Reasonableness on the Clapham Omnibus: Exploring the Outcome-Sensitive Folk Concept of Reasonable

Abstract

This paper presents a series of studies (total N=579) which demonstrate that folk judgments concerning the reasonableness of decisions and actions depend strongly on whether they engender positive or negative consequences. A particular decision is deemed more reasonable in retrospect when it produces beneficial consequences than when it produces harmful consequences, even if the situation in which the decision was taken and the epistemic circumstances of the agent are held fixed across conditions. This finding is worrisome for the law, where the reasonable person standard plays a prominent role. The legal concept of reasonableness is outcome-insensitive: whether the defendant acted in a reasonable fashion or not depends exclusively on her context of action, no matter how things play out. Folk judgments of reasonableness are thus inconsistent with the legal concept of reasonableness. Problematically, in common law jurisdictions, the decision whether a defendant’s behavior was reasonable or not is frequently (though not necessarily) delegated to a lay jury.

Full text

Reasonableness on the Clapham Omnibus:
Exploring the Outcome-Sensitive Folk
Concept of Reasonable
Markus Kneer
Abstract The reasonable person standard is of great importance to US criminal and
tort law. According to the law, whether or not an agent acted reasonably does not
depend on features of the outcome which are not under her control. Mock juror
attributions of reasonableness, however, are shown to be outcome-dependent. A
series of experiments reveals that this outcome-dependence does not constitute a
bias, since the very folk concept of reasonableness is outcome-sensitive. Conse-
quently, the law makes a mistaken assumption as to what kind of concept of
reasonableness lay jurors will apply in court. This conceptual misalignment, it is
argued, could lead to serious injustice in US trials.
Keywords Reasonable person standard · Negligence · Outcome effect · Legal
liability · Experimental jurisprudence
1 Introduction
This paper presents a series of studies which demonstrate that folk judgments
concerning the reasonableness of decisions and actions depend strongly on whether
they engender positive or negative consequences. A particular decision is deemed
more reasonable in retrospect when it produces beneficial consequences than when it
produces harmful consequences, even if the situation in which the decision was
I would like to thank Elliot Ash, Clark Barrett, Teneille Brown, Joshua Knobe, James Macleod, Dan
Simon, Alex Stremitzer, Doron Teichman, Kevin Tobia, the participants of the workshop
Experimental Methods in Law organized by Christoph Engel, the members of the Guilty Minds
Lab Zurich and the Editors for very helpful comments and suggestions. Research on this paper
was supported by an SNSF Ambizione grant for the project Reading Guilty Minds
(PZ00P1_179912).
M. Kneer (*)
University of Zürich, Zürich, Switzerland
e-mail: markus.kneer@uzh.ch
©The Author(s), under exclusive license to Springer Nature Switzerland AG 2022
P. Bystranowski et al. (eds.), Judicial Decision-Making, Economic Analysis of Law
in European Legal Scholarship 14, https://doi.org/10.1007/978-3-031-11744-2_3
25
Penultimate draft.
Please consult the final version:
https://doi.org/10.1007/978-3-031-11744-2_3

taken and the epistemic circumstances of the agent are held fixed across conditions.
This finding is worrisome for the law, where the reasonable person standard plays a
prominent role. The legal concept of reasonableness is outcome-insensitive:
Whether the defendant acted in a reasonable fashion or not depends exclusively on
her context of action, no matter how things play out. Folk judgments of reasonable-
ness are thus inconsistent with the legal concept of reasonableness. Problematically,
in common law jurisdictions, the decision whether a defendant’s behavior was
reasonable or not is frequently (though not necessarily) delegated to a lay jury.
What could, and should, be done about this problem depends on what explains the
outcome-sensitivity of folk judgments of reasonableness.
26 M. Kneer
Prima facie, the outcome-sensitivity of reasonableness judgments allows for
different interpretations. On a bias interpretation, perceived reasonableness post
hoc is inappropriately distorted by outcome information. This view comes in two
flavours: The unmediated bias view, according to which the effect of outcome
information on reasonableness judgments is (predominantly) direct. Alternatively,
there is the mediated bias view, according to which the distorting effect of outcome
information is mediated at least partially by another factor. What kind of factor?
Perceived probability is, perhaps, the most plausible candidate. On this view, a
decision is deemed more or less reasonable because the (beneficial or harmful)
consequences of said decisions are deemed more or less probable post hoc
depending on what actually happened. Or, in a nutshell, the outcome bias in
reasonableness judgments could be reduced, at least in parts, to the hindsight bias.
Importantly, however, the fact that outcome impacts reasonableness judgments
need not constitute a bias in the first place. The folk might deem it appropriate to
assess the reasonableness of decisions and actions inter alia on the basis of the
consequences these decisions and actions produce. In such a case, it is more
adequate to treat the folk concept of reasonable as inherently outcome-dependent,
and the outcome-dependence of reasonableness judgments would not only be
consistent with but due to a correct deployment of the concept.
In this paper, I will explore which interpretation of the outcome-sensitive folk
judgments of reasonableness is correct and what this means for the law, in particular
in common law systems where lay jurors (i.e. the folk) play a key role in adjudicating
whether a defendant acted in a reasonable fashion. To do so, I will begin with
theoretical preliminaries, focusing on the scope of the reasonable person standard,
the legal concept of reasonable, and general problems that can arise when the law
relies on jurors. A brief survey of previous pertinent empirical work motivates the
experiments. The bulk of the paper reports three studies that suggest that the folk
concept of reasonableness is outcome-dependent and explores what that means for
the law—in particular in countries where questions pertaining to reasonableness are
frequently decided by lay juries.

Reasonableness on the Clapham Omnibus: Exploring the Outcome-Sensitive... 27
1.1 The Reasonable Person Standard
Many of the law’s demands are tied to what it is reasonable to believe, feel, judge
and do.
1
Negligence, for instance, is attributed if a reasonable person would have
been aware of a substantial risk in the defendant’s circumstances.
2
Self-defense can
be claimed if a reasonable person would have justifiably feared for their life in the
defendant’s situation.
3
But reasonableness also plays a prominent role in cases
concerning, for instance, excessive fees, rates or prices,
4
statutory rape,
5
deceptive
advertising,
6
restraints on trade,
7
delivery on contractually agreed terms,
8
or the
public dissemination of private information.
9
Consequently, the reasonable person is
an authority on everything from the appropriate degree of emotional distress when
under attack to which kinds of information are deemed too embarrassing to have
disclosed in public.
The reasonable person standard, Gardner argues, “exists to allow the law to pass
the buck, to help itself pro tempore to standards of justification that are not them-
selves set by the law, and which therefore are only as good as the standards of
justification used by the person or people to whom the buck is passed.”Recurrence
to the reasonable person allows the law to significantly expand its scope without the
introduction of countless specific and controversial rules. Instead the buck is passed
to the fact-finder, frequently a jury in common law countries, “who is invited to use
extra-legal standards to bridge the gap from legal rule to legal ruling.”
10
1
For discussion, see e.g. Fletcher (1985), Gardner (2001,2015), Moran (2003,2010), Westen
(2008), Miller and Perry (2012), Zipursky (2014), and Tobia (2018).
2
Model Penal Code §2.02. (d).
3
UK Criminal Law Act, 1967, c. 58, §85 sch. 3(1) (Eng. & Wales) (“A person may use such force
as is reasonable in the circumstances in the prevention of crime [...].”); United States v. Peterson
483 F 2d 1222 (1973); NY Penal Code Art. 35; see also Baron (2011).
4
See, e.g., 16 U.S.C. §824e(a) (2012) (stating that the Federal Energy Regulatory Commission
must set “the just and reasonable rate”); 16 U.S.C. §2621(18)(B) (2012) (stating that states should
ensure “a reasonable rate”of return when authorizing utility companies). More generally, see
Berger (1977).
5
Model Penal Code §213.6(1).
6
See, e.g., Fink v. Time Warner Cable, 714 F.3d 739, 741 (2d Cir. 2013); Elias v. Hewlett-Packard
Co., 950 F. Supp. 2d 1123, 1131 (N.D. Cal. 2013).
7
See e.g. Leegin Creative Leather Prod., Inc. v. PSKS, Inc., 551 U.S. 877, 885 (2007), (“[T]he
Court has repeated time and again that §1‘outlaw[s] only unreasonable restraints.’”); Standard Oil
Co. of N.J. v. United States, 221 U.S. 1, 60 (1911) (stating that “the standard of reason”regards anti-
competitive trade restraints under the Sherman Act).
8
See e.g. Mayfield v. Koroghli, 184 P.3d 362, 366 (Nev. 2008) (stating that parties must deliver on
contracts within a “reasonable time”).
9
Restatement (Second) of Torts §652D (1977).
10
Gardner (2015), p. 36.

r
28 M. Kneer
1.2 Two General Threats to Trial by Jury
When the law “passes the buck”to lay juries, they need the relevant competences in
the tasks assigned to them. One competence regards the grasp of the particular
concept at stake as defined by the law. Jurors must know (often following instruc-
tions) how expressions such as “reasonable”, but also “intention”,“cause”,o
“probability”are understood in the legal context. Another competence regards
their application: Jurors must assess the reasonableness of the action, and other
relevant features of the case at issue in responsible and unbiased fashion. In trials
where jurors are involved, justice is thus under potential threat from at least two
sources: unnoticed conceptual divergence between the legal and the folk concepts at
stake, and cognitive bias.
Naturally, once identified, each problem requires a different remedy. If the law
takes itself to operate with a folk concept of e.g. reasonableness, causation, or
intention, and this assumption turns out mistaken (the actual folk concept differs
from the presumed folk concept), we must work towards conceptual alignment.On
the other hand, if conceptual alignment exists, yet cognitive biases distort the
assessment of reasonableness, causation, mens rea and the like, strategies to debias
such judgments must be devised. Given the vast reach of the reasonable person
standard and serious concerns as to whether the folk concept of, e.g. intention or
causation fit the ones in use in the law,
11
it is of interest to test whether folk
judgments of what is reasonable are consistent with the legal concept of reasonable-
ness. In order to do this, we will begin with a closer examination of the legal meaning
of “reasonable”.
1.3 The Legal Concept of Reasonableness
What is the legal concept of reasonableness, and what should it be? These questions
have received considerable attention over the last two decades. One debate explores
the conceptual category or nature of reasonableness: Does “reasonable”denote a
descriptive or a normative concept? More precisely, must we understand that which
is reasonable descriptively as what an average or ordinary person would do? Or
must we rather understand it as what an upright or ideal person would do, in which
case the concept of reasonableness would be normative?
12
Perhaps neither interpre-
tation is correct tout court, and the expression “reasonable”is ambiguous and takes
different meanings in different legal contexts.
13
It could even have a hybrid meaning
11
See e.g. Alicke (1992,2000), Nadelhoffer (2006), Macleod (2019), Kneer and Bourgeois-
Gironde (2017), Tobia (2020b), and Güver and Kneer (2022).
12
For discussion, see Ripstein (2001), Gardner (2001,2015), Moran (2003), Westen (2008), Miller
and Perry (2012), Zipursky (2014), Tilley (2017), and Tobia (2018).
13
Zipursky (2014).

with both descriptive and normative features.
14
A distinct debate examines whether
it is appropriate to hold everyone, no matter their particular individual or “subjec-
tive”features to account with respect to a single “objective”standard.
15
Reasonableness on the Clapham Omnibus: Exploring the Outcome-Sensitive... 29
The descriptive/normative and the subjective/objective debates are fascinating
and of considerable practical importance to the law. I will set them aside in this
paper, and turn to a third question which I will label the potential outcome-sensitivity
of the folk concept of reasonableness.
16
In recent work on practical rationality
broadly conceived, I found judgments of what is reasonable, what is rational, and
what should be done, to depend on outcome valence. In two situations A (negative
outcome) and B (positive outcome), in which two agents have the same information
and act with the same frame of mind, A’s decision is judged significantly less
reasonable than B’s.
17
These results are at odds with the legal concept of reason-
ableness (or at least the correct application thereof): Whether or not a decision,
belief, emotional reaction or action is reasonable does not depend on the conse-
quences, but the situation in which the agent finds herself. Or, to put it somewhat
more technically, the semantics of the legal expression “reasonable”is sensitive to
the context of decision or action, not the ex post context of assessment where the
consequences are known.
18
The Model Penal Code, for instance, is perfectly explicit
about this with respect to criminal negligence.
19
The short paragraph emphasizes
that what matters is the agent’sepistemic perspective and circumstances and not
some context ex post, and this is stressed not only once, but twice (see italics):
A person acts negligently with respect to a material element of an offense when he should be
aware of a substantial and unjustifiable risk that the material element exists or will result
from his conduct. The risk must be of such a nature and degree that the actor’s failure to
perceive it, considering the nature and purpose of his conduct and the circumstances known
to him, involves a gross deviation from the standard of care that a reasonable person would
observe in the actor’s situation. (MPC, 2.02. (d), italics added)
14
Tobia (2018).
15
For discussion, see. e.g. Wells (1982), Kelman (1991), Armour (1994), Moran (2003,2010), Lee
(2007), Simons (2008), and Nourse (2008).
16
This question, whose practical importance might even exceed the one of the other two debates has
received next to no attention thus far. Two exceptions are the excellent empirical papers by Baron
and Hershey (1988) and Kamin and Rachlinski (1995). The latter is limited to probability judgments
in tort negligence, the former explicitly includes “reasonable”as a dependent variable, though
mentions a plethora of other expressions in the same scale. Both studies have very low power, and
neither of the two employs a clean within-subjects design, which—as will be argued below—is
essential to a proper understanding of the folk concept of reasonableness.
17
Kneer (in preparation-a), see also Baron and Hershey (1988).
18
Kneer (in preparation-b), for the semantic framework, see MacFarlane (2014).
19
The same emphasis on the agent’s circumstances is found in the definition of recklessness (see
MPC 2.02 (c)), where the “reasonable person”is substituted with “the law-abiding person”). Many
commentators, however, treat negligence and recklessness as very similar, and the “law-abiding
person”as a close cousin of the “reasonable person”. Both negligence and recklessness are a matter
of running an unreasonable risk, the difference lying in whether the defendant did so knowingly or
not. For discussion, see e.g. Dressler (1995) and Simons (2002).

Section 220.3 (regarding manslaughter) of the Model Penal Code further elucidates
how “reasonableness”must be understood. Criminal homicide is a case of man-
slaughter if the killing “committed under the influence of extreme mental or emo-
tional disturbance for which there is reasonable explanation or excuse. The
reasonableness of such explanation or excuse shall be determined from the viewpoint
of a person in the actor’s situation under the circumstances as he believes them to
be.”
20
s
30 M. Kneer
The Restatement (3rd) of Torts on Negligence §3 further demonstrates the point:
A person acts negligently if the person does not exercise reasonable care under all the
circumstances. Primary factors to consider in ascertaining whether the person’s conduct
lacks reasonable care are the foreseeable likelihood that the person’s conduct will result in
harm, the foreseeable severity of any harm that may ensue, and the burden of precautions to
eliminate or reduce the risk of harm. (italics added)
What matters for the evaluation of reasonable care are the foreseeable likelihood of
harm, and the foreseeable severity of harm—in short, the foresseable risk. The law’s
insistence on the foreseeability of risk thus limits the factors relevant for the
assessment of reasonable care or negligence to features of the defendant’s circum-
stances ex ante. What actually happens, and what becomes known ex post,i
irrelevant
21
—the legal meaning of “reasonable”is outcome-independent.
22
Ima-
gine, for a moment, that it were not. Then the vast area of the law in which judgments
of reasonableness take centre stage would be a matter of pure outcome-luck.
23
Under
conditions of risk and uncertainty, any action—no matter how rigorously thought
through and carefully executed—can produce harmful consequences. But if reason-
ableness, on its legal understanding, were outcome-dependent, then the reasonable
person standard would just be strict liability in fanciful disguise: No matter how
reasonable the decision ex ante, if the defendant is unlucky and the consequences
dire, her decision or action is likely to be judged unreasonable ex post.
20
For discussion, see e.g. Low (1987) and Dressler (1995).
21
Given that the law explicitly addresses the severity of the foreseeable consequences, one might
think that the legal concept of reasonable care (and negligence) is outcome-sensitive. And in some
sense, of course, it is: Holding probability fixed, it is considerably more unreasonable to embark on
an action where a human life is at risk than an action that might lead to a broken finger. Importantly,
however, the outcome-sensitivity thus understood regards possible outcomes. In the main text,
when I write about outcome-dependence, I am interested in the influence of actual outcomes—
whether or not the possible bad outcome does come to pass—on the assessment of the defendant’s
behaviour. Differently put, when I write about the “outcome-dependence of reasonableness”, what
is meant is not the ex ante, or epistemic, reading thereof, but the ex post, and thus partially
metaphysical understanding of “outcome-dependence”. Thanks a lot to Kevin Tobia for pointing
out the need for clarification in this regard, and for drawing my attention to the passage from
Tort law.
22
For further discussion, see Kamin and Rachlinski (1995) and Rachlinski (1998).
23
On moral luck in general, see Nagel (1979), Williams (1993), and Hartman (2017). With a view to
the law, see Enoch (2007,2010); Enoch and Marmor (2007), for empirical work, see Martin and
Cushman (2016) and Kneer and Machery (2019).

Reasonableness on the Clapham Omnibus: Exploring the Outcome-Sensitive...31
1.4 Identifying the Problem
The legal concept of reasonableness is outcome-insensitive—what is, and is not
reasonable, must be evaluated with regards to the context of decision or action, not
post hoc, at a later context when the consequences are known. Preliminary empirical
findings, however, suggest that folk judgments of reasonableness are inconsistent
with the legal concept of reasonableness. People’s assessments of what is reasonable
depend strongly on outcomes. This, we said, might be the consequence of one of two
more general potential problems: On the one hand, there might be conceptual
alignment between the legal and the folk understanding of reasonableness, yet folk
judgments are distorted by an outcome bias. To put it differently: As presumed by
lawmakers and legal practitioners, the legal meaning of “reasonable”is the same as
the ordinary language meaning of the expression. But despite conceptual alignment,
people might find it difficult to disregard outcomes. An effect of this sort is also
found, for instance, when it comes to the ascription of causation and mens rea in
between-subjects experiments.
24
Hence, if reasonableness judgments were distorted,
we should work towards a better understanding of their underlying “mechanics”,so
as to devise means to alleviate the bias.
An alternative to the bias view is the following: The law might wrongly assume
conceptual alignment. Whereas the law operates with an outcome-insensitive and
dynamically invariant concept of reasonableness, the folk concept might simply
differ—without the outcome-sensitivity of “reasonable”constituting a bias. The folk
concept of punishment, for instance, is strongly sensitive to outcomes,
25
without this
clearly suggesting that anything is amiss: People tend to judge deserved punishment
differently across outcomes even in cases in which they assess blame and wrongness
identically.
26
If this possibility were born out, the law, with its perpetual insistence
that what is reasonable just is what the “man on the Clapham omnibus”or “the
average Joe”takes it to be, would be completely mistaken, because the women and
men on the Clapham bus in fact have a different concept of reasonableness than what
the law takes them to have. The law, that is to say, would unwittingly operate with a
technical concept of reasonableness, yet assume that the legal concept is the
non-technical folk concept.
1.5 Structure of the Paper
The paper proceeds as follows: In Sect. 2I will say a few more words about extant
research regarding outcome effects in empirical legal scholarship and experimental
24
Alicke (1992,2000); Alicke and Rose (2010); Nadelhoffer (2006); Kneer and Bourgeois-Gironde
(2017); Bourgeois-Gironde and Kneer (2018).
25
Cushman (2008).
26
Kneer and Machery (2019).

32 M. Kneer
jurisprudence.
27
Sections 3–5report three experiments based on the well-known TJ
Hooper torts negligence case, as well as a medical malpractice scenario. The results
suggest that the folk concept of reasonableness is outcome-sensitive, dynamically
variant across ex ante and ex post contexts of assessment and that these effects are
not due to the hindsight bias. Section 6concludes by arguing that, out of the two
threats to trial by jury we are faced with conceptual misalignment. It is thus of utmost
practical importance for the law to come to understand that it is operating with a
concept of reasonableness that differs radically from the folk concept and to take
precautions against the likely injustice this generates on a daily basis.
2 Outcome Effects
2.1 Intention and Knowledge
Culpability in Western criminal law requires both an actus reus (a “guilty act”) and a
mens rea (a “guilty mind”). Since neither a bad intention nor a harmful accident, by
themselves, engender criminal responsibility (Coke’s law), the actus reus and the
mens rea must be established independently and treated as conceptually distinct.
Findings from psychology and experimental philosophy raise a worry for this neat
picture of two independent constitutive judgments feeding into culpability assess-
ment: Outcome is known to influence all-things-considered judgments such as
blame, culpability and punishment. It might thus be the case, that certain features
of the actus reus—such as outcome valence and severity—affect judgments as to
whether a defendant acted with a guilty mind.
There is, in fact, ample evidence that this is precisely the case at the more
inculpating spectrum of the mens rea catalogue. Outcome valence (good v. bad)
has an impact on attributions of intentionality
28
as well as knowledge and belief
29
.
Furthermore, there is considerable evidence that the severity of harm correlates
positively with mens rea ascription. For instance, the more severe an undesired,
27
Experimental jurisprudence is a young discipline which explores philosophical topics regarding
the law by empirical means. See e.g. Kneer and Bourgeois-Gironde (2017), Bourgeois-Gironde and
Kneer (2018), Prochownik et al. (2020) and Tobia (2020a)onmens rea attribution, Sommers
(2019) on the notion of consent, MacLeod (2019), Knobe and Shapiro (2021) and Güver and Kneer
(2022) on legal causation, Lidén et al. (2019) on the confirmation bias in guilt assessments,
Strohmaier et al. (2020) on the hindsight bias in the law, Donelson and Hannikainen (2020) and
Hannikainen et al. (2021) on Fuller’s theory of law, Struchiner et al. (2020) on the Fuller/Hart
debate, or Tobia (2020b) on ordinary meaning and legal interpretation to name but a few. For
reviews see Prochownik (2021) and Tobia (2022).
28
The Knobe effect, see Knobe (2003a,b); for reviews, see Feltz (2007); Cova (2016).
29
The epistemic side-effect effect, Beebe and Buckwalter (2010); Beebe and Jensen (2012).

Reasonableness on the Clapham Omnibus: Exploring the Outcome-Sensitive... 33
yet knowingly occasioned harm, the higher the likelihood that people will judge it
intentional
30
or that they will judge it as known or foreseen.
31
These findings are of importance for common law countries, where the mens rea
question in criminal trials is frequently decided by lay juries. The situation might be
similarly problematic in countries relying exclusively (or near-exclusively) on
professional judges: The Knobe effect and severity effects have also been found to
arise among professional judges, lawyers, and law students across several
countries.
32
2.2 Recklessness, Negligence and the Reasonable Person
Standard
The reasonable person standard features prominently in the lower echelons of the
mens rea catalogue. An agent is criminally liable when grossly deviating from
standards of reasonable care, be it knowingly (thus recklessly) or unwittingly (thus
negligently).
33
Although the topic has not been explored much in the empirical
literature, there is at least some evidence that negligence attributions are outcome-
sensitive.
34
In between-subjects designs, people are more inclined to ascribe negli-
gence to an unlucky, well-intentioned agent whose actions lead to a bad outcome,
than to a lucky agent, whose otherwise identical actions do not. Interestingly, the
outcome effect on negligence is mediated by perceived probability: People deem a
negative outcome more likely in retrospect than a neutral one and thus tend to be
more willing to ascribe negligence in the former than in the latter case. This finding
dovetails with Fischhoff’s (1975) hindsight bias, i.e. the change in perceived
probabilities of events as they unfold. Naturally, this is bad news for the law.
35
As
the above-quoted passage from the Model Penal Code (2.02 (d)) makes clear, the law
cares about the probabilities, and hence the risks, at the context of action, not some
later context when the consequences are known. But if perceptions of probabilities
30
Kneer and Bourgeois-Gironde (2017); García Olier and Kneer (in preparation); Prochownik et al.
(2020); Tobia (2020a,b).
31
Kneer et al. (in preparation).
32
Kneer and Bourgeois-Gironde (2017); Bourgeois-Gironde and Kneer (2018) for French judges
and lawyers respectively; Prochownik et al. (2020) for German law students; Kneer et al. (in
preparation) for professionals in the UK, Poland, the Netherlands and Brazil; the Knobe effect—
though not severity effect—also arises for US professionals, Tobia (2020a,b).
33
Cf. MPC 2.02 (c,d), and Dressler, 2015 §10.04.D, see also Homicide Act 1957 s3 and Gardner
(2015). For discussion of the reasonable person standard with respect to negligence, see
e.g. Fletcher (1970), Hurd and Moore (2002), Moore and Hurd (2011), Alexander et al. (2009)
and Stark (2016).
34
Kamin and Rachlinski (1995); Kneer and Machery (2019); Frisch et al. (2021); Kneer and
Skoczeń(2022).
35
For discussion see Rachlinski (1998); Teichman (2014); Wittlin (2016); Hoffman (ms.);
Strohmaier et al. (2020); Kneer and Skoczeń(2022).

¼ ¼
34 M. Kneer
and risks are outcome-dependent, it comes as no surprise that evaluations of
negligence and recklessness—in contrast to what the law requires—are, too.
The experiments presented below aspire to break new ground in several ways:
First, and in contrast to Kamin and Rachlinski
36
and Kneer and Machery,
37
the
dependent variable of Experiment 1 (and the follow-up experiments) is reasonable-
ness per se, not merely a related variable, negligence or probability. The implications
of the results are thus considerably broader in scope, since the reasonable person
standard is not limited to negligence cases. Second, Experiment 2 explores whether
the outcome-dependence of reasonableness judgments is a fundamental feature of
the folk-concept of reasonable, or a bias, as is standardly simply assumed. Third,
Experiment 3 investigates whether the outcome-dependence of reasonableness judg-
ments is but a consequence of the hindsight bias. This is done by holding perceived
probability fixed across outcomes, so as to explore whether a direct effect of
outcome on reasonableness remains under well-controlled circumstances. In sum,
the paper aspires to make considerable progress not only as to whether outcome
influences attributions of reasonableness, but what explains the outcome depen-
dence, how it is to be interpreted and what all this means for the law.
3 Experiment 1
The first experiment employs a between-subjects design to explore whether people
deem a decision more reasonable ex post if the consequences are positive than when
they are negative (holding all else fixed across conditions). The scenario is based on
a famous negligence case, The TJ Hooper.
38
In that case, the plaintiff lost two barges
and their cargo in a storm, claiming that this was due to the fact that the defendant’s
tugboats—the TJ Hooper and the Montrose—did not have reliable radio equipment,
which could have warned them of the storm.
3.1 Participants
I recruited 178 participants on Amazon Mechanical Turk. The IP address was
restricted to the United States. As preregistered,
39
I excluded participants failing
an attention check and those who responded to the first question (which required
reading the scenario) in less than 15 s. 151 participants remained (70 female, age
M 43 years, SD 14 years).
36
Kamin and Rachlinski (1995).
37
Kneer and Machery (2019).
38
The T.J. Hooper, 53 F.2d 107 (S.D.N.Y. 1931), for discussion see e.g. Epstein (1992).
39
https://aspredicted.org/blind.php?x¼ke2ia5. The data for all three experiments is available under
https://osf.io/3qudf/

3.2 Methods and Materials
Participants were presented with one of two conditions of the following scenario.
The two conditions differed only in terms of outcome (positive v. negative; labels in
bold omitted):
Mr Garcia runs a flourishing logistics business with two dozen cargo jets. Until now, he has
exclusively focused on air transport. He decides to expand to shipping and buys five cargo
ships.
For ships, there is a large variety of security systems available. They track other ships, make
detailed weather predictions, and minimize risks for ships in distress. Garcia decides to get a
package deal from a provider who installs several different types of maritime security
systems on his cargo freighters.
The provider offers different plans for servicing the security systems and updating the
software. There are four options: A monthly check, a bimonthly check, a check every six
months, and a plan where the systems are only checked and updated once a year. The yearly
plan is by far the cheapest, the more regular the checks get, the more expensive the plan.
Out of the four options, Garcia decides to take the second-cheapest. The security systems on
his boats are thus checked and updated every six months.
Positive outcome: Garcia’s cargo shipping business has many clients. 15 months after
commencing operations, there has not been a single security incident on any of his cargo
freighters.
Reasonableness on the Clapham Omnibus: Exploring the Outcome-Sensitive... 35
Negative outcome: Garcia’s cargo shipping business has many clients. 15 months after
commencing operations, one of Garcia’s cargo ships crashes into a rock due to the
malfunctioning of one of its security systems. The ship sinks. The 40 tons of crude oil it
was transporting are spilled into the ocean. The damage to the environment is very serious.
Participants were asked to what extent they agreed with the claim “It was reasonable
for Garcia to choose the six-month service plan.”Responses were collected on a
7-point Likert scale anchored at 1 (completely disagree) and 7 (completely agree).
Next, participants were asked “[h]ow much praise or blame, if any, does Garcia
deserve for choosing the six-month servicing plan?”. The 7-point Likert scale was
anchored at 1 (lots of praise) and 7 (lots of blame), the midpoint 4 being labelled
“neither praise nor blame”.
3.3 Results
There was a pronounced effect of outcome valence on judgments of reasonableness
ex post. In the positive outcome condition, where there has not been a single security
incident, mean reasonableness (M ¼5.24, SD ¼1.43) was significantly higher than
in the negative outcome condition, in which one of the ships crashes (M ¼3.96,
SD ¼1.88), t(1,149) ¼4.70, p<.001, Cohen’sd¼.76, a medium-large effect. As
regards blame and praise, the situation was reversed. Mean blame/praise ascriptions
in the negative outcome condition (M ¼4.92, SD ¼1.36) significantly exceeded

36 M. Kneer
Fig. 1 Mean
reasonableness ex post and
mean blame/praise
attributions across
conditions (positive
v. negative outcome). Error
bars denote standard error of
the mean
mean blame/praise in the positive outcome condition (M ¼3.69, SD ¼1.05),
t(1,149) ¼"6.24, p<.001, Cohen’sd¼1.01, a large effect. Figure 1represents
the results graphically.
3.4 Discussion
Consistent with previous findings,
40
outcome had a significant and pronounced
effect on judgments of reasonableness and blame. The results raise a serious
worry: The law employs an outcome-insensitive concept of reasonableness, but
people’spost hoc evaluations of whether an agent’s decisions were reasonable are
strongly sensitive to outcomes. Given these results, the question arises which of the
two threats to trial by jury discussed in the introduction is at stake. Are we facing a
performance error in the application of a folk concept of reasonableness that is
identical to the legal one in so far as it is outcome-insensitive (if correctly
employed)? Or might there be tacit conceptual divergence, i.e. might it be the case
that the folk concept of reasonableness just is outcome-sensitive, and thus differs
from the legal concept?
Just like our first experiment, the little work that touches upon the reasonable
person standard in existence cannot answer this question. Kamin and Rachlinski
(who test for probability and negligence rather than reasonableness more generally)
ran a between-subjects design, Baron and Hershey
41
presented participants with a
whole battery of scenarios with good and bad outcomes. In order to better understand
what drives the outcome effect in assessments of reasonableness ex post, and thus to
identify which of the two threats to jury trial is at stake, a different design is required:
We should explore whether people, when assessing a decision or action ex ante—i.e.
before the consequences are known—are willing to change their evaluations of
40
Baron and Hershey (1988); Kneer (in preparation-a).
41
Baron and Hershey (1988).

¼ ¼
¼
Reasonableness on the Clapham Omnibus: Exploring the Outcome-Sensitive...37
reasonableness ex post once the consequences have been revealed. If people, in the
novel design, assess reasonableness by and large the same ex ante and ex post, this
would indicate that their concept or reasonableness is, like the legal one, outcome-
independent. The fact that, in between-subjects designs, we find a pronounced
outcome effect could thus be interpreted as a performance error or bias. If, on the
other hand, people change their views of whether an action was reasonable ex post in
contrast to what they judged ex ante on the basis of the outcome, then the folk
concept is simply not what the law presumes it to be. In that case, the folk concept of
reasonableness (qua folk concept) is appropriately sensitive to outcome, and the
outcome effect in Experiment 1 the consequence of its proper application.
4 Experiment 2
The goal of Experiment 2 is to investigate whether the outcome-sensitivity of
reasonableness judgments ex post constitutes a performance error or else, whether
the folk concept of reasonableness is simply outcome-sensitive (in contrast to the
legal concept). If people judge reasonableness ex ante and ex post the same, we
would have evidence in favour of the former hypothesis. If people’sex post
judgments of reasonableness diverge from their ex ante judgments depending on
outcome, we would have evidence in favour of the latter hypothesis.
4.1 Participants
I recruited 180 participants on Amazon Mechanical Turk. The IP address was
restricted to the United States. As preregistered,
42
I excluded participants failing
an attention check and those who responded to the first question (which required
reading the scenario) in less than 15 s. 148 participants remained (71 female, age
M 43 years, SD 13 years).
4.2 Methods & Materials
The materials were similar to those in Experiment 1. The main difference was that,
once Garcia’sgeneral situation has been explained, participants are merely told that
Garcia considers choosing the second-least expensive security plan, according to
which the boats are serviced every six months. They then had to state whether they
agree or disagree with the claim that “[i]t is reasonable for Garcia to choose the
42
https://aspredicted.org/blind.php?x hp8zx4.

¼ ¼
38 M. Kneer
Fig. 2 Mean reasonableness ex ante and ex post across outcomes. Error bars denote standard error
of the mean
six-month service plan”on a 7-point Likert scale. Differently put, participants rated
reasonableness at the context of decision, i.e. reasonableness ex ante. On the next
screen, participants were told that Garcia in fact chose the six-month plan, and they
were presented with either the positive outcome (smooth sailing) or negative out-
come (one of the boats crashes). They were then asked to what extent they agree or
disagree with the claim that it was reasonable for Garcia to choose that plan (7-point
Likert scale), and to what extent they would attribute praise or blame to him for
deciding as he did (7-point Likert scale, questions were framed and responses
labelled as in Experiment 1). The mixed design experiment thus allows us to contrast
perceived reasonableness ex ante and ex post for each of the two conditions, and
reasonableness ex post and blame across outcomes.
4.3 Results
4.3.1 Within-Subjects Results
In this experiment, people rated the reasonableness of Garcia’s decision both before
the outcome was revealed, and again once outcomes had been presented. The
purpose of the experiment was to explore whether, on average, people change
their assessment of the reasonableness of a decision in light of its consequences
once the latter are revealed. This was indeed the case for both outcomes. In the
positive outcome condition, people deemed it significantly more reasonable for
Garcia to choose the six-month plan ex post (M ¼5.40, SD ¼1.53) than they did
ex ante (M ¼5.12, SD ¼1.62), paired-samples t-test, t(1,77) ¼"2.58, p¼.012,
Cohen’sd¼.29, a small effect, see Fig. 2. In the negative outcome condition, people
deemed it significantly less reasonable to choose the six-month plan ex post
(M ¼4.09, SD ¼1.99) than they did ex ante (M ¼5.43, SD ¼1.48),
t(1,69) 6.38, p<.001, Cohen’sd.76, a medium-large effect.

¼" ¼
Reasonableness on the Clapham Omnibus: Exploring the Outcome-Sensitive... 39
4.3.2 Between-Subjects Results
The first part of the scenario (at which the outcome is still withheld) was identical
across conditions. Consequently, assessments of reasonableness ex ante should not
differ, which is indeed what we find (independent samples t-test, t(1,146) ¼"1.22,
p¼.223, two-tailed). Participants across the different conditions then viewed the
distinct outcomes on a second screen, and had to assess reasonableness ex post. The
impact of outcome on reasonableness was significant, t(1,146) ¼4.52, p<.001,
Cohen’sd¼.74, a medium-large effect, see Fig. 3. The same held for blame/praise,
t(1,146) 6.43, p<.001, Cohen’sd1.06.
4.3.3 Comparison Across Designs
The data collected so far allows a further analysis: We can explore whether reason-
ableness assessments ex post differ across subjects who first rated reasonableness ex
ante (a sort of priming technique for debiasing purposes, results from Experiment 2)
and subjects exclusively confronted with questions of reasonableness ex post
(Experiment 1). The same analysis can be conducted for blame. A 2 (outcome:
positive v. negative) #2 (design: ex ante assessment or priming v. no ex ante
assessment or no priming) ANOVA reveals the expected significant main effect for
outcome (F(1,295) ¼42.43, p<.001). Importantly, though, there was neither a
significant main effect for experiment design (F(1,295 ¼.49, p¼.484) nor a
significant outcome*design interaction (F(1,295) ¼.005, p¼.941). For blame,
the pattern of results was the same: A significant main effect for outcome (F
(1,295) ¼79.94, p<.001), a nonsignificant main effect for design (F
(1,295) ¼.27, p¼.601), as well as a nonsignificant interaction (F(1,295) ¼.01,
p¼.941). This is quite remarkable: Priming people by aid of an ex ante assessment
Fig. 3 Mean reasonableness ex post and mean blame/praise attributions across conditions (positive
v. negative outcome). Error bars denote standard error of the mean

40 M. Kneer
of the reasonableness of Garcia’s decision—and thus potentially immunizing them
against an outcome effect or hindsight bias—had no effect whatsoever. Note that the
probability of finding a mid-size effect of f¼.25 with the recruited sample size of
N¼299 across the two experiments and an alpha of .05 exceeds 99% to find such an
effect, if there is one to be found (power analysis conducted with G*Power 3.1).
4.4 Discussion
Consistent with the results of Experiment 1, and broadly in tune with those of Baron
and Hershey
43
and Kamin and Rachlinski,
44
we found that ascriptions of reason-
ableness (or related variables) are outcome-dependent. None of these previous
studies, however, address the central question of whether the outcome effect testifies
to a biased ascription of an outcome-insensitive concept of reasonableness, or
whether the folk-concept of reasonableness simply is outcome-dependent (and
thus correctly applied). The design of Experiment 2 sheds some light on the matter,
because participants judge Garcia’s decision both ex ante and ex post. If the outcome
effect on reasonableness were a bias, then people would presumably not change their
assessment once the consequences have been revealed. Interestingly, they do: People
judge a decision that leads to negative consequences as less reasonable ex post than
ex ante, and a decision that leads to positive consequences as more reasonable ex
post than ex ante. Furthermore, the two-step procedure (first ex ante, then ex post
assessment) does not mitigate the outcome effect on reasonableness ex post at all, as
our comparison between Experiment 1 (only ex post assessments) and Experiment
2 (assessment both ex ante and ex post) reveal. Together, the results suggest that the
folk concept of reasonableness is genuinely outcome-dependent, and thus diverges
radically from the legal concept, and from what the law assumes about the folk
concept.
There is a final worry. Kamin and Rachlinski
45
as well as Kneer and Machery
46
have argued that the outcome effect on negligence is due to an outcome effect on
perceived probability post hoc. In other words, the outcome effect on negligence is
mediated by perceived probability post hoc. A similar hypothesis is possible for
reasonableness more broadly conceived: The fact that the probability of harm post
hoc is judged higher when harm actually does occur than when it does not might lead
people to evaluate a decision that leads to harm as less reasonable than one that does
not. Whereas an outcome effect on the perceived probability of the sort invoked here
43
Baron and Hershey (1988).
44
Kamin and Rachlinski (1995).
45
Kamin and Rachlinski (1995).
46
Kneer and Machery (2019).

¼
Reasonableness on the Clapham Omnibus: Exploring the Outcome-Sensitive... 41
is standardly interpreted as a bias,
47
arationalist interpretation is also possible.
48
Especially in cases where probabilities are fuzzy and hard to assess, it makes sense to
use outcomes as an indicator of an event’s probability: Whether a decision or action
did lead to a particular outcome or not constitutes useful information for the
assessment (and potential reassessment) of whether it was likely that it would lead
to that sort of outcome. While this is an interesting question to pursue in further
detail, our final experiment focuses on an even more fundamental one: Is there, at
least sometimes, a direct or unmediated effect of outcome on reasonableness ex
post? In other words, can an outcome effect on reasonableness ex post be observed
even if we were to hold perceived probability ex post constant across conditions? If
so, this would show that the outcome-sensitivity of reasonableness cannot be
reduced to a hindsight effect—whether the latter is best interpreted as a bias or not.
5 Experiment 3
The goal of the final experiment consists in exploring the effect of outcome on
reasonableness in situations where the risk can be clearly quantified ex ante. The
vignette explicitly specifies the probability of a harmful outcome, which, paired with
a manipulation check focusing on probability, allows us to minimize the influence of
the hindsight effect on reasonableness attributions. That way, we can explore
whether the outcome-effect on reasonableness is tied to perceived probability, or
whether outcome can influence perceived reasonableness directly.
The vignette itself is based on a duty of care case—a type of case where
reasonableness also plays a central role. Manufacturers, for instance, must exercise
reasonable care in providing consumers with safe products, doctors are under the
obligation to prevent reasonably foreseeable harm to their patients. In cases of this
sort, it is once again up to the fact-finder to decide what, exactly, constitutes a
reasonable measure of care and what does not.
5.1 Participants
I recruited 219 participants on Amazon Mechanical Turk. The IP address was
restricted to the United States. As preregistered,
49
I excluded participants failing
an attention check and those who responded to the first question (which required
47
For instance by Kamin and Rachlinski (1995), as well as nearly all the papers on the “hindsight
bias”surveyed by Rachlinski (1998). Note how rare it is to read about a hindsight effect rather than a
hindsight bias. The problem is discussed in more detail by Kneer and Machery (2019), who explore
the rationalistic interpretation of the hindsight effect at length. For excellent discussion, see also
Teichman (2014), Wittlin (2016) and Brown (2022).
48
I’d like to thank Joshua Knobe for pressing me on this important point.
49
https://aspredicted.org/blind.php?x ng28n4.

Besides responding to the two main questions, participants also had to answer the
following question regarding probability: “On a scale of 0% (completely unlikely) to
100% (certain), how high was the probability that Fluxis would cause damage to the
liver?”. As explained above, this question served as a manipulation check. In order to
explore the impact of outcome on reasonableness independently of a hindsight
reading the scenario) in less than 15 s. As also preregistered (and detailed below), I
also excluded all participants who, after having learned the outcome, did not report
the precise probability of a harmful consequence, which was specified explicitly in
the scenario. This ensured that the outcome effect on reasonableness ex post was not
driven by the hindsight bias (i.e. a diverging assessment of probability ex ante v. ex
post). 125 participants remained (55 female, age M¼41 years, SD ¼13 years).
5.2 Methods and Materials
Like Experiment 2, Experiment 3 also took a 2 (between-subjects factor outcome:
positive v. negative) #2 (within-subjects factor reasonableness:ex ante v. ex post)
design. All participants were presented with the following first part of the vignette
and had to rate the doctor’s decision to prescribe a certain drug on a 7-point Likert
scale:
Ms Woods goes to the hospital since she has been experiencing severe abdominal pain for
several days. Dr Miller examines her carefully and concludes that she has an inflammation of
the intestines.
Dr Miller is thinking about prescribing a new drug called Fluxis, which has proven very
effective with the kinds of inflammation Ms Woods has. As most drugs, it has some possible
side-effects. According to recent studies, there’s a 3% probability that Fluxis will cause
damage to the liver.
Q: To what extent do you agree or disagree with the following claim:
42 M. Kneer
“It is reasonable for Miller to prescribe Fluxis.”(1: completely disagree –7: completely
agree)
Having responded to the reasonableness ex ante question, participants were ran-
domly assigned to one of the two outcomes (labels in square brackets omitted):
[Good outcome] Dr Miller prescribes the drug Fluxis to Ms Woods. The drug works
perfectly. The inflammation is cured within days, and Ms Woods has no side-effects
whatsoever.
[Bad outcome] Dr Miller prescribes the drug Fluxis to Ms Woods. The drug does not work
as hoped. The inflammation persists and Fluxis causes serious damage to Ms Woods’liver.
Q1: To what extent do you agree or disagree with the following claim:
“It was reasonable for Miller to prescribe Fluxis.”(1: completely disagree –7: completely
agree)
Q2: How much praise or blame, if any, does Miller deserve for prescribing Fluxis to Ms
Woods? (1: lots of praise, 4: neither praise nor blame, 7: lots of blame).

effect, those participants who did not respond that the probability of liver damage
was 3% (as specified in the vignette) were excluded.
¼
¼" ¼
Reasonableness on the Clapham Omnibus: Exploring the Outcome-Sensitive... 43
Fig. 4 Mean reasonableness ex ante and ex post across outcomes. Error bars denote standard error
of the mean
5.3 Results
5.3.1 Within-Subjects Results
In the positive outcome condition, people deemed it significantly more reasonable to
prescribe the drug ex post (M ¼5.98, SD ¼1.38) than they did ex ante (M ¼5.45,
SD ¼1.40), paired-samples t-test, t(1,61) ¼"4.40, p<.001, Cohen’sd¼.56, a
medium-sized effect. In the negative outcome condition, people deemed it signifi-
cantly less reasonable to prescribe the drug ex post (M ¼4.51, SD ¼1.76) than they
did ex ante (M ¼5.48, SD ¼1.27), t(1,62) ¼7.04, p<.001, Cohen’sd¼.89, a
large effect, see Fig. 4.
5.3.2 Between-Subjects Results
Across conditions, outcome had a significant impact on reasonableness ex post,
independent-samples t-test t(1,123) ¼5.28, p<.001, Cohen’sd¼.94, a large
effect.
50
The same held true for blame, t(1, 123) ¼"6.06, p<.001, Cohen’s
d1.08, a large effect, see Fig. 5.
51
50
Unsurprisingly, there is no significant difference across conditions for reasonableness ex ante
(t(1,123) ¼".103, p¼.918), since the two conditions are—at that point—still identical (outcomes
have not yet been revealed).
51
As expected, all contrasts remain strongly significant when one leaves those participants who did
not respond with the exact probability of a harmful outcome in the dataset (N ¼157). There was no
significant effect of outcome on perceived probability post hoc,t(1,155) .715, p.476. This

suggest that, in explicitly specifying the probability in the scenario, we managed to limit its potential
effect even without excluding all those who participants who did not report it correctly.
44 M. Kneer
Fig. 5 Mean reasonableness ex post and mean blame/praise attributions across conditions (positive
v. negative outcome). Error bars denote standard error of the mean
5.4 Discussion
Replicating the between-subjects results from Experiments 1 and 2, we found that
decisions leading to a negative consequence are judged less reasonable post hoc than
those that lead to positive consequences. Replicating the within-subjects results from
Experiment 2, people judge the reasonableness of a decision producing harm as less
reasonable ex post than ex ante (before the outcome was known), and vice versa for
positive outcomes. These findings suggest that the folk application of the concept of
reasonableness is not subject to a performance error, but that the folk work with a
concept of reasonableness which is outcome-dependent. Is this outcome-dependence
driven by a hindsight effect, as some scholars like Kamin and Rachlinski
52
suggest?
Perhaps, in certain situations, this might also be the case. However, even when one
carefully controls for potential hindsight bias, we find a strong direct effect of
outcome on reasonableness.
6 Conclusion
In their capacity as fact-finders determining what is, and is not, reasonable, jurors,
we said, must have two types of competences: They must have concept competence,
i.e. understand and use the legal concept of reasonableness. And they must possess
52
Kamin and Rachlinski (1995).

Reasonableness on the Clapham Omnibus: Exploring the Outcome-Sensitive... 45
application competence, that is, they must not fall prey to cognitive biases when
deciding what is reasonable and what is not.
The legal concept of reasonableness is outcome-insensitive and dynamically
invariant. In other words, whether a decision produces a positively or negatively
valenced outcome must not have an impact on whether it is judged reasonable, and a
decision that is reasonable ex ante remains reasonable ex post, even if it produces
undesirable consequences. In line with previous, vaguely related studies, however,
all three experiments suggest that folk attributions of reasonableness are strongly
sensitive to outcome. In principle, this could be due to one of two reasons: First,if—
as assumed by the law—the legal concept of reasonableness just is the folk concept
of reasonableness, then the potential jurors fall prey to an outcome bias, as they tend
to also do in the attribution of mens rea. Or else, second, folk ascriptions might
actually be true to the folk concept of reasonableness—a concept that is very
different from the legal concept of reasonableness and from what the law takes the
folk concept to be. Experiments 2 and 3 suggest that this is in fact the case: Even in
within-subjects designs, people find it perfectly perfectly appropriate to revise their
ex ante assessments of reasonableness ex post. It thus appears that outcome-
sensitivity is built into the ordinary language expression “reasonable”and the
concept it denotes. The law seems to wrongly assume conceptual alignment and
unwittingly operate with a technical (outcome-insensitive) concept of
reasonableness.
Hardnosed advocates of the bias interpretation could, broadly in line with Kamin
and Rachlinski,
53
object to our preliminary conclusion in the following way: What
is, and is not reasonable is (at least frequently) tied to the notion of risk, i.e. the
product of probability and severity of outcome. Given that probability assessments
are subject to the exceptionally robust and pervasive hindsight bias, the variation of
reasonableness assessed ex ante vs. ex post might simply be a consequence of
distorted probability judgments. The findings from Experiment 3, however, suggest
that this objection is on the wrong track.
54
Even if we hold perceived probability
fixed across the ex ante context of decision and the ex post context of assessment on
the one hand, as well as across differently valenced outcomes on the other, our
results from Experiments 1 and 2 replicate perfectly. Naturally, this leaves open the
possibility that the outcome sensitivity and dynamic variance of reasonableness is
exacerbated by the hindsight bias in the many situations where probabilities are
unknown or hard to quantify. However, it demonstrates that the outcome-sensitivity
and dynamic variance of reasonableness does not constitutively depend on perceived
probability. The folk concept seems to be directly sensitive to outcomes.
53
Kamin and Rachlinski (1995).
54
It does not matter whether the effect of hindsight on perceived probability is viewed as a bias or as
arational heuristic (as discussed in Sect. 4.4). The point here is simply that the effect of outcome on
reasonableness is, at least in parts, independent of probability judgments (no matter how they are
interpreted).

¼
46 M. Kneer
Over the last two decades, considerable effort has been devoted to a better
understanding of what the legal concept of reasonableness does and should mean,
and how potential pitfalls in the application of the reasonable person standard can be
avoided. Perhaps the subjective/objective and the normative/descriptive debates
have received the most attention, and they are, without a doubt, of great practical
importance. The potential outcome-sensitivity of reasonableness ascriptions, by
contrast, has received next to no attention, and there is no literature on the dynamic
sensitivity of reasonableness at all. This constitutes a serious shortcoming because
there is no legal case invoking the reasonable person standard where these two
features do not matter. It is thus hard to overstate the practical importance of the
present findings and more research—both theoretical and empirical—is required to
explore how conceptual alignment can be achieved so as to prevent further legal
injustice due to the law’s misunderstanding of the folk concept of reasonableness.
References
Alexander L, Ferzan KK, Morse SJ (2009) Crime and culpability: a theory of criminal law.
Cambridge University Press
Alicke MD (1992) Culpable causation. J Pers Soc Psychol 63(3):368
Alicke MD (2000) Culpable control and the psychology of blame. Psychol Bull 126(4):556
Alicke M, Rose D (2010) Culpable control or moral concepts? Behav Brain Sci 33(4):330
Armour JD (1994) Race ipsa loquitur: of reasonable racists, intelligent Bayesians, and involuntary
Negrophobes. Stanford Law Rev:781–816
Baron M (2011) The standard of the reasonable person in criminal law. In: Duff RA, Farmer L,
Marshall SE, Renzo M, Tadros V (eds) The structures of criminal law. Oxford University Press,
pp 11–36
Baron J, Hershey JC (1988) Outcome bias in decision evaluation. J Pers Soc Psychol 54(4):569
Beebe JR, Buckwalter W (2010) The epistemic side-effect effect. Mind Lang 25(4):474–498
Beebe JR, Jensen M (2012) Surprising connections between knowledge and action: the robustness
of the epistemic side-effect effect. Philos Psychol 25(5):689–715
Berger SR (1977) Court awarded attorneys’fees: what is reasonable. Univ Pa Law Rev 126:281
Bourgeois-Gironde S, Kneer M (2018) Intention, cause, et responsabilité: Mens Rea et effet
Knobe. In: Ferey S, G’Sell F (eds) Causalité, responsabilité et Contribution à la dette de
réparation. Une Perspective interdisciplinaire, Bruylant
Brown T (2022) Minding accidents. https://dc.law.utah.edu/cgi/viewcontent.cgi?article¼1325&
context scholarship
Cova F (2016) The folk concept of intentional action: empirical approaches
Cushman F (2008) Crime and punishment: distinguishing the roles of causal and intentional
analyses in moral judgment. Cognition 108(2):353–380
Donelson R, Hannikainen IR (2020) Fuller and the folk: the inner morality of law revisited. In:
Lombrozo T, Knobe J, Nichols S (eds) Oxford studies in experimental. Oxford University Press
Dressler J (1995) Understanding criminal law. Matthew Bender
Enoch D (2007) Luck between morality, law, and justice. Theor Inq Law 9(1):23–59
Enoch D (2010) Moral luck and the law. Philos Compass 5(1):42–54
Enoch D, Marmor A (2007) The case against moral luck. Law Philos 26(4):405–436
Epstein RA (1992) The path to “The TJ Hooper”: the theory and history of custom in the law of tort.
J Leg Stud 21(1):1–38
Feltz A (2007) The Knobe effect: a brief overview. J Mind Behav:265–277

6¼
Reasonableness on the Clapham Omnibus: Exploring the Outcome-Sensitive... 47
Fletcher GP (1970) Theory of criminal negligence: a comparative analysis. Univ Pa Law Rev 119:
401
Fletcher GP (1985) The right and the reasonable. Harv Law Rev:949–982
Frisch LK, Kneer M, Krueger JI, Ullrich J (2021) The effect of outcome severity on moral
judgement and interpersonal goals of perpetrators, victims, and bystanders. Eur J Soc Psychol
51(7):1158–1171
García Olier J, Kneer M (in preparation) What drives the Knobe effect: valence or severity?
Gardner J (2001) The mysterious case of the reasonable person. Univ Toronto Law J 51(3):273–308
Gardner J (2015) The many faces of the reasonable person. Law Q Rev 131(1):563–584
Güver L, Kneer M (2022) Causation and the silly norm effect. In: Magen S, Prochownik K (eds)
Advances of experimental philosophy of law (forthcoming). Bloomsbury Press
Hannikainen IR, Tobia KP, De Almeida GDF, Donelson R, Dranseika V, Kneer M et al (2021) Are
there cross-cultural legal principles? Modal reasoning uncovers procedural constraints on law.
Cogn Sci 45(8):e13024
Hartman RJ (2017) In defense of moral luck: why luck often affects praiseworthiness and blame-
worthiness, vol 38. Taylor & Francis
Hoffman M (ms.) The psychology of the trial judge
Hurd HM, Moore MS (2002) Negligence in the air. Theor Inq Law 3:333
Kamin KA, Rachlinski JJ (1995) Ex post ex ante. Law Hum Behav 19(1):89–104
Kelman M (1991) Reasonable evidence of reasonableness. Crit Inq 17(4):798–817
Kneer (in preparation-a) The folk concept of practical rationality
Kneer (in preparation-b) “Reasonable”and “rational”are assessment-sensitive
Kneer M, Bourgeois-Gironde S (2017) Mens rea ascription, expertise and outcome effects: profes-
sional judges surveyed. Cognition 169:139–146
Kneer M, Machery E (2019) No luck for moral luck. Cognition 182:331–348
Kneer M, SkoczeńI (2022) Outcome effects, moral luck and the hindsight bias. Cognition
(to appear). https://doi.org/10.2139/ssrn.3810220
Kneer M, Hannikainen IR, Zehnder M, Almeida G, Aguiar F, Bystranowski P, Dranseika V, Janik
BM, Garcia Olier J, Güver L, Liefgreen A, Tobia K, Próchnicki M, Rosas A, SkoczeńI,
Strohmaier N, Struchiner N (in preparation) Outcome effects on mental state ascriptions across
cultures
Knobe J (2003a) Intentional action and side effects in ordinary language. Analysis 63(3):190–194
Knobe J (2003b) Intentional action in folk psychology: an experimental investigation. Philos
Psychol 16(2):309–324
Knobe J, Shapiro SJ (2021) Proximate cause explained: an essay in experimental jurisprudence.
Univ Chic Law Rev 88:165–236
Lee CKY (2007) Murder and the reasonable man: passion and fear in the criminal courtroom, vol
37. NYU Press
Lidén M, Gräns M, Juslin P (2019) ‘Guilty, no doubt’: detention provoking confirmation bias in
judges’guilt assessments and debiasing techniques. Psychol Crime Law 25(3):219–247
Low PW (1987) The model penal code, the common law, and mistakes of fact: recklessness,
negligence, or strict liability. Rutgers Law J 19:539
MacFarlane J (2014) Assessment sensitivity: relative truth and its applications. OUP, Oxford
Macleod JA (2019) Ordinary causation: a study in experimental statutory interpretation. Ind Law J
94:957
Martin JW, Cushman F (2016) The adaptive logic of moral luck. The Blackwell companion to
experimental philosophy, pp 190–202
Miller AD, Perry R (2012) The reasonable person. NYUL Rev 87:323
Moore MS, Hurd HM (2011) Punishing the awkward, the stupid, the weak, and the selfish: the
culpability of negligence. Crim Law Philos 5(2):147–198
Moran M (2003) Rethinking the reasonable person: an egalitarian reconstruction of the objective
standard. Oxford University Press

48 M. Kneer
Moran M (2010) The reasonable person: a conceptual biography in comparative perspective. Lewis
Clark Law Rev 14:1233
Nadelhoffer, T. (2006) Bad acts, blameworthy agents, and intentional actions: some problems for
juror impartiality. Philos Explor 9(2), 203–219
Nagel T (1979) Mortal questions, vol 89(3). Cambridge University Press
Nourse V (2008) After the reasonable man: getting over the subjectivity/objectivity question. New
Crim Law Rev 11(1):33–50
Prochownik KM (2021) The experimental philosophy of law: new ways, old questions, and how not
to get lost. Philos Compass 16(12):e12791
Prochownik K, Krebs M, Wiegmann A, Horvath J (2020) Not as bad as painted? Legal expertise,
intentionality ascription, and outcome effects revisited
Rachlinski JJ (1998) A positive psychological theory of judging in hindsight. Univ Chic Law Rev
65(2):571–625
Ripstein A (2001) Equality, responsibility, and the law. Cambridge University Press
Simons KW (2002) Dimensions of negligence in criminal and tort law. Theor Inq Law 3(2)
Simons KW (2008) Self-defense: reasonable beliefs or reasonable self-control? New Crim
Law Rev: Int Interdiscip J 11(1):51–90
Sommers R (2019) Commonsense consent. Yale Law J 129:2232
Stark F (2016) Culpable carelessness: recklessness and negligence in the criminal law. Cambridge
University Press
Strohmaier N, Pluut H, van den Bos K, Adriaanse J, Vriesendorp R (2020) Hindsight bias and
outcome bias in judging directors’liability and the role of free will beliefs. J Appl Soc Psychol
Struchiner N, Hannikainen I, Almeida G (2020) An experimental guide to vehicles in the park.
Judgm Decis Mak 15(3)
Teichman D (2014) The hindsight bias and the law in hindsight. In: Zamir E, Teichman D (eds) The
Oxford handbook of behavioral economics and the law. Oxford University Press, pp 354–373
Tobia KP (2018) How people judge what is reasonable. Ala Law Rev 70:293
Tobia KP (2020a) Law and the cognitive science of ordinary concepts. In: Handbook on law and the
cognitive sciences. Cambridge University Press, forthcoming
Tobia KP (2020b) Testing ordinary meaning: an experimental assessment of what dictionary
definitions and linguistic usage data tell legal interpreters, vol 133. Harv Law Rev
Tilley CC (2017) Tort Law Inside Out, 126 YALE L.J. 1320, 1327
Tobia K (2022) Experimental jurisprudence. Univ Chic Law Rev 89 (forthcoming)
Wells C (1982) Swatting the subjectivist bug. Crim Law Rev:209–220
Westen P (2008) Individualizing the reasonable person in criminal law. Crim Law Philos 2(2):
137–162
Williams B (1993) Moral luck. In: Statman D (ed) Moral luck. SUNY Press
Wittlin M (2016) Hindsight evidence. Colum Law Rev 116:1323
Zipursky BC(2014) Reasonableness in and out of negligence law. Univ Pa Law Rev 163:2131