AUTONOMY, REASON AND BIAS IN CONTRACT LAW
Jaap Hage
jaap.hage@maastrichtuniversity.nl
www.jaaphage.nl
Abstract
The cognitive sciences provide us in high frequency with insights in the functioning of the human
mind. One of these insights is that human decision making is often not rational. The law has tools to
deal with lapses of rationality, but it works on the basis of an image of mankind according to which
humans are, and typically act, rationally. This raises the question whether law should be more
adaptive to the stream of new insights produced by the cognitive sciences. The present article aims to
answer this general question for an important field of law: the law of contract. It shows that there are
good reasons to assume that the conclusion of a contract is often not a rational event.
The law can follow two strategies in dealing with the insight that contracts, and – extrapolating a bit
– juridical acts in general – are often not created rationally. One strategy is to use the tools which law
already has, for dealing with irrationality for newly discovered kinds of cases such as the anchoring
bias. Another strategy is to redesign substantial parts of the law to take into consideration that
subjective preferences often do not reflect objective interests. Which strategy is the best depends on
the frequency of situations in which subjective preferences do not reflect objective interests.
Keywords: anchoring bias, autonomy, contract, interest, preference, rationality, reliance
1. Introduction
This article is part of a larger project that aims to take stock of recent developments in the cognitive
sciences and their possible implications for law. The focus of the present contribution is on the
implications for contract law if it proves to be the case that human decision making in general, and in
contracting in particular, is less rational than is often assumed. A possible consequence of this finding
might be that contract law should be completely rewritten, or perhaps even abolished.
The autonomy of legal subjects is the cornerstone of private law in general, and contract law in
particular. Private law deals with the mutual relations between private legal subjects. Originally1 the
role of state organs was confined to the enforcement of the legal positions and relations that legal
subjects created themselves. For example, legal subjects determine who owns what, and the role of
the state is confined to protecting those property rights. The much-heralded freedom of contract
provides another illustration. The law of contract makes it possible for cooperating legal subjects to
modify their legal positions by undertaking contractual obligations. The freedom of contract involves
both the power of legal subjects to create the obligations that they want to undertake, and the
permission to do so.2 By using this freedom, legal subjects express their autonomy3, and this
1
2
After the heyday of the night-watch state and free-market liberalism, the interference of the governance in
what were originally considered to be private relations has increased. As a result, topics that used to belong
to private law only, have partly become the object of public law too. Labor relations provide an illustration
of this shift from private to public.
Hage 2010.
-1-
empowerment of legal su je ts auto o
goes ha d i ha d ith the assu ptio that this
4
autonomy is exercised rationally. As we will see in section 2.4, support for this link between
autonomy and rationality in actual contract law can be found in the situations where the law limits
the freedom of contract.
Psychological and neurological research of the last decennia has shown that human decision making
is often less rational than was traditionally assumed. Investigations (inspired) by Festinger, Ryle and
Gazzaniga have made it clear that humans sometimes confabulate the motives that made them act.5
In a number of papers, Kahnemann and Tversky have shown that human judgments are often based
on heuristics that are less than fully rational, and that they suffer from biases.6 Moreover, recent
research into the brain has begun to reveal the neural mechanisms that implement decision making;
specifically, it has revealed that the parts of the brain closely related to our emotions, actually play
an important role in the decision-making process.7
The assumption that adult and mentally healthy persons have always had the capacity to act
rationally lies at the heart of contract theory. However, in light of recent these psychological and
neurological findings, I will try to show in this article, that a thorough reconsideration of contract
theory and of the foundations of contract law is necessary. To this purpose, I will discuss the
a ho i g ias . This ias is one of the biases that were identified by Kahnemann and Tversky. I will
argue that the existence of this bias and of other biases should at least lead to an adaptation of the
doctrine of contract formation. Perhaps it is even better to completely replace contract law by rules
that immediately aim to promote the interests of legal subjects, rather than empowering legal
subjects to promote their interests on the basis of their subjective preferences.
In section 2 I will argue that the central role of the will or intention of contract partners is best
considered as a stand-in for the interests of contract parties, and also that the present law is
primarily focused on interests rather than on autonomy, but that this focus has been inconsistent.
Section 3 discusses the anchoring bias, and shows that many more contracts than only those based
on misrepresentation, duress, mistake, or fraud, are the result of irrationality.
Section 4 deals with the consequences the law should attach to the fact – if there is one – that the
behavior of legal subjects is often irrational. In a conservative approach, irrationality raises the
question whether and under which conditions contracts that were irrationally concluded should be
avoidable. In a more radical approach, the question becomes whether the social practice of
contracting should even be supported by law. This article is summarized and concluded in section 5.
2. The justification of contractual obligations
Traditionally, contract law assigns an important role to the will or intention of the contracting parties,
when it comes to determining whether a contract was concluded. The point of the present section is
to argue that the will or intention of the contract partners is a heuristic stand-in for the interests of
3
4
5
6
7
Larenz and Wolf 2004, p. 393.
The lose elatio ship et ee auto o
a d atio alit is a o e sto e of Ka t s philosoph of p a ti al
easo . It is ell o th e plo i g hethe this ele e t of Ka t s philosoph has i flue ed the
development of contract doctrine in continental Europe, but this possible connection will not be elaborated
here.
Wegner 2002, pp. 171-186.
Kahnemann 2011.
Damasio 1994; Cushman, Young and Greene 2010; Sapolsky 2017.
-2-
the partners. In the end, the interests count, not the will or intention. This is already the case in
contract law as it presently is, although it is not always clearly visible. We can discover the role of
will, intention, and interests, if we ask what justifies the existence of contractual obligations. Why
should the partners in a contract have obligations, merely because they promised each other to do
something?
Let me introduce an example that will facilitate the discussion that follows. Anne offers her house for
sale to Bob fo € 0,000. Bob makes a counter-offe of €
,
, hi h is a epted
Anne. In this
example, the questions that need to be answered are why Anne has the obligation to transfer the
house to Bob and why Bob has the obligation to pay Anne €
,
. The answers to these questions
belong to one of three kinds of answers. The first kind emphasizes the mutual reliance that is created
by making contractual promises. For example, Anne promised Bob to transfer her house and Bob s
reliance on this promise justifies that Anne has the obligation to perform. The second kind of answer
refers to the autonomy of the contract partners. Anne s auto o
gi es he the po e to u de take
contractual obligations by expressing her will to do so. Anne s a epta e of Bob s ou te -offer is an
exercise of this autonomy, by means of which she undertook the obligation to transfer the house.
This justifies the existence of the obligation to transfer. The third kind of answer does not focus on
concrete contracts, but on the social practice of contracting, which has been laid down in contract
law. Concrete contractual obligations can then be justified by invoking the practice: there is a rule
which attaches the obligation to transfer to the contract.
In the subsections 2.1 and 2.2 it will be argued that the first of these two reasons do not cut ice. The
third reason, invocation of the social practice of contracting, is the right one, but presupposes a
justification of this very practice. This justification is the topic of the subsections 2.3 – 2.5, and in
studying it, we will discover that interests play a more important role than will or intention.
2.1
Reliance on specific promises
One classic answer to the question of why contract partners derive obligations from their contract is
that a contractor induces her partner to do something which he would otherwise not have done, by
means of the contract. The contract creates reliance, and this reliance is awarded with an obligation
on the partner who induced it.8 At first sight, this reliance theory of contractual obligation has its
attractions, but on closer inspection the theory fails. It will turn out that reliance can only be justified
if contracts lead to obligations. However if that is the case, the reliance cannot be the reason why
contracts create obligations.
Suppose that Anne told Bob that her house is somber when it is cloudy and that on a cloudy day she
would happily transfer her house to anyone offering at least €
,
. On a cloudy day, Bob offers
Anne €
,
a d elies o it that Anne will transfer the house. Based on this reliance, Bob sells his
o
house to C fo €
,
, hi h is considerably less than the market value. Is Anne under an
obligation towards Bob to transfer her house to him, agai st pa e t of €
,
?
Reliance only leads to obligations if it is justified. This raises the question of under which
circumstances reliance counts as justified and leads to obligations. Perhaps it might be argued that in
the above example, Bob was justified in relying on Anne s e a k. Pe haps Anne is the kind of
woman whose remarks are a reliable predictor of what she will do. However, reliance on a
trustworthy predictor of future behavior is not the relevant kind of reliance for the creation of
8
Atiyah 1986; see also Smits 1995.
-3-
obligations. Bob should not rely on a mere prediction of Anne s futu e eha io . This becomes
immediately clear if the prediction is not based on the behavior of Anne herself, but on that of some
other person. “uppose that Ge aldi e, A e s siste , k o s A e ell, a d a
ake t ust o th
p edi tio s of A e s futu e eha io . Bo s elia e that A e ill t ansfer the house would, legally
speaking9, not be justified if Geraldine predicted the transfer.
Reliance based on trustworthiness would not even be justified if Anne herself would have made this
prediction of her future behavior. Bob should rely on it with justification, not merely that Anne will
transfer her house, but that Anne placed herself under an obligation to transfer the house. That is
only possible, however, if Anne could place herself under this obligation independent of Bob s
reliance. However, if Anne could do this, it means that the obligation does not depend on the
reliance, and is not justified by it. It is rather the other way around: the reliance is justified by the
obligation which came into existence independently of the reliance. Because contracts lead to
obligations, a contract partner is justified in relying on the assumption that other contract party will
perform.
2.2
Autonomy, will and intention
The second kind of answer to the question of why contracts lead to obligations, is that because
contractual obligations are undertaken voluntarily and they are willed by the contract partners, they
are in that sense expressions of their autonomy.
The problem with this answer is that it is either vague, or unconvincing. To see how it can be vague,
one only needs to consider the circumscription of autonomy as given in the Stanford Encyclopedia of
Philosophy:
...to be autonomous is to be one's own person, to be directed by considerations, desires,
conditions, and characteristics that are not simply imposed externally upon one, but are part of
hat a so eho e o side ed o e's authe ti self. 10
The reference to what can somehow e o side ed o e s authentic self illustrates that this notion of
autonomy is in need of elucidation. In the absence of such elucidation, this kind of autonomy cannot
play a serious role in the justification of contractual obligations.
Similarly problematic is the view that contractual obligations are justified because they were willed.
The very notion of the will is vague. The will is not a conscious event. For instance, it would not make
se se to sa A i ute ago, I willed to sell my house .11 A better view is that the will is a disposition to
act in a particular way. It akes se se to sa I al a s illed a ted to ake a t ip to Ve i e . That
would mean that the person with this will, will make the trip as soon as a suitable opportunity
presents itself. However, if the will is such a disposition, it is not clear why having a will justifies a
contractual obligation. The person who expresses a will to take a trip to Venice does not need a
contract for his motivation to make the trip. Moreover, he has no interest in concluding such a
contract: why add an obligation to an already existing motivation? The other contract party may have
an interest in the existence of an enforceable obligation and therefore in a contract. However, it does
9
10
11
The alternative for legally justified reliance is epistemically justified reliance. The very point of the present
example is that if a person is epistemically speaking justified in relying on something, this does not
automatically mean that this person is also legally speaking justified in relying. Legally justified reliance
presupposes contract law, while epistemically justified reliance does not.
Christman 2017.
Ryle 1949, pp. 61-67.
-4-
not seem to make sense that this obligation would exist for the mere reason that the first contractor
was motivated to make the trip. We can conclude that the will that some action will be performed is
not a good reason to justify the existence of an obligation.
Perhaps the view that contractual obligations exist because they were undertaken voluntarily12 is
more promising than the idea that these obligations exist because the obligated behavior was willed.
The idea that contractual obligations are undertaken voluntarily may seem to be helpful, at least if
the notion of voluntary action is used in the traditional sense of an action not being caused by
conditions which should not have caused it.13 However, it is unclear why the fact that an obligation
was undertaken voluntarily shows that the obligation itself is justified. If an agent undertakes an
obligation, then this agent has an obligation, no matter whether the obligation was undertaken
olu ta il . U de taki g is i this espe t a su ess-verb: one can only undertake a task or an
obligation if one has this task or obligation after having undertaken it. If voluntariness is a condition,
it is a condition for the capability to undertake, not for the existence of what was undertaken.
If this sounds too abstract, perhaps the following example can illustrate my point. If a mountaineer
climbs a mountain, she will reach the top by definition of hat li i g a ou tai
ea s. It does
not matter for this purpose whether she did this voluntarily. Whether she climbed voluntarily or not,
does not matter for the issue of whether she reached the top. However, on some definitions of
climbing, involu ta
li i g is i possi le. A pe so
ho li ed i olu ta il
as a tuall
dragged to the top of the mountain. The voluntary nature of the event determines whether we call it
climbing or being dragged, but not whether the mountain top was reached. Similarly, the
voluntariness of undertaking an obligation does not determine whether an obligation came into
e iste e. Ho e e , it a e ele a t fo
hethe the e e t ould e alled u de taki g a
obligation, rather than having an obligation imposed.
The same reason which makes that the voluntariness of undertaking obligations does not justify the
existence of these obligations also applies to the intentionality of undertaking obligations. An agent
can contract with the intention to undertake in that way a particular obligation. The precise contents
of the obligations are strongly influenced by the intentions of the contract partners. However, the
fact that intentions influence the contents of the obligations, does not prove that these intentions
also justify the existence of the resulting obligations. Contracting with the intention to create
particular obligations only makes sense if the contracting parties already know that their behavior
will lead to the intended obligations. This knowledge presupposes the existence of a social practice
according to which contracting leads to obligations. If this practice exists, it justifies the existence of
contractual obligations. Then however, there is no need for additional intentions.
Perhaps the mountain climbing example can make this clearer again: intentionally climbing a
mountain is only possible for a person who knows that her behavior will lead her to the mountain
top. An intentional mountain climber can only know this, if his climbing will bring him actually to the
top. This means that reaching the top cannot depend on the intention to do so.
I conclude that specific contractual obligations are not justified by the autonomy, will, or intention of
the contracting parties. Together with the finding that these obligations are not justified by reliance
either, the conclusion must be that the traditional theories of the justification of contractual
obligations fail if they are interpreted as theories about the justification of specific obligations.
12
13
Raz 1972.
The alternative is that an action is voluntary if it is based on a free will. When it comes to vagueness, this
notion of a free will can compete with the notion of autonomy: we hardly have a clue what it might be.
-5-
However, this does not mean that they also fail as theories of why the social practice of contracting is
justified.
2.3
Preferences and interests
The social practice of contracting consists of a set of rules that specify which kinds of actions count as
contracting, which obligations or other legal consequences result from valid contracts, and which
legal steps can be undertaken in case a contractual obligation is not performed. For my present
purposes, the details of these rules – which vary from one jurisdiction to another - are not relevant.
Even if there exists a social practice of contracting which justifies contract-based obligations, the
uestio
a e aised hethe su h o ligatio s a e eall justified. The a s e depe ds o
whether the social practice itself is justified. If the practice is bad, it cannot justify obligations, but if
the practice is a good thing, it can justify obligations. It turns out that if contractual obligations are
justified by invoking the social practice of contracting, the success of this justification depends on the
justification of the practice itself. This leads us to the question whether the existence of a social
practice of contracting is justified. In answering this question, both reliance and autonomy play a
role.
It is efficient i the e o o i se se of effi ie
, if there is a social practice which makes it possible
for agents to undertake obligations in exchange for obligations undertaken by other agents. To see
why, it is useful to introduce the notion of a Pareto-improvement, named after the Italian economist
Vilfredo Pareto. A Pareto-improvement in the division of goods (in a broad sense) over a population
is a change in the distribution which makes at least some members of the population better off,
while it makes nobody worse off. A voluntary exchange will normally14 lead to a Pareto improvement
for both15 parties involved. For example, if Anne alues he house o €
,
, hile Bob values the
house o €
,
, a sale of the house f o Anne to Bob fo a a ou t i et ee €
,
a d
€
,
ill lead to a Pa eto-improvement. If goods are distributed over a population in such a way
that no Pareto-improvement is possible anymore – any exchange will make somebody worse off –
the distribution is said to be Pareto-efficient.16
Normally, in a voluntary sale both parties will improve their situations if the exchange takes place
immediately, and nobody runs the risk that the other party will not perform. However, it is often
better if the performance of one or both of the parties takes place later. For instance, Bob only needs
the house three months from now, but if he can rely on it that the house will be transferred then, he
can already sell his own house to Carol, and use the money to pay Anne. At the same time, Anne can
count on receiving the money after three months, and already book the cruise around the world that
she always wanted to make. Such delayed exchanges can only function in a reliable way, if there exist
mechanisms to enforce them. Reputation is one such mechanism, contract law is another. By making
enforcement of contracts and therefore also reliance possible, the practice of contracting facilitates
delayed exchanges and will therefore contribute to a Pareto-efficient division of goods and services
over society. At least, so it seems.
14
15
16
Late , I ill a gue that the lause o all is u iall i po ta t he e.
I ignore multi-party contracts, but including them would not change the argument considerably.
More
details
on
Pareto-improvements
and
Pareto-efficiency
can
be
found
https://en.wikipedia.org/wiki/Pareto_efficiency and on
http://www.econlib.org/library/Enc/bios/Pareto.html (last visited on 31-12-2017).
-6-
on
2.4
If preferences and interests do not match
Pareto-improvements have to do with making members of the population better or worse off. What
precisely makes a participant better or worse off is not specified in this characterization. There are
both subjective and objective tests for what makes an agent better off.17 The typical subjective test is
preference satisfaction18: the situation of an agent is improved if more of his actual preferences are
satisfied to a higher degree.19 A more objective variant of this test takes only informed preferences
into account.20 The typical objective test is based on the interests of an agent, regardless of whether
the agent prefers these interests to be satisfied. This test leaves the question open what precisely
ou t as a age t s i te ests. A t aditio al a s e to this uestion is that an agent has only one
ultimate interest: maximization of happiness.21 A related answer is the maximization of well-being.22
There are also pluralist theories according to which there are several interests which must somehow
be combined.23
Contracts are based on subjective preference satisfaction. They are voluntary transactions, and it
seems unlikely that an agent will voluntarily engage in a transaction which violates his actual
preferences. Legal subjects do not contract to realize the preferences which they would have under
ideal circumstances. Either they do not know what these ideal preferences would be, and can
therefore not act upon them, or they know their ideal preferences, but then these have become their
actual preferences, since ideal preferences are the preferences which persons would actually have
under ideal circumstances. Further, legal subjects do not contract to realize what is good for them,
unless they have the subjective preference to do so.
There are several possible reasons why the subjective preferences of a person do not reflect his real
interests. First, a subjective preference may be short-lived, and perhaps last only a few minutes. A
far-fetched example would be an agent who is under hypnosis and has a preference which
disappears as soon as the hypnosis ends. Second, the agent may suffer from a general lack of
capacity to form adequate subjective preferences. Obvious examples are young children, some aged
people, and persons suffering from a mental handicap or illness. Third, a subjective preference may
be badly informed and would have been different if well-informed. An example would be if Bob
wants to buy Anne s house, e ause he elie es that the e ill e a good ail a o e tio f o
the house to his working place. However, the railway company intends to end this railway
connection. If Bob had known this, he would not have wanted to buy the house.
All three examples have in common that an agent has preferences which do not (necessarily) reflect
his interests. Often, the law takes this discrepancy into account. A juridical act performed under
hypnosis will sometimes be avoidable, and the same holds for juridical acts performed by very young
persons, o
pe so s ho a e fo othe easo s ot apa le to fo
thei
ill . This last
17
18
19
20
21
22
23
From here on, I will ignore the possibility that an agent is worse off as the result of some transaction, but
the argument on when an agent is better off applies mutatis mutandis also to being worse off.
Hansson and Grüne-Yanoff 2017.
I will ignore the trade-off between the satisfaction of more preferences and a higher degree of satisfaction
of less preferences. For a discussion of this trade-off, see Hage 2005.
See Hansson and Grüne-Ya off
o filte ed p efe e es.
Bentham 1789, chapter I, section 2.
See the discussion of welfarism in Crisp 2017, and Griffin 1986.
Mason 2015.
-7-
formulation is telling, since almost all persons are capable to form a will.24 The problem in the
mentioned cases is not that people cannot form a will, but that their will may not reflect their
interests. From the fact that the law attaches consequences to situations in which the subjective
preferences of contract partners do not reflect their real interests, we may conclude that what
counts in law is not only people s subjective preferences, but also their interests.
Perhaps, we should even go one step further. It is possible to interpret the importance which the law
attributes to the subjective preferences of persons as an indication that these preferences are seen
as evidence for their interests. In this interpretation of law, only interests count, and the subjective
preferences that people have are relevant as indicators of what their interests are. Indeed, it is easier
to establish subjective preferences than to establish interests, and substituting subjective
preferences for real interests may in many cases be a responsible epistemic strategy.25 However, as
soo as it e o es lea that the su je ti e p efe e es a de iate f o a pe so s real interests,
the law provides tools to correct the legal consequences of juridical acts based on unrepresentative
preferences. An example from Dutch law would be the following. If a 13-year child old child uses
their savings to buy reasonably-priced study books, the contract will be normally valid. However, if
they use the money to buy a motor cycle on impulse, the contract can be vitiated.26
The tool that is most often used for this purpose is the possibility to avoid juridical acts that do not
serve the interests of the agents who performed these acts. The details obviously differ between
jurisdictions. In Dutch law, avoidability is – under conditions - available to minors, persons under
wardship, persons suffering from a mental disorder, o pe so s a ti g u de a defe t of o se t ,
such as mistake, fraud, threat and undue influence.27 28
If a contract does not serve the interests of one of the contract parties29, it will often be in his
interest to avoid the contract. If the other contract partner relied on his performance, this other
pa t s interests will typically be violated. For example, Bob receives word that the railway company
will end the train connection between his working place and Anne s house. The a uisitio of Anne s
house no longer serves Bob s i te ests, a d he a a t to a oid the sales o t a t fo the house.
However, Anne has booked the cruise around the world that she always wanted to make and if the
sale is discontinued, she may have a financial problem. More in general, if it becomes easy to avoid
contracts if it turns out that they do not serve the interests of one of the contract parties, it becomes
more difficult to rely on contracts. Contracting becomes less attractive, fewer contracts will be
concluded, and fewer Pareto-improvements will be realized. Allowing avoidance to protect private
24
25
26
27
28
29
A caveat is i pla e he e, e ause the e p essio fo
o e s ill suggests that ill-formation is under
control of the agent. Even if this is so eti es the ase, e ofte it is ot. O e s ill is typically something
that just happens to a person.
As Jan Smits pointed out to me, in a meta-ethical non-cognitivist view, there may not be real interests over
and above subjective preferences. In that case the satisfaction of subjective preferences coincides with the
satisfaction of real interests. Such a non-cognitivist view is in my opinion not very attractive, however, since
it does not even distinguish between subjective and informed preferences.
Art. 1:234 S. 3 Burgerlijk Wetboek.
Smits 2014, pp. 91-100 and 159-175.
Notably, the conditions for avoidance typically do not directly mention the interests of the person who is
empowered to avoid. Instead the conditions tend to refer to the mental condition of the agent which made
it difficult for him to safeguard his interests himself.
If a contract disadvantages both parties, they can terminate the contract consensually.
-8-
interests damages the public interest, and this damage needs to be balanced against the interest of
one of the contract partners in avoiding the contract.30
2.5
Conclusion on the justification of contractual obligations
We have considered three possible justifications for the existence of contractual obligations. The first
two justifications invoke reliance and autonomy and we have seen that they both presuppose the
existence of the very obligations that they aim to justify. Reliance that a contract will be performed is
only justified if it may be assumed that a contractual obligation exists. The autonomy to create
obligations by means of contracts presupposes the possibility to create obligations by contracting.
The third justification invokes the social practice of contracting. A particular contractual obligation is
justified because the contract is an event that according to this practice leads to obligations. The
social practice of contracting can only justify particular obligations if this practice actually exists and if
it is justified itself. Both reliance and autonomy play their role here again, but this time in an indirect
way. The protection of reliance is a reason for recognizing contractual obligations because this
recognition and the enforcement that goes with it, facilitates the use of this practice, which itself
leads to Pareto-improvements. Autonomy consists in part in empowerment, and the social practice
of contracting empowers legal subjects to arrange their own affairs by creating obligations.
However, the role of autonomy in this connection is in need of further justification. It turns out that
the law does not unconditionally recognize the autonomy of legal subjects to undertake contractual
obligations. The power to undertake these obligations is limited when there is reason to assume that
it will not be used in the interest of the subject exercising this power, as when he is a minor, or
mentally ill. This suggests that the aim of the law in advancing individual autonomy does not have
this autonomy as its final goal, but rather sees it as an instrument to promote individual interests.
Moreover, the use of this instrument is limited by the interests of other legal subjects.
We can see this in the possibility to avoid some contracts. The reasons to avoid a contract always
relate to the circumstances of concrete cases. The social practice of contracting, laid down in
contract law, provides rules which provide that, in general contracts lead to the obligations that the
parties intended to undertake by means of the contract. This practice is assumed to be justified
because it leads to a more Pareto-efficient society, and reliance pays an important role in this
connection. The justification of concrete obligations need not refer to other circumstances than that
the parties followed the procedure for undertaking contractual obligations. If one of the parties
nevertheless wants to argue that one or more of the obligations do not hold, it must refer to the
special circumstances of the concrete case. In this connection, an important argument is that the
intention of the party that wants to nullify the contract did not properly reflect his interests. This
means that if contracts are seen as reflections of the tension between will and reliance, the role of
reliance is most important for the justification of contracting as a social practice, while the will has its
most important function in justifying exceptions to the general practice in concrete cases. This role of
the will derives its importance from the fact that the intentions of contracting parties do not always
reflect the interests of these parties. In the next section I will use the anchoring bias as a tool to
30
Actually it is slightly more complicated, because it is also a public interest that private interests are
safeguarded. To determine what serves the public interest best, it is necessary to balance the decreased
number of Pareto-improvements against the better protection of private interests. It depends on many
circumstances what the best outcome of this balancing operation will be. There exists a lot of theory
dealing with this issue under the headings of duties to inquire and to inform. See Kronman 1978 and SeftonGreen 2005.
-9-
demonstrate that this discrepancy between intention and interest is more common than used to be
assumed.
3. The anchoring bias
The anchoring bias involves a situation in which people who must answer a question or take a
decision that involves quantities, tend to let their answer or decision be influenced by a quantity that
was placed in their mind just before. This occurs even if this quantity is completely irrelevant for the
question or decision. For example31, a wheel of fortune was rigged to make 10 or 65 the only possible
outcomes. After a spin of the wheel – which gave 10 or 65 as result – students of the University of
Oregon were asked whether the percentage of African states as members of the UN was larger or
smaller than the result of the wheel. This planted either the number 10 or the number 65 in their
minds. After that, the students also had to answer the question what is, in their opinion, the
percentage of African members of the UN. Students who had 10 as outcome of the wheel estimated
the percentage of African members states on the average to be 25. However, students who started
from 65 estimated the percentage of African members states on the average to be 45. Somehow the
stude t s esti atio s e e d a
to a ds the out o e of the heel of fo tu e, e e though the
students were very much aware that this outcome was not relevant for their estimation. The result
of the heel fu tio ed as a a ho fo the judg e t a out the pe e tage of Af i a UN e e
states. This a ho iased the judg e t of the stude ts, a d that e plai s the a e a ho i g ias .
Other examples of the anchoring bias are the fact that the price a potential buyer is willing to offer
for a house is influenced by the asking price, or that a car driver who just left the highway and must
now drive at a lower speed, may be tempted to drive too fast.
It turns out that there are two explanations for the existence of the anchoring bias, which
complement each other. The one explanation is that there is often uncertainty about the correct
answer to a question or decision. An anchor influences the answer, because the respondent takes
the a s e as sta ti g poi t fo adaptatio s to a ds the o e t a s e . If Anne asks €
,
fo
her house, Bob ill thi k this a ou t too high, a d adapt it do
a d to €
,
. That ight e a
correct price (uncertainty) and since the reasoning started from a higher price, it stops at the highest
amount in a range of possibly correct prices. If Bob would have started the negotiations with an offer
of €
,
, Anne would have adapted that amount upwards until the range of possibly correct
p i es has ee ea hed. Let us assu e that this a ge goes f o €
,
to €
,
. The , if
Anne sta ts adapti g f o €
,
, she ill a i e at €
,
as the fi st possi le o e t p i e. It is
well possible that she will give this as a counter-offer, which may be accepted by Bob. In this way, the
first price mentioned in the negotiations about the sale of a house influences the price for which the
house will be sold. It turns out that this anchoring effect even influences the estimations of real
estate agents who are aware of the existence of the anchoring bias.
The second explanation for the existence of the anchoring bias is that mentioning an amount before
a question is asked, somehow influences the answer to the question by association. Where the
anchor exercises a conscious influence on the answer in the first explanation, it has an unconscious
influence in the second explanation. The example of the percentage of African UN members
illustrates this. The students who had to estimate this percentage will not have taken the outcome of
31
The present discussion, and all examples are based on Kahnemann 2011, pp. 119-128, which contains
references to more specific research.
- 10 -
the wheel of fortune as starting point for conscious adaptations. It is more likely that the result of the
wheel unconsciously affected their judgements. Nevertheless, even where there is clearly no
connection between the answer that must be given and a number planted in the mind of the
espo de t to a uestio , the pla ted u e still p i es the a s e .32
What is the relevance of this anchoring bias for contract law? If an anchor influences the decision
that somebody makes, it depends on the anchor what the decision will be. There are then two
possibilities. One is that no matter what the anchor, the decision will fall within a range of outcomes
that are equally good for the decision maker. Then the anchor does not influence the quality of the
decision for the decision maker. In case of contracts this means that the anchor does not influence
the (dis)advantages which a contract has for a party. Then the anchoring bias is not relevant for
contract law. Although this is theoretically possible, it will not often happen that an anchor has no
influence on the attractiveness of the resulting contract for a contract partner.
The other possibility seems more likely. Anchoring means that a contract will influence the interests
of the contracting parties. The most frequent situation will be that the party whose decision has been
influenced by the anchor will conclude a less advantageous contract than he would have done
without the anchor. If Anne asks €
, 0 for her house, this will most likely influence the amount
that Bob is illi g to pa fo the house. I ou e a ple the pa ties ag eed o a a ou t of €
,
.
Ho e e , the
ight ha e ag eed o a lesse a ou t, fo i sta e €
,
, if Bob s illi g ess to
pay had not been influenced by the asking price. Then, the anchoring effect costs Bob € ,
. Bob s
ill to pa €
,
did ot efle t his i te est, hi h ould ha e ee se ed ette if Bob had
ought the house fo €
,
.
4. Coping with irrationality
The anchoring bias illustrates the more general phenomenon that human decision making is often far
from rational. People take decisions which they know, or at least could easily have known, not to
reflect their interests optimally.33 The law has traditionally recognized a number of these situations,
and has provided tools to limit the effects of irrationality. However, recent psychological, behavioraleconomic, and neurologically research has shown that the situations in which people act irrationally
are more frequent than was traditionally assumed. How should law deal with the results of this
research?
For the purpose of exposition, I will ignore the possible answer that law should do nothing with these
results. Moreover, I will confine myself to considering only two extreme ways of taking the recent
insights into account. One way is to expand the use of the existing tools. If people are more often
irrational than was traditionally assumed, the law might consider more contracts as null and void,
and allow contract partners more often to avoid contracts they irrationally concluded. Of course,
these possibilities should still take the interests of the other contract partners and of trade in general
into account. The precise outcome of this approach cannot be described here, but it is possible to
characterize this approach. It is a continuation of present contract law, with minor adaptations to
accommodate new scientific insights. Most likely, this is what will actually happen, because law is –
for good reasons – a conservative institution.
32
33
This u o s ious p i i g effe t also pla s a ole i diffe e t setti gs tha those of a ho i g. “ee
Kahnemann 2011, pp. 52-58.
The assumption here is that rationality consists, at least to a large extent, i p o oti g o e s o i te ests.
- 11 -
The question should be raised, however, whether the recent scientific insights reveal that our image
of mankind as a rational decision maker who sometimes lapses, is fundamentally wrong. Perhaps
human beings are not very good at promoting their own interests. Perhaps the satisfaction of
subjective preferences is not at all a good way to create a Pareto-efficient, or otherwise good,
distribution of goods. There are no satisfactory answers to these questions yet, but the recent
scientific developments justify that serious attempts will be made to find such answers. Moreover,
the willingness should in principle exist, if it turns out that subjective preferences are not reliable
indicators of interests, to strongly limit the role of autonomy in contract law. It is even imaginable
that the very phenomenon of contracts, where subjective preferences play a crucial role, should be
replaced by transactions which better reflect the interests of legal subjects. If that would happen, it
would be a very radical change in law indeed. Radical, but also exciting, and worthy of additional
study.
5. Conclusion
The cognitive sciences provide us in high frequency with insights in the functioning of the human
mind. One of these insights – perhaps not even very new – is that human decision making is often
not rational. The law has tools to deal with lapses of rationality, but it works on the basis of an image
of mankind according to which humans are, and typically act, rationally. This raises the question
whether law should be more adaptive to the stream of new insights produced by the cognitive
sciences.
The present article aims to answer this general question for an important field of law: the law of
contract. It shows in section 3, on the basis of the anchoring bias, that there are good reasons to
assume that the conclusion of a contract is often not a rational event. Yet, by making the satisfaction
of subjective preferences a cornerstone of contracts, the law seems to presuppose the opposite. Is it
therefore necessary to change the law considerably, or is it possible to limit the changes or even to
ignore the new insights from the cognitive sciences?
These questions cannot be answered without insight into the essence of a field of law, in our case, of
contract law. Section 2 aimed at providing this insight into the essence of contract law by discussing
three possible justifications of the phenomenon that contracts lead to legally enforceable
obligations. It turned out that two of these justifications, the ones that focused on particular
obligations, are unsatisfactory. The justification of particular obligations on the basis of a general
social practice of contracting fared better. However, the discussion of this kind of justification also
made clear that the social practice of contracting implicitly assumes that subjective preferences are a
stand in for objective interests. This assumption can only be made if subjective preferences generally
are formed in a rational manner. However, this very assumption is the topic of the doubts created by
the cognitive sciences.
The law can follow two strategies34 in dealing with the insight that contracts, and – extrapolating a
bit – juridical acts in general – are often not created rationally. One strategy is to use the tools which
law already has for dealing with irrationality for newly discovered kinds of cases such as the
anchoring bias. Another, is to redesign substantial parts of the law to take into consideration that
subjective preferences often do not reflect objective interests. Which strategy is the best depends on
34
Of course, more strategies are possible, but they will often be intermediates on a scale of which the two
mentioned possibilities are the extremes.
- 12 -
the frequency of situations in which subjective preferences do not reflect objective interests. Is it still
rational to use subjective preferences as indicator for objective interests?
The last question cannot be answered here, but suggests a line of research that has become
important in the light of the developing cognitive sciences. We need to find out which
presuppositions the law makes about the functioning of the human minds. Then we can check
whether these presuppositions still hold. If the presuppositions turn out to be outdated, the law
must be modified on the basis of insight into the essence of its subfields and on the basis of our
recent insights into the functioning of the mind.
References
Ati ah
. P.“. Ati ah, Fulle a d the Theo
Press 1986, pp. 73-92.
of Co t a t , i Essays on Contract, Oxford: Clarendon
Bentham 1789. Jeremy Bentham, An Introduction to the Principles of Morals and Legislation, edited
by J.H. Burns and H.L.A. Hart, London: Methuen 1970, first edition 1789.
Christman 2017. Joh Ch ist a , Auto o i Mo al a d Politi al Philosoph , The Stanford
Encyclopedia of Philosophy (Winter 2017 Edition), Edward N. Zalta (ed.), URL =
<https://plato.stanford.edu/archives/win2017/entries/autonomy-moral/>.
Crisp 2017. ‘oge C isp, Well-Bei g , The Stanford Encyclopedia of Philosophy (Fall 2017 Edition),
Edward N. Zalta (ed.), URL = <https://plato.stanford.edu/archives/fall2017/entries/well-being/>.
Cushman, Young and G ee
. Fie Cush a , Lia e You g a d Joshua D. G ee e, Multi-System
Mo al Ps holog , i Joh M. Do is ed. , The Moral Psychology Handbook, Oxford: Oxford University
Press 2010, pp. 47-71.
Damasio 1994. Antonio Damasio, Descartes’ Error. E otio , reason and the human brain, New York:
Putnam 1994.
Griffin 1986. James Griffin, Well-being. Its Meaning, Measurement and Moral Importance, Oxford:
Clarendon Press 1986.
Hage
. Jaap C. Hage, Co pa i g Alte ati es , i Studies in Legal Logic, Dordrecht: Springer
2005, pp. 101-134.
Hage
. Jaap Hage, ‘e htstheo etis he a al se a de pa tijauto o ie i het o t a te e ht.
T ee di e sies a o ta ts ijheid , i Ilse “a o ed. , Evolutie van de basisbeginselen van het
contractenrecht, Antwerpen: Internsentia 2010, pp 241-261.
Hansson and Grüne-Yanoff 2017. Sven Ove Hansson and Till Grüne-Ya off, P efe e es , The
Stanford Encyclopedia of Philosophy (Winter 2017 Edition), Edward N. Zalta (ed.),
URL = <https://plato.stanford.edu/archives/win2017/entries/preferences/>.
Kahnemann 2012. Daniel Kahnemann, Thinking, fast and slow, London: Penguin 2012.
Kronman 1978. A to T. K o a , Mistake, Dis losu e, I fo atio , a d the La of Co t a ts ,
Journal of Legal Studies VII,1 (1078), pp. 1-34. Also in Posner and Parisi 2002, pp. 351-384.
Larenz/Wolf 2004. Karl Larenz and Manfred Wolf, Algemeiner Teil des bürgerlichen Rechts, 9th ed.,
München: Beck 2004.
- 13 -
Maso
. Eli o Maso , Value Plu alis , The Stanford Encyclopedia of Philosophy (Summer
2015 Edition), Edward N. Zalta (ed.),
URL = <https://plato.stanford.edu/archives/sum2015/entries/value-pluralism/>.
‘az
. Joseph ‘az, Volu ta o ligatio s a d o
Society¸Supplementary Volume 46, pp. 79-101.
ati e po e s , i Aristotelian
Ryle 1949. Gilbert Ryle, The Concept of Mind, London: Hutchinson 1949.
Sapolsky 2017. Robert M. Sapolsky. Behave. The Biology of Humans at Our Best and Worst, London:
Bodley Head 2017.
Sefton-Green 2005. Ruth Sefton-Green, Mistake, Fraud and Duties to Inform in European Contract
Law, Cambridge: Cambridge University Press 2005.
Smits 1995. J.M. Smits, Het vertrouwensbeginsel en de contractuele gebondenheid, Arnhem: Gouda
Quint 1995.
Smits 2014. Jan M. Smits, Contract Law. A comparative introduction, Cheltenham: Edward Elgar 2014.
“ its
. Ja “ its, O e ee ko ste e ht , i Fokke Fe hout ed. , Het Nederlandse recht. Een
Maastrichtse inleiding, Maastricht: Uitgeverij Gianni 2017, pp. 155-179.
Wegner 2002. Daniel M. Wegner, The Illusion of Conscious Will, Cambridge (Mass.): The MIT Press
2002.
- 14 -