Content uploaded by Malte Dold
Author content
All content in this area was uploaded by Malte Dold on Jul 17, 2024
Content may be subject to copyright.
1
Individual Autonomy and Public Deliberation in
Behavioral Public Policy
Nathanaël Colin-Jaeger
1
& Malte Dold
2
July 17, 2024
Abstract: The most successful concept in behavioral public policy (BPP) is nudging, which involves
altering choice architecture to leverage people’s biases and heuristics to promote welfare-improving
behaviors. However, in recent years, nudging has faced criticism. This article addresses a specific
critique: while nudges may enhance welfare, they often fail to promote autonomy. Several authors have
raised this concern, yet there is no unified definition of autonomy in BPP. This article delves into the
various meanings of autonomy in the BPP literature: freedom of choice, agency, and self-constitution.
It focuses on autonomy as self-constitution, which acknowledges instrumental rationality but also
considers substantive rationality, i.e., people’s ability to reason about their goals, aspirations, and
identities. The article explores epistemic, normative, and psychological challenges of autonomy as self-
constitution and suggests that public deliberation in mini-publics could mitigate some of these
challenges. Moreover, it emphasizes that an autonomy-centric BPP should shift its focus from reframing
individual choice situations (i-frame interventions) to enabling public deliberation about institutional
choices (s-frame interventions).
Keywords: Autonomy, Behavioral Public Policy, Deliberation, Paternalism, Reasoning
1
American University of Paris, France, ncolinjaeger@aup.edu.
2
Pomona College, USA, malte.dold@pomona.edu.
2
1. Introduction
Behavioral Public Policy (BPP) is a relatively young field. The International Behavioral Public
Policy Association (IBPPA) held its third annual meeting at the University of Cambridge in
June 2024. The first issue of its leading field journal, Behavioral Public Policy, was published
in 2017. Notwithstanding its young age, BPP has had a substantial impact on policy practice,
as is evidenced in the success of the Behavioral Insights Teams (BIT) and similar organizations
around the world (OECD, 2017).
A core contributor to BPP’s ongoing success is the idea of nudging which is explicitly
built on behavioral insights to help people overcome problems of bounded rationality or
willpower (Thaler and Sunstein, 2008). Its proponents describe a nudge as “any aspect of the
choice architecture that alters people’s behavior predictably without forbidding any options or
significantly changing their economic incentives.” (Thaler and Sunstein, 2008, p. 6). Nudges
straddle the line between libertarianism and paternalism. They uphold individual autonomy by
ensuring that individuals can easily opt for the non-nudged option if they so wish.
Simultaneously, they adopt a paternalistic stance, as they involve choice architects guiding
individuals towards welfare-improving outcomes by reshaping the decision-making
environment. Thaler and Sunstein provide a limited discussion of what constitutes individual
welfare. Yet, they emphasize the importance of respecting the subjective goals of individuals
when they say that nudges aim to “make choosers better off, as judged by themselves” (Thaler
and Sunstein, 2008, p. 5).
As is the case with other successful policy ideas, the idea of nudging has been defended
and criticized by others.
3
This article does not rehash this debate but focus on a strand in the
BPP literature that has advocated for a shift away from the narrow welfarist focus of nudging
toward a more autonomy-oriented interpretation of BPP (Banerjee et al., 2024; Hargreaves
Heap, 2023; Dold and Lewis, 2023). This shift has been motivated by the observation that
nudges often leverage people’s cognitive biases to prompt behavioral shifts. Instead of
providing reasons for choosing the nudged option, nudges exploit the same decision-making
flaws that led to the initial biased choice. Altering the default in retirement savings plans is a
prime example (Beshears et al., 2009). Here, the nudge doesn’t make people more competent
in financial matters or empower them to reflect on their preferences; rather, the nudge exploits
people’s tendency to go with the status quo, thereby steering them toward higher savings rates.
The individual being nudged often remains unaware of the intervention’s purpose. Choice
architects reshape the decision environment from the top down, aiming to steer individuals
towards options deemed best by the architect. Consequently, nudges often “leave citizens ‘in
the dark’, incapable of internalizing and owning the process of behavior change” (Banerjee et
al., 2024, p. 1). While the autonomy-oriented strand in BPP is united in this critique of the
nudging approach, it has not come up with a unified understanding of what constitutes
autonomy (Vugts et al., 2020). In Section 2, we discuss the various understandings of autonomy
proposed in the literature. We focus on a more expansive understanding of autonomy as self-
constitution and the idea that behavioral insights can contribute to a discussion of interventions
that enhance people’s capability to reflect on their identity-shaping choices. A core goal of BPP
that focuses on autonomy as self-constitution is to help people become aware of situational and
social factors shaping their belief and preference formation processes and, in doing so, increase
their capacity to form intentions in a self-reflective manner and act freely on them. We admit in
3
For a summary of the defense, see Schmidt and Engelen (2020); for a summary of the critique, see Rizzo and
Whitman (2020).
3
Section 3 that this proposal for a ‘thicker’ notion of autonomy in BPP is not without its
problems. In particular, it faces epistemic, normative, and psychological challenges.
Epistemically, a concern is the question whether policy designers can gain knowledge of to the
kind of identity formation processes individuals favor. Normatively, a concern is whether an
autonomy-oriented BPP can avoid falling back onto a form of autonomy paternalism that might
reflect the goals of academically trained policy experts but miss people’s perspectives.
Psychologically, a concern is whether individuals have the cognitive means and motivation to
engage in transpositional exercises to form their identities in a self-reflective manner.
In Section 4, we discuss ways in which public deliberation can help mitigate these
challenges. Public deliberation follows the concept of “talking as a decision procedure”
(Schauer, 1999); it transcends mere information exchange (although information pooling is
undeniably important to the deliberative process), highlighting instead discussion, disputation,
and the transformation of people’s views (Elster, 1986/2010). If properly designed, public
deliberation can enhance participants’ reasoning abilities by exposing them to diverse beliefs
and evaluative standards, thereby making them aware of their positionality (Sen, 2009; Mercier
and Sperber, 2011; Dold, 2024). In essence, public deliberation can enhance individuals’
autonomy and aid them in their processes of self-constitution. This helps tackle the
psychological challenge. Moreover, public deliberation can take various forms, such as mini-
publics, citizen juries, deliberative polling, or consensus conferences. In this article, we do not
advocate for a specific institutional form of public deliberation but rather underscore its
constructive aspects to also tackle the epistemic and normative challenges mentioned before.
By involving citizens in the policy design process, public deliberation alleviates the epistemic
burden on experts in the context of BPP. Furthermore, by relying on the buy-in of affected
citizens, it addresses the normative concern of autonomy paternalism. We acknowledge that
public deliberation in the form of citizen participation in public policy processes encounters
counterarguments. We draw upon the empirical literature on public deliberation to address
some of them and advocate for a well-designed deliberative process. Section 5 concludes.
2. Autonomy in BPP: Freedom of Choice, Agency, and Self-Constitution
While policy proposals in BPP typically aim at increasing people’s welfare – e.g., by nudging
them to save more or eat fewer fatty foods – they typically also rely to a large degree on the
idea of autonomy: the consumer is supposed to make the ultimate decision and is free to ignore
the nudge.
4
Both the proponents of nudging, such as Thaler and Sunstein (2008), and
contemporary critics who favor different behaviorally-informed interventions, such as Banerjee
et al. (2024), are advocating for policies that preserve or promote personal autonomy. However,
autonomy is a broad church. Following Vugts et al. (2020), one can distinguish at least three
conceptions of autonomy in BPP: autonomy as freedom of choice, autonomy as agency, and
autonomy as self-constitution. Most debates in BPP revolve around the first and second
conception, while we argue in this article that the third conception ought to be properly taken
into account. We argue that understanding autonomy as self-constitution requires BPP scholars
to focus on the institutional and political conditions for autonomy (s-frame), and not only on
the individual conditions for welfare-improving choices (i-frame).
5
If autonomy as self-
constitution is valuable, an institutional turn in BPP is warranted. But before discussing
institutional implications, let us first depict the three conceptions of autonomy.
4
There are exceptions, which directly endorse the paternalistic view, such as Conly (2013).
5
On the distinction between i-frame and s-frame policy interventions in the context of BPP, see Chater and
Loewenstein (2023).
4
The first conception of autonomy is freedom of choice. This is the main conception of
autonomy that we find in BPP, in particular in the early discussions of nudging (Thaler and
Sunstein, 2008) and its subsequent refinements in response to critics (e.g., Sunstein, 2015, 2016,
2017). The main idea is that behavioral policies should respect individuals’ freedom to choose
among different options, such that even when people are nudged, it is possible for them to
switch at no or a low cost to another option. In other words, nudges do not take options off the
table or significantly alter their relative prices. Nudges and other interventions are seen as
legitimate if they are not coercive and people are left free to go their own way.
The typical cases where nudges are invoked are cases of weakness of the will – when
individuals wish to lose weight but succumb to their urgent needs when they see chocolate at
the cafeteria cashier – or lack of information – when they have to choose a retirement plan but
get lost in the complexity of the choice menu. In these cases, nudges can help individuals to act
according to their goals (in this case, losing weight or saving effectively for retirement). Nudges
include opting people automatically into employer-matched retirement savings plans or
rearranging items in the cafeteria to make them less visible, thus helping people to choose
healthy options. The hope is in both cases that individuals would accept the nudge since it steers
them toward their ‘true’ preferences (losing weight, having a good retirement plan) against their
seemingly biased preferences (succumbing to temptation, saving less).
6
Kuyer and Gordijn (2023) argue that most discussions on the ethics of nudging (and
BPP more broadly) occur within this conception of autonomy as freedom of choice.
Interestingly, it is not just the proponents of nudging who subscribe to this understanding of
autonomy but also some of its critics. For example, Sugden (2018) criticizes Thaler and
Sunstein’s approach based on the argument that they do not provide a realistic psychological
model of what constitutes ‘true’ preferences and, in doing so, do not take the inherent context-
dependent nature of preferences seriously. Nevertheless, Sugden argues that autonomy should
be understood as the freedom to choose among options, such that autonomy is increased when
the opportunity set is enlarged.
The second conception of autonomy in BPP acknowledges that freedom of choice is
undoubtedly valuable (in the sense that people care about it), but not self-standing. It should be
complemented with a focus on agency understood as decision competence (Dold, 2023). This
view attracted considerable attention over the past years, particularly by proponents of the
boosting approach (Grüne-Yanoff and Hertwig, 2016; Hertwig and Grüne-Yanoff, 2017;
Banerjee et al., 2024). One idea underlying boosting is that BPP should not just be concerned
with the number of opportunities people are confronted with, but also their capacity to choose
competently among them (Dold and Rizzo, 2021). In this second conception of autonomy as
agency, autonomy requires not only having options and being free from coercion but having
the internal capacity to reflect on what options are best to achieve a given personal goal (Vugts
et al., 2020). A further difference with the first conception of autonomy is that this conception
concerns the internal cognitive processes of agents, as it requires BPP to increase people’s
mental capacities to reason instrumentally and choose competently (McKenzie et al., 2018;
Dold and Lewis, 2023).
Proponents of this notion of autonomy criticize nudges for violating autonomy as
agency because they (largely) target or exploit shallow cognitive processes – the fast, automatic,
and intuitive type 1 processes (Kahneman, 2011) – to alter people’s behavior but do not try to
improve their cognitive processes by appealing to rational and effortful type 2 thinking (Saghai,
6
Sunstein (2015, 2016) extends this approach, stating the importance of individualizing such interventions to
realize personal autonomy rather than imposing a preconceived conception of the good on individuals.
5
2013). Nudges largely fail to treat people as reasoning subjects, diminish the control over their
choice process, and ultimately undermine their agency (Hausman and Welch, 2010; Lanzing,
2019; Noggle, 2018). According to proponents of the second notion of autonomy, specific
BPPs, such as debiasing and boosts, should help people choose the ‘right’ means for a given
end by helping them “build cognitive capability and motivation through the regular activation
of reflective processes” (Banerjee et al., 2024).
Boosts modify people’s behavior by making them more competent regarding a specific
problem, typically by teaching them the application of simple decision rules (Grüne-Yanoff and
Hertwig, 2016; Grüne-Yanoff, 2021). They consist, for instance, of special training regarding
issues individuals face: self-control training to avoid procrastination or training in fast-and-
frugal decision trees to break down complex decisions, e.g., in the medical context (Gigerenzer,
2024). Boosts encompass domain-specific training (e.g., how to exercise and remain motivated)
or general skills (e.g., statistical literacy for various decisions). Boosts do not bypass people’s
system 2 thinking because training ought to be accepted and chosen by individuals. Boosted
individuals select themselves into training activities that enable them to achieve their goals (e.g.,
becoming more reliable, more statistically trained, etc.).
Debiasing is a second possibility to increase autonomy as agency, consisting of
procedures not to make biases disappear – which is probably impossible – but to make them
explicit (Banerjee et al. 2024). A typical example is food labeling and mixed framing. A brand
could sell their food with a 75% fat-free label, which could be accompanied by a 25% fat label,
thereby helping people to make a better decision. More expansive debiasing techniques work
in institutional or organizational settings where time-costly interventions, such as role-play, are
available. In a role-play scenario, different positions on a topic can be predetermined to avoid
false unanimity or fallacies. If some individuals are required to play a pessimistic role on a
project, this can help counterbalance optimism bias. These techniques can also be attainable for
individuals. For example, when confronted with an important choice, people can justify their
decisions by writing up pro and contra arguments on a given topic. In these situations, taking
the time to make reasons explicit allows people to realize biases in their intuitive approach to a
decision problem. Enhancing autonomy via boosts and debiasing techniques complements the
aforementioned freedom of choice view of autonomy. However, it still regards the primary
objective of BPP as bolstering individuals’ instrumental reasoning – namely, selecting the most
appropriate means to achieve their given ends. Within this context, the biggest threat to
instrumental reasoning is subtle manipulation (e.g., in the form of sublime advertisements) that
leads people to choose suboptimal means for their given ends (Vugts et al., 2020).
We propose to go a step further defending a third account of autonomy as self-
constitution. According to the philosopher Christine Korsgaard (2009), self-constitution
requires conceiving yourself as the cause of your actions. Korsgaard (2009, p. xi) claims: “For
an action is a movement attributable to an agent as its author, and that means that whenever you
choose an action—whenever you take control of your own movements—you are constituting
yourself as the author of that action, and so you are deciding who to be.” Seeing your action as
emanating from yourself means recognizing yourself as a cause rather than the action as the
production of something supposedly within me (such as a counterfactual ‘true’ preference) or
in work through me (such as an external, manipulative influence). BPP interventions like
nudges typically target some given desirable goals but do not necessarily respect agency as self-
constitution; they exploit my biases and in this sense work through me. Moreover, assuming
hypothetical, counterfactual preferences I might not have right now but will be nudged towards,
nudges target something supposedly within me without necessarily recognizing myself as a
whole (Korsgaard 2009, 18).
In contrast, autonomy as self-constitution entails the production of oneself through
active decisions. We shape our identity through our choices. In this sense, autonomy does not
6
only entail the capacity to choose what means we want to pursue for given ends (instrumental
reasoning), but also the capacity to decide which goals we want to pursue in the first place and
which person we want to become over time (substantive reasoning). This notion of autonomy
does not imply that individuals need to be free from all external influences or to choose every
action consciously to be autonomous. Agents can recruit routines through conscious intentions,
leading to non-conscious behavior (Clarke, 2010; Schlosser, 2019). In this sense, autonomy as
self-constitution is compatible with non-conscious behavior that is the consequences of some
prior intentional and deliberate decisions. Sociology and social psychology have taught us that
individuals are not free from external influences when making choices. Yet, self-constitution
means that people are able to reflect on processes of preference and belief formation and the
ways they are affected by social and psychological influences. In this regard, the ideas of
indoctrination or mental conditioning are seen as threats to autonomy: they undermine people’s
capacity to be the authors of their own lives since they make it difficult for them to be aware of
where their preferences and beliefs come from (Vugts et el., 2020; Dold, 2024).
In the existing BPP toolkit, autonomy as self-constitution is addressed by the idea of
nudge+ (Banerjee and John, 2021). Nudge+ mixes the idea of traditional Thaler-Sunstein
nudges with deliberation about their use and the goals of citizens. Conscious reflection is used
to overcome the potentially harmful effects of nudges (e.g., reactance or subtle manipulation).
For Banerjee and John, nudge+ can take many forms depending on the coupling of nudges with
reflection (simultaneous or sequential). Deliberation plays an essential role for nudge+: they
allow people to reflect not only on an institutional intervention but also on their preferences and
what nudging means for their identity formation processes. Other BPP interventions that go
beyond the idea of nudge+ would be s-frame institutional reforms that expose people to multiple
viewpoints and challenge their preconceived notions on certain topics through education,
experiments in living and public deliberation (Dold, 2024). Education plays a crucial role in
contributing to discussions about the core elements of our identity, but also for understanding
that “diversities can take many distinct forms and that we have to use our reasoning to decide
how to see ourselves.” (Sen 2003). The concept of experiments in living (Mill 1859/2003)
highlights the significance of encountering diverse life experiences. This can be achieved either
through unintentional exposure to a variety of lifestyles or by deliberately immersing oneself
in different cultures, practices, and political systems. Due to space constraints, this article does
not delve deeper into education and experiments in living as important means to foster
autonomy as self-constitution but will instead focus on the central role of public deliberation in
the next section. Table 1 gives an overview of the dimensions of the three conceptions of
autonomy discussed so far.
Table 1: Conceptions of autonomy in BPP
Autonomy I
Freedom of Choice
Autonomy II
Agency
Autonomy III
Self-Constitution
Main
Target
Behavioral Outcomes
Instrumental Reasoning (IR)
Substantive Reasoning (SR)
Task of
BPP
Steer people’s behavior
in a certain predefined
welfarist direction
Increase competence, i.e.,
capacity to choose means
𝑚𝑖"from {𝑚1… 𝑚𝑛} for given
goal 𝑥
Increase agentic capability, i.e.,
ability to choose goal
𝑥𝑖"
from
{𝑥1… 𝑥𝑛}
that is self-endorsed and
authentic
7
Threats
Coercion
Subtle manipulation
Indoctrination and mental
conditioning
Policy
Nudges, opportunities
Boosts, debiasing
Nudge+, public deliberation,
experiments in living
Level
i-frame intervention
i-frame intervention
i- and s-frame interventions
3. The Threefold Challenge of Autonomy as Self-Constitution
The idea of self-constitution is appealing to those who contend that emancipation and
experimentation are valuable components of a fulfilled life (Dold and Lewis, 2023). Autonomy
as self-constitution means that individuals shape themselves through choices rather than merely
satisfying some given goals. However, in this conception of autonomy as self-constitution it
might not be straightforward for external observers to gain knowledge of the type of social
environments (e.g., pluralistic-open vs. homogenous-closed) individuals favor for their identity
formation processes. Moreover, this conception encounters the problem of justifying what kinds
of identity formation processes should be supported and for whom; autonomy as self-
constitution risks reintroducing social planner-like conceptions of the good life that might be
normatively problematic. Finally, even if these two challenges can be resolved, individuals
might be overwhelmed by the cognitive demands of substantive reasoning and the idea of
exposing their identity and goals to critical scrutiny. Consequently, there is (at least) a threefold
challenge of using autonomy as self-constitution in BPP:
(i.) Epistemic challenge: How can policymakers gain knowledge of the social
environments conducive to the identities individuals aspire to hold?
(ii.) Normative challenge: How can policymakers avoid reverting to paternalism
when fostering certain processes of identity formation over others?
(iii.) Psychological challenge: Do individuals have the cognitive means to engage in
transpositional exercises to form their identities in a self-reflective manner?
Considering these challenges, it appears that an autonomy-oriented BPP encounters similar
issues as the standard welfarist justification for BPP. In the latter case, it is problematic if people
are nudged into behavioral outcomes that do not respect individuals’ actual choices. In the
former case, guiding the process of self-constitution might be equally problematic for the same
reason: direct interventions might be illegitimate, and meta-interventions on the social
environments that provide the background against which individuals form their identities can
be illegitimate because they unduly steer or restrict the range of possible identities an agent may
have access to. The process view of the self (which is inherent to autonomy as self-constitution)
presents a specific challenge: if neither the policymaker standpoint nor any particular individual
standpoint seems relevant, from which standpoint may we overcome this problem? Neither the
“I” perspective of the chooser (due to context-dependence and cognitive limitations) nor the
“She” perspective of the policymaker (due to the epistemic and normative problems stated
above) offers a satisfying solution.
A promising attempt to tackle the epistemic and the normative challenges has been made
by Lecouteux and Mitrouchev (2023) in an article defending the “view from manywhere”,
which is deemed normatively superior to the “view from nowhere” (the “She” perspective of
the planner) and the “view from somewhere” (the “I” perspective of the choosing individuals).
8
The view from manywhere is based on an intra-subjective dialogical account of the different
identities of a given individual. Following Lecouteux and Mitrouchev (2023), an individual A
may generate a set of context-dependent identities with possibly different and competing
interests (think of a person’s identity A' in the context when she is drinking during the evening
and her identity A'' in the context of a sober but slightly hung-over context the next
morning). Lecouteux and Mitrouchev argue that what should have normative status is not the
choice of one identity over the others (e.g., A'' over A'), but what we may call the deliberative
process that generates preferences through the open-ended confrontation of interests and
perspectives.
We sympathize with this approach, as the idea of achieving a view from manywhere has
a long tradition in liberal thought from Adam Smith’s idea of the impartial spectator (Smith,
1859) to Amartya Sen’s account of transpositional reasoning (Sen, 2009). Nevertheless, we
have several concerns with the approach – at least when it is understood narrowly. First, it fails
to address the aforementioned psychological challenge because it imposes a heavy cognitive
burden on individuals, requiring them to undertake the complex deliberative process of
reconciling different perspectives within their multiple identities on their own. Second, even if
individuals are psychologically capable of generating a view from manywhere, it can still be
significantly constrained by imagination, experience, social networks, and other factors that
may hinder individuals from achieving the normative goal of comprehending a broader range
of influences on their behavior and preference formation. The Lecouteux-Mitrouchev approach
remains mostly at the individual level as an inner deliberation, while self-constitution does not
limit itself to strictly individual deliberation; in fact, self-constitution occurs reflexively within
social and institutional contexts. In the next section, we argue that public deliberation helps
alleviate these concerns.
4. Public Deliberation: Autonomy Through Self-Government
By public deliberation we refer to the idea that “people come together, based on equal status
and mutual respect, to discuss political issues they face and, based on those discussions, decide
on the policies that will affect their lives” (Bächtiger et al., 2018, p. 2). Public deliberation
theorists emphasize that deliberation is above all a mode of self-governance; they consider it
the privileged mode of public decision-making for members of a society who value equality
and justifiability.
7
Traditionally, public deliberation in a representative democracy manifests
itself through discussions held in assemblies or within judicial proceedings. Yet, recent
literature on deliberation emphasizes alternative, more experimental forms of deliberation, such
as mini-publics (Niemeyer, 2011; Fishkin, 2018; Landemore, 2020). Mini-publics can take
different forms such as citizen juries (randomly selected citizens who deliberate on a specific
issue, e.g., how to make the city more inclusive), deliberative polling (a statistically
representative group of citizens randomly selected to participate in a moderated discussion on
particular issues and propose recommendations, e.g., citizens conventions like the French
Citizen Convention for Climate), or consensus convention (when a group of citizens and experts
meet to discuss a particular issue to reach a consensus). In this article, we do not advocate for
a specific institutional form of mini-publics. However, we want to note that public deliberation
can and should happen in diverse institutional forms and at various levels (federal, state, and
local).
A significant challenge of current BPPs, such as nudges and boosts, is that they are
designed by experts who are supposed to nudge or boost people in specific directions (based on
7
See, for instance, Habermas (1996), Benhabib (1996), Cohen and Sabel (1997), Gutman and Thompson (1997),
Bohman (2004).
9
what experts imagine to be in the people’s interest – be it welfare or autonomy). This implies
that experts need to know a lot about people’s preferences, including possible trade-offs, such
as between people’s welfare and autonomy concerns. Nudges might increase welfare through
automatization and unconscious procedures and boost interventions might enhance people’s
autonomy by teaching them simple heuristics that improve the quality of their instrumental
reasoning. But how can BPP experts gain knowledge about such trade-offs and avoid falling
back onto a form auf welfare or autonomy paternalism? We argue that deliberation and the
inclusion of citizens in designing the policies can help meet these concerns.
There are three core functions of deliberative mechanisms displayed in the empirical
literature (Niemeyer et al., 2023): the transparency and accountability function, the reduction
of cognitive burden function, and the enhancement of cognitive abilities function. The first
function of deliberation is to increase transparency and accountability which allows us to take
up the normative challenge raised above. The reason-giving requirement of deliberation is what
allows communication between different individuals with different perspectives and interests
(Habermas, 1996; Gutman and Thompson, 1997; Bohman, 2004; Landemore, 2011). This ties
into the second function which is the reduction of the epistemic burden on behalf of the
policymakers because it allows individuals to participate in the policy-making process and
signal the trade-offs they want to accomplish. Since deliberative forums (typically) produce a
set of concrete and actionable policies, the epistemic burden for policymakers is decreased: they
can look at the outcomes of deliberative processes and do not need to infer values or aspirations
from observed market behavior or noisy surveys. This allows us to take up the epistemic
challenge raised above. The third crucial function is that participation in deliberative forums
enhances people’s reasoning capacities. Deliberation often has a transformative effect on
people’s beliefs due to the discovery of information as well as on their evaluative standards due
to people’s exposure to other people’s views and values (Knight and Johnson, 2011); by
recognizing their own individual positionality people broaden their views and become aware of
factors that influence their substantive reasoning (Habermas 1996). In this sense, people do not
only learn new information based on the diversity of public beliefs but are led to develop new
preferences and values based on the diversity of evaluative standards.
8
Argumentation makes
us aware of hidden assumptions and influences and can thus modify our views. If we take these
insights seriously, we see how public deliberation answers better than individual deliberation
the psychological challenge mentioned in the previous section. The empirical literature on
public deliberation suggests indeed that participation in public forums such as mini-publics has
a transformative effect that facilitates the view from manywhere, enabling a broader
transpositional view on matters that can contribute to more informed ways of self-constitution
and individual identity formation (Niemeyer 2011; Niemeyer et al 2023). Yet, deliberative
arenas pose new problems for BPP. They may shift the locus of BPP from individual reasoning
to collective discussion (Counterargument I). Moreover, collective deliberation may not
necessarily help individuals improve their cognitive abilities and make better decisions for
themselves (Counterargument II).
Counterargument I
Why should BPP instruments transition from individual reasoning to public deliberation? Isn’t
BPP primarily concerned with individual autonomy and welfare? To counter these points, we
wish to underscore that BPP does not shape individuals’ choice environments in isolation but
has also consequences for the structure of the social environment within which people interact
8
The capacity to learn from diversity is very important and a cornerstone of the epistemic democracy view, as
defended by Estlund (2008), Landemore (2012, 2020), or Goodin and Spiekerman (2018). The position often refers
to the “diversity trumps ability theorem” derived from Hong and Page (2004).
10
and choices take place. Moreover, BPP is public and concerns individuals as citizens as much
as private individuals. Inner deliberation, highlighted by Lecouteux and Mitrouchev (2023),
may be insufficient to assess BPP’s legitimacy and acceptability. This does not imply that
public deliberation should be ‘holistic’, i.e., solely concerned with the welfare of the society as
a whole. It should instead acknowledge that BPP impacts the relations between individuals and
their interdependent identity formation processes as much as their ability to reason about ends
or achieve a given end competently. Thus, we contend that public deliberation is a crucial
component of a conception of autonomy as self-constitution. The process view of autonomy as
self-constitution relies on the notion that for an intervention to be legitimate, it should align
with individuals’ values and aspirations. To achieve autonomy as self-constitution, BPP can
help people navigate the process of identity formation in a transpositional manner, which cannot
be dictated from the outside by a social planner or determined strictly individually. Instead, it
should, at least in some cases, involve a social process of identity formation, i.e., public
deliberation.
Public deliberation possesses the same advantages as the multiple identities model of
Lecouteux and Mitroutchev (2023) but with real individuals representing different framings of
a problem at the same time. Instead of individual deliberation, where one individual weighs
various reasons by imagining different current and future versions of herself, it is collective
deliberation among many individuals with different perspectives that fosters multiple-frame
reasoning (Bermudez, 2020). The relation between individual deliberation and collective
deliberation is not disjunctive. Individual reasoning benefits from the confrontation of views in
public deliberation (Colin-Jaeger et al., 2022). It also helps answer the normative question: how
can we identify a locus of legitimacy if neither the “I” perspective nor the “She” perspective is
satisfying? Thus, the appeal of public deliberation lies in its potential to replace an expert-driven
agenda and democratize BPP (Liskow and Markowits, 2023).
Counterargument II
How can one make sure that deliberation transforms individual views in the “right” direction,
i.e., that it helps them transcend their positionality and avoid producing entrenched or parochial
preferences and beliefs? Many critics argue that individuals are poor reasoners, and deliberation
may produce adverse effects. Rather than enhancing autonomy by making choices more
deliberate, it can create new pervasive influences and forms of social control. The criticisms
are based on the assertion that individuals often possess low levels of information (Lippmann,
1925; Brennan, 2016), the observation of adverse effects of collective deliberation on human
intelligence (Schumpeter, 1943; Sunstein, 2002), or the incapacity of collective deliberation to
improve individual decision-making (Müller, 2019). These results are not surprising given the
number of cognitive biases and reasoning errors discussed in psychology and behavioral
economics (Kahneman 2011). In addition, Banerjee et al. (2024) recognize the possibility of
intentional manipulation of deliberative assemblies by third parties. Considering these
concerns, it seems unclear at best whether deliberation is the appropriate route to enhancing the
legitimacy of BPP interventions. Nevertheless, recent empirical studies on mini-publics offer
good arguments to counter the claim that deliberation cannot, in principle, justify the positive
transformative claim made above.
9
What the critics show us is that attention should be paid to
the context and design of deliberation. Two primary arguments can be formulated in response
to the critics of deliberation.
9
That is the claim that it increases people’s agentic capacity of transpositional reasoning. On this point, see Dold
(2024).
11
First, most empirical works casting doubts on public deliberation studies individual
reasoning: “Much political science survey research demonstrates citizen incompetence in terms
of solitary reasoning. Many pertinent psychological experiments involve decontextualized tasks
with no interaction and no supportive environment providing participants with adequate
information. But deliberation involves reasoning together, not individually.” (Niemeyer et al.,
2023, p. 2). As Mercier and Sperber (2011, 2017) argue, human reasoning – understood as a
deliberate process, where one or more premises lead to a conclusion – is best when exercised
in small teams or groups with flat hierarchies. One possible explanation developed by Mercier
and Sperber (2017) is that human reasoning evolved in groups and was selected as a
communication tool rather than an individual task. Group reasoning displays better results than
individual reasoning on most issues, including bias detection, information search, and even
depolarization when people are facing a non-abstract situation of problem-solving. Group
reasoning can sometimes even separate the production of arguments from their evaluation,
which leads to better handling of confirmation biases (Mercier and Sperber, 2011, p. 63).
10
Second, deliberation not only performs better in addressing individual biases, but it can
also be enhanced through proper design. Niemeyer et al. (2023) emphasize the significance of
group building and its impact on the quality and effectiveness of deliberation, extending beyond
the conventional design elements of deliberative forums (such as diverse expertise, participant
representativeness, process facilitation, etc.). Group building encompasses the idea that
members of a group debate about the rules and roles within the group before the actual debate
about substantive issues starts. Groups that display high levels of group building, particularly
when they can develop their norms and rules of decision-making beforehand, share more
information, are less polarized, and face complex issues more efficiently.
For an example where public deliberation could help increase individual autonomy,
consider potential BPPs regarding the consumption of addictive products, such as tobacco,
alcohol, or fatty foods. The BPP literature has suggested many ingenious ways to counter the
self-control issues involved – from nudges (Thaler and Sunstein, 2008) to sin taxes
(O’Donoghue and Rabin, 2006) to motivational boosts (Hertwig and Grüne-Yanoff, 2017).
These suggestions have in common that they involve an expert who designs the nudge,
calibrates the sin tax, and teaches the boost from a “She” perspective. Moreover, in the nudge
and sin tax case, the affected individuals are not forming deliberate intentions themselves
regarding how much tobacco, alcohol, or fatty foods they want to consume (in other words,
their autonomy as self-constitution is sidestepped). In contrast, if the issue of consumption of
addictive products is opened to public deliberative discourse, we allow the individual to be
exposed to multiple positions on the issue. Relevant considerations to inform the discussion
might include scientific findings on the addictive properties and physical consequences of
specific goods and substances, as well as research on the propensity of different types of people
for addictive behavior and the broader societal impact of their actions. Various solutions to the
problem are possible, depending on whether addiction is viewed primarily as a matter of
biochemical propensities, personality traits, material incentives, or socioeconomic structure.
There will be both settled and contested facts concerning the implications of different policies.
A shared representational framework constructed in the process of public deliberation can help
render these aspects mutually intelligible, reduce the epistemic burden on behalf of
10
Mercier and Sperber (2011, 2017) emphasize that we engage in argumentative reasoning within groups, where
one person presents their claim, and others evaluate it. When constructing an argument, our primary focus tends
to be on persuasiveness rather than accuracy. Our main goal is to convince others, often leading us to selectively
use evidence that supports our claim (a form of motivated reasoning). However, when we are the recipients of an
argument, we aim to avoid being misled and thus practice “epistemic vigilance”: we are on the lookout for signs
of inaccuracy to reject the unconvincing arguments of others. This asymmetry benefits group reasoning and
decision-making, particularly when diverse viewpoints are present.
12
policymakers and reduce the psychological burden on behalf of citizens who can form their
goals and aspirations without having to find different evaluative standards themselves.
Furthermore, the participatory component of deliberation counters the normative challenge
from above; it makes citizens design the policies, which increases policy acceptability and
produces long-term compliance, as some studies suggest (Der Does and Jacquet, 2021; Fishkin
et al., 2024).
11
5. Conclusion
This article discussed autonomy as self-constitution, characterized by the competent mastery of
the process of identity production through transpositional reasoning (Section 2). We discussed
epistemic, normative, and psychological challenges of this view of autonomy (Section 3). We
then argued that the “democratization of BPP” through deliberative mini-publics can help
address, or at least mitigate, these challenges. In particular, they can assist individuals in
reaching a view from manywhere by considering opinions and evaluative standards offered by
other people, and investigating what kind of social environment they find conducive to their
processes of identity formation. We have proposed to switch – at least partly – the attention of
BPP discussions from i-frame interventions (such as nudges and boosts) to the question of s-
frame interventions i.e., questions of institutional design of public deliberation.
In doing so, our article introduces a twofold switch away from traditional BPP: from
instrumental reasoning to substantial reasoning, and from interventions at the individual level
to intervention in the institutional framework. If the goal of BPP is to enhance individual’s
autonomy, and autonomy requires not only to be able to have more opportunities but also to be
able to reason to become the author of one’s own life, it requires the capacity to co-construct
one’s environment that will help one think about future preferences and goals.
We suggest emphasizing deliberation to democratize BPP. Admittedly, some
individuals may not want to subject their process of identity formation to collective deliberation.
Autonomy entails the freedom to resist collective decisions. Although this view has sometimes
been used to negate the relevance of deliberation and public reasoning altogether (Sugden,
2018), we deem it important and sympathize with it. Autonomy involves the capacity to reason
independently and avoid blindly obeying others or accepting their collective decisions without
critical evaluation. A proper focus on autonomy in BPP calls for more work at the intersection
of political economy, institutional analysis, and political philosophy to cope with this issue.
Such interdisciplinary work needs to assess how various institutional experiments can enhance
a diversity of perspectives and enable pluralistic processes of identity construction while
avoiding the danger of specifying a unique way of life others must follow.
11
If deliberative individuals choose to put forward routines or delegate the framing of the consumption of addictive
substances to a third party, it is taking the idea of autonomy as self-constitution seriously. The decision not to
reason before every potential encounter with addictive substances and to delegate is made deliberately and
deliberatively. As mentioned in the text above, routinized choice and delegation need not be opposed to autonomy.
13
References
Aligica PD, Tarko V (2012) Polycentricity: From Polanyi to Ostrom, and Beyond. Governance
25:237–262.
Aligica PD, Sabetti F (eds) (2014) Choice, rules and collective action: The Ostroms on the study
of institutions and governance. ECPR Press, Colchester
Bächtiger A, Dryzek JS, Mansbridge J, Warren ME (2018) Deliberative democracy: An
introduction. In: Bächtiger A, Dryzek JS, Warren ME, Mansbridge J (eds) The Oxford
handbook of deliberative democracy. Oxford University Press, Oxford, pp 1–34
Banerjee S, Grüne-Yanoff T, John P, Moseley A (2024) It's time we put agency into
behavioural public policy. Behav Public Policy.
Banerjee S, John P (2021) Nudge plus: Incorporating reflection into behavioral public policy.
Behav Public Policy 1–16
Benhabib S (1996) Democracy and difference. Princeton University Press, Princeton
Bermúdez JL (2020) Frame it again: New tools for rational decision-making. Cambridge
University Press, Cambridge
Beshears J, Choi JJ, Laibson D, Madrian BC (2009) The importance of default options for
retirement saving outcomes. Soc Secur Policy Chang Environ 167–195
Bohman J (2004) Realizing deliberative democracy as a mode of inquiry: Pragmatism, social
facts, and normative theory. J Speculative Philos 18(1):23–43
Blumenthal-Barby JS, Burroughs H (2012) Seeking better health care outcomes: The ethics of
using the 'nudge'. Am J Bioeth 12(2):1–10
Button M (2018) Bounded rationality without bounded democracy: Nudges, democratic
citizenship, and pathways for building civic capacity. Perspect Polit 16(4):1034–1052
Chater N, Loewenstein G (2023) The i-frame and the s-frame: How focusing on individual-
level solutions has led behavioral public policy astray. Behav Brain Sci 46
Colin-Jaeger N, Dold M, Gascoin A (2022) Realism and deliberation in normative political
economy: The fruitful intellectual dialogue between Rawls and Buchanan. In: Badei S,
Griveaux A (eds) The positive and the normative in economics. Routledge, London, chap 6, pp
108–128
Conly S (2013) Against autonomy: Justifying coercive paternalism. Cambridge University
Press, Cambridge
Cudd A (2014) Adaptations to oppression: Preference, autonomy and resistance. In: Oshana M
(ed) Personal autonomy and social oppression. Routledge, New York
14
Delmotte C, Dold M (2022) Dynamic preferences and the behavioral case against sin taxes.
Const Polit Econ 33(1):80–99
Der Does R, Jacquet V (2021) Small-scale deliberation and mass democracy: A systematic
review of the spillover effects of deliberative. Polit Stud 71(1):218–237
Dold M (2023) Individual Agency in Behavioral Public Policy: New Knowledge Problems. RG
Working Paper. DOI: 10.13140/RG.2.2.12235.77603
Dold M (2024) Amartya Sen on Adaptation: Explications and Implications. RG Working Paper.
DOI: 10.13140/RG.2.2.12081.60008
Dold M, Lewis P (2023) A neglected topos in behavioural normative economics: The
opportunity and process aspect of freedom. Behav Public Policy 7(4):943–953
Dold M, Rizzo M (2021) The limits of opportunity-only: context-dependence and agency in
behavioral welfare economics. J Econ Methodol 28(4):364–373
Elster J (1986/2010) The market and the forum: Three varieties of political theory. In: The idea
of the public sphere: A reader
Engelen B, Nys T (2020) Nudging and autonomy: Analyzing and alleviating the worries. Rev
Philos Psychol 11(1):137–156
Fabian M, Alexandrova A, Coyle D, Agarwala M, Felici M (2022) Respecting the subject in
wellbeing public policy: Beyond the social planner perspective. J Eur Public Policy
Fishkin J, Bolotnyv V, Lerner J, Siu A, Bradburn N (2024) Can deliberation have lasting
effects? Am Polit Sci Rev
Frankfurt H (1971) Freedom of the will and the concept of person. J Philos 68(1):5–20
Fumagalli R (2016) Decision sciences and the new case for paternalism: Three welfare-related
justificatory challenges. Soc Choice Welf 47(2):459–480
Gigerenzer G (2024) From bounded rationality to ecological rationality. In: Gigerenzer G,
Mousavi S, Viale R (eds) The Herbert Simon Companion. Elgar Publishing
Grüne-Yanoff T (2012) Old wine in new casks: Libertarian paternalism still violates liberal
principles. Soc Choice Welf 38(4):635–645
Grüne-Yanoff T (2021) Boosts: A remedy for Rizzo and Whitman's Panglossian fatalism. Rev
Behav Econ 8(3–4):285–303
Grüne-Yanoff T, Hertwig R (2016) Nudge versus boost: How coherent are policy and theory?
Minds Mach 26(1):149–183
Gutmann A, Thompson D (1997) Democracy and disagreement. Harvard University Press,
Cambridge, MA
15
Habermas J (1996) Three normative models of democracy. In: Benhabib S (ed) Democracy
and difference: Contesting the boundaries of the political. Princeton University Press,
Princeton
Hargreaves Heap SP (2023) Mill's constitution of liberty: An alternative behavioural policy
framework. Behav Public Policy 7(4):933–942
Hausman DM, Welch B (2010) Debate: To nudge or not to nudge. J Polit Philos 18(1):123–
136
Hausman DM (2018) Efficacious and ethical public paternalism. Rev Behav Econ 5(3–
4):261–280
Hertwig R, Grüne-Yanoff T (2017) Nudging and boosting: Steering or empowering good
decisions. Perspect Psychol Sci 12(6):973–986
Hukkinen JI (2016) Addressing the practical and ethical issues of nudging in environmental
policy. Environ Values 25(3):329–351
Kahneman D (2011) Thinking, fast and slow. Farrar, Straus and Giroux, New York
Knowles C (2022) Beyond adaptive preferences: Rethinking women's complicity in their own
subordination. Eur J Philos 30:1317–1334
Korsgaard C (2009) Self-constitution. Oxford University Press, Oxford
Kuyer P, Gordijn B (2023) Nudge in perspective: A systematic literature review on the ethical
issues with nudging. Rationality Soc
Lafont C (2015) Deliberation, participation, and democratic legitimacy: Should deliberative
mini-publics shape public policy? J Polit Philos 23(1):40–63
Lecouteux G, Mitroutchev I (2023) The view from manywhere: Normative economics with
context-dependent preferences. Econ Philos 1–23
Lepenies R, Malecka M (2015) The institutional consequences of nudging—Nudges, politics,
and the law. Rev Philos Psychol 6(3):427–437
Lewis P, Dold M (2020) James Buchanan on the nature of choice: Ontology, artifactual man
and the constitutional moment in political economy. Camb J Econ 44(5):1159–1179
Loewenstein G, Bryce C, Hagmann D, Rajpal D (2015) Warning: You are about to be
nudged. Behav Sci Policy 1(1):35–42
McGinnis M (1999) Polycentric governance and development. University of Michigan Press,
Ann Arbor, MI
McKenzie CR, Sher S, Leong LM, Müller-Trede J (2018) Constructed preferences,
rationality, and choice architecture. Rev Behav Econ 5(3–4):337–360
16
Mercier H, Sperber D (2011) Why do humans reason? Arguments for an argumentative
theory. Behav Brain Sci 34(2):57–74
Mercier H, Sperber D (2017) The enigma of reason. Harvard University Press, Cambridge,
MA
Mill JS (1859/2003) On liberty. In: Bromwich D, Kateb G (eds) Yale University Press, New
Haven and London
Muldoon R (2018) Norms, nudges, and autonomy. In: Boonin D (ed) The Palgrave handbook
of philosophy and public policy. Palgrave, London, pp 225–233
Niemeyer S, Veri F, Dryzek J, Bächtiger A (2023) How deliberation happens: Enabling
deliberative reason. Am Polit Sci Rev (Online version)
Niemeyer S (2011) The emancipatory effect of deliberation: Empirical lessons from mini-
publics. Polit Soc 39(1):103–140
OECD (2017) Behavioural insights and public policy: Lessons from around the world. OECD
Publ.
Ostrom E (2010) Beyond markets and states: Polycentric governance of complex economic
systems. Am Econ Rev 100(3):641–672
Ostrom V, Tiebout C, Warren R (1961) The organization of government in metropolitan
areas: A theoretical inquiry. Am Polit Sci Rev 55(4):831–842
Rizzo M, Whitman G (2020) Escaping paternalism. Cambridge University Press, Cambridge
Schauer F (1999) Talking as a decision procedure. In: Deliberative politics: Essays on
democracy and disagreement. Oxford University Press, pp 17–27
Schmidt AT (2019) Getting real on rationality: Behavioral science, nudging, and public policy.
Ethics 129(4):511–543
Sen AK (2003) The importance of basic education. The Guardian
Smith A (1859) The theory of moral sentiments. Clarendon Press, London
Sugden R (2018) Community of advantage. Oxford University Press, Oxford
Sunstein C (2015) Choosing not to choose: Understanding the value of choice. Oxford
University Press, New York
Sunstein C (2016) The ethics of influence: Government in the age of behavioral science.
Cambridge University Press, New York
Sunstein C (2020) Behavioral welfare economics. J Benefit-Cost Anal 11(2):196–220
Taylor R (2017) Exit Left. Oxford University Press, Oxford
17
Thaler R, Sunstein C (2008) Nudges. Yale University Press, New Haven
Thiel A, Blomquist W, Garrick D (2019) Governing complexity: Analyzing and applying
polycentricity. Cambridge University Press, Cambridge
Vugts A, Van den Hoven M, De Vet E, Verweij M (2020) How autonomy is understood in the
discussion on the ethics of nudging. Behav Public Policy 4(1):108–123
Wilkinson TM (2013) Nudging and manipulation. Polit Stud 61(2):341–355
