This article belongs to the debate » Choice Architecture in Democracies
07 Januar 2015

Constitutional Limits to Paternalistic Nudging in Germany

Nudges with paternalistic aims pose special legal problems in liberal States. Surprisingly, the discussion on regulation-by-nudging has not focused on the constitutional limits to nudging. Although the property rights of firms potentially infringed by nudging measures are dealt with in the literature and by (international) courts (e.g. the tobacco cases), the potential infringement of the rights of those being nudged is neglected. But judges may at one point be confronted with a nudge regulation challenged by the individuals being nudged; and even before reaching a court, the legality of nudging should be scrutinised by legislators.

I explore the legal limits of paternalistic nudging under the German Constitution, especially the right to freedom of action and self-determination under Art. 2 (1) German Basic Law, judging different types of nudges by the proportionality principle. At issue is the question of how much paternalistic nudging and what types of paternalistic nudges the fundamental rights protection in Germany permits. The analysis can be applied, mutatis mutandis, to non-paternalistic nudges (targeting externalities and public goods/bads).

1. The Tool Box of Paternalistic Measures

The legal system has different instruments at its disposal to regulate with a paternalistic intention: (1) it can manipulate choice invisibly (choice manipulation), including invisible default rules, (2) it can command or prohibit certain choices explicitly (prohibitions), or (3) it can support choice (choice support). Supportive instruments can take the form of either a measure providing facilitation for an isolated decision by an individual, or a communicative process. An isolated choice facilitation instrument is limited to slowing down the decision of an isolated individual during the decision-making process. It is a stop sign aiming at inducing auto-deliberation. From an autonomy point of view, enabling learning is crucial. These instruments supporting choice can be put in place by different measures, e.g. information, self-engagement help (as with casinos), (open) default rules and regret mechanisms. An initial approach is to give individuals full, accessible information. However, individuals usually communicate before they take important decisions – and arguably only the latter should be the target of regulation in liberal States. To the best of my knowledge, communicative choice facilitation instruments have been largely ignored by legal scholars and behavioural economists addressing paternalism. A communication-oriented regulation of biases establishes rules in support of rationality and autonomy, and improves the basis for individual choice through communication and deliberation. This kind of regulation provides especially strong stop signs, capable of influencing preferences and cognition while guaranteeing autonomy, e.g. patient commissions in hospitals.

2. A Proportionality Assessment

The German constitutional scope of rights protection, is, in contrast to European protection of human rights, without any gaps – any state measure can be assessed under Art. 2 (1) GG (right to freedom of action and self-determination). This right has a very broad scope of protection and is subsidiary to other, more specific rights. It includes the right to use the freedom in a way which contradicts (in the eyes of third persons) the well-understood interests of the individual, including the “freedom to illness”.[2] Any nudging measure thus falls under the scope of protection of the right. Nudging regulation can be counted as an interference with the right, since a broad concept of ‘interference’, which may even include indirect, non-targeted effects on an individual’s sphere of liberty emanating from governmental action is used; a fortiori targeted, even if invisible nudges must count as an interference.

Can paternalistic nudges be constitutionally justified? The proportionality assessment consists of a four-pronged test (for laws which are necessary in my view as a basis for paternalistic nudging measures): legitimate aim of the measure, suitability of the measure and least restrictive means tests and proportionality strictu sensu. Paternalism targeting preferences (autonomy) needs to be subjected to greater scrutiny than paternalism targeting cognition (rationality). If, taking Mill´s bridge example, the wanderer wants to cross the decrepit bridge after being informed (correcting cognition) about the danger, that preference needs to be respected.

The first test is the legitimate aim of a nudge. Liberal paternalism remains unclear whether it´s goal is some kind of “objective” welfare (e.g. a healthy body), full rationality or autonomy. In liberal societies, I submit, the goal of a nudge must be the enablement of autonomy or rationality, it cannot be an “objective” idea of welfare of individuals. Other voices even deny this, holding that any sort of paternalism is unjustified.[3] The GFCC has been firm in protecting freedom of will against commonly assumed preferences and rationality.[4] Thus, nudges targeting only the welfare of individuals without due account of their autonomy are problematic under German doctrine. The purported aim of a measure was not always well scrutinized by the GFCC and it accepted third party effects (e.g. the protection of the community of the insured), choosing a slippery slope.

Invisible and thus manipulating nudges may be suitable if welfare is the goal but they are not fostering autonomy or rationality and even inhibit learning processes. Arguably, invisible nudges, including invisible default rules are thus unsuitable in contrast to choice supporting nudges. The regulation has to choose the least intrusive means – measured in terms of autonomy. This implies that open information, educational campaigns etc. always trump hidden nudging or manipulation. Isolated choice support is less intrusive than communicative choice support which in turn is less intrusive than a prohibition which in turn is less intrusive than invisible manipulation.

This article draws on other articles of mine: ‚Begrenzte Rationalität und Paternalismusgefahr. Das Prinzip des schonendsten Paternalismus‘ in M Anderheiden et al (eds), Paternalismus und Recht (Tübingen, Mohr Siebeck, 2006), ibid ‚Das deliberative Element juristischer Verfahren als Instrument zur Überwindung nachteiliger Verhaltensanomalien. Ein Plädoyer für die Einbeziehung diskursiver Elemente in die Verhaltensökonomik des Rechts‘ in C Engel et al. (eds), Recht und Verhalten Beiträge zu Behavioral Law and Economics (Tübingen, Siebeck/Mohr, 2007), ibid. ‚Judge the Nudge: In Search of the Legal Limits of Paternalistic Nudging in the EU‘ in A Alemanno and AL Sibony (eds), Nudging and the Law. What Can EU Law Learn from Behavioural Sciences? (Oxford, Hart, forthcoming 2014).

[2] GFCC Decision of October, 7 1981, 2 BvR 1194/80, para. 45.

[3] C Hillgruber, Der Schutz des Menschen vor sich selbst (1992); K Möller, Paternalismus und Persönlichkeitsrecht (2005); K Fischer, Die Zulässigkeit aufgedrängten staatlichen Schutzes vor Selbstschädigung (1997).

[4] GFCC Decision of March, 23 2011, 2 BvR 882/09, para. 55: There is „keine ‚Vernunfthoheit‘ staatlicher Organe über den Grundrechtsträger dergestalt, dass dessen Wille allein deshalb beiseite gesetzt werden dürfte, weil er von durchschnittlichen Präferenzen abweicht oder aus der Außensicht unvernünftig erscheint“.

SUGGESTED CITATION  van Aaken, Anne: Constitutional Limits to Paternalistic Nudging in Germany, VerfBlog, 2015/1/07,, DOI: 10.17176/20181005-151036-0.

One Comment

  1. Martin Holterman So 11 Jan 2015 at 03:05 - Reply

    Interesting. I would have thought that the whole point of nudging is that you don’t formally limit anyone’s freedom to do anything, so that it’s not in need of constitutional justification.

Leave A Comment


1. We welcome your comments but you do so as our guest. Please note that we will exercise our property rights to make sure that Verfassungsblog remains a safe and attractive place for everyone. Your comment will not appear immediately but will be moderated by us. Just as with posts, we make a choice. That means not all submitted comments will be published.

2. We expect comments to be matter-of-fact, on-topic and free of sarcasm, innuendo and ad personam arguments.

3. Racist, sexist and otherwise discriminatory comments will not be published.

4. Comments under pseudonym are allowed but a valid email address is obligatory. The use of more than one pseudonym is not allowed.