to steer taxpayer behaviour will likely be a fact of life in the future. Therefore, it is essential that tax administrations that stimulate the subconscious of taxpayers to enforce certain behaviour are well aware of the limits of their actions. When nudging is used to improve the tax administration’s service and communication, this is a good development. The current Dutch tax law and the principles of good governance, however, offer insufficient protection against the dangers of other types of nudges. This article mentions three tests that can be used to judge whether the taxpayer’s autonomy is being affected and whether a specific legal standard is desirable.

[1] R.H. Thaler and C.R. Sunstein, Nudge, Improving decisions about health, wealth and happiness, London, Penguin Books, 2009, p. 6.

[2] R.H. Thaler and C.R. Sunstein, Nudge, Improving decisions about health, wealth and happiness, London, Penguin Books, 2009, p. 6.

[3] R.H. Thaler and C.R. Sunstein, Nudge, Improving decisions about health, wealth and happiness, London, Penguin Books, 2009, p.109.

[4] R.H. Thaler and C.R. Sunstein, Nudge, Improving decisions about health, wealth and happiness, London, Penguin Books, 2009, Chapter 15.

[5] P.H.J. Essers, Nudging, Weekblad voor Fiscaal Recht (WFR) 2017/105.

[6] J.M. van Rooij and D. Geurts, Hoe de belastingheffer de mens ontdekt. Een praktijkperspectief op de relatie tussen belastingregels en belastinggedrag, RegelMaat 2014 (29)2, p. 90.

[7] R.H. Thaler and C.R. Sunstein, Nudge, Improving decisions about health, wealth and happiness, London, Penguin Books, 2009, p.244-246.

[8] J.M. Van Rooij and D. Geurts, Hoe de belastingheffer de mens ontdekt. Een praktijkperspectief op de relatie tussen belastingregels en belastinggedrag, RegelMaat 2014 (29)2; R.H. Thaler and C.R. Sunstein, Nudge, Improving decisions about health, wealth and happiness, London, Penguin Books, 2009.