Download Economic Behavior and Evolutionary vs. Behavioral Perspectives

Economic Behavior and Evolutionary vs. Behavioral Perspectives
Ulrich Witt
Evolutionary Economics Group
Max Planck Institute of Economics
Kahlaische Str. 10, 07745 Jena, Germany
[email protected]
Abstract
Behavioral economics focuses mainly on how limitations of the human cognitive apparatus,
risk attitudes, and human sociality affect decision making. The former two lead to deviations
from rationality standards, the latter to deviations from rational self-interest. Some of these
research interests are also shared by evolutionary psychology which, however, explains the
observed deviations by features of the human genetic endowment conjectured to have
evolved under fierce selection pressure in early human phylogeny. Important as the decisionmaking theoretical perspective of the two disciplines is, it will be claimed to be too narrow to
fully account for what an evolutionary approach has to say about economic behavior. This is
particularly true regarding the questions of what choices are made and what causes them,
i.e., the motivational aspects of behavior. In the language of economics this points to the
agents= preferences. This article discusses how preferences relate to the human genetic
endowment, how they can change over time, and what conclusions have to be drawn from
extending the analysis to the motivational side.
Keywords
Behavioral economics; Decision heuristics; Development; Evolutionary economics; Growth;
Motivation; Preferences; Welfare
2
Over the past years, the canonical microeconomic utility maximization paradigm has gotten a
serious rival in the form of behavioral economics (see, e.g., Kahneman 2003; Sent 2004;
Santos 2011; see also the contributions in Camerer et al. 2004). Backed by a broad adoption
of an experimental methodology and supported by neuroscience research, behavioral
economics has established a naturalistic approach to economic behavior. As such it can be
seen as an important reorientation of the theory of economic behavior towards a
“consilience” of scientific knowledge (Wilson 1998). The evolutionary perspective suggested
by Wilson as the overarching element is, however, still missing. The question of whether and
how economic behavior is influenced by genetic dispositions is not addressed in behavioral
and canonical economics alike.
For canonical economics the neglect is not surprising, given that its very core--the idea
of utility maximization--was invented during the “marginalist” or “subjectivist” revolution in
economics at the end of the 19th century to mimic Newtonian classical mechanics. It is an
irony of history that this happened exactly at the same time as the Darwinian revolution in the
sciences started to challenge the Newtonian worldview. In Darwin’s theory, conjectures
about the descent of man and the implications for interpreting human behavior are
straightforward. Not so in a theory that “consists in applying the differential calculus to the
familiar notions of … utility, ... [and] demand ... [as] the complete theory of almost every other
science involves the use of that calculus,” as Jevons (1897, pp. 4-5) put it in allusion to
Newtonian physics. Here the explanation of human behavior was--and is to the present day-forced into the Procrustean bed of a “mechanics of utility and self-interest” (p. 23).
Regarding behavioral economics, the lack of an evolutionary perspective is more
surprising, given that important parts of the behavioral repertoire of animals and presumably
also of humans are innate, i.e., develop as an expression of their genes. The lack may be
due to the fact that behavioral economics originated from a critique of the abstract canonical
theory of economic behavior and its excessive claims regarding the rationality of economic
decision makers, particularly in organizations. A first wave of behavioral studies of actual
economic decision making in organizations, conducted by the Carnegie School (March and
Simon 1958; March 1978), was thus inspired by the concept of bounded rationality and
explored its implications.
Motivated in a similar way, the behavioral approach to economics took a more
individualistic turn with the seminal papers by Kahneman and Tversky (1979) and Grether
and Plott (1979). Since then the focus was, and still is, mainly on how individual choices are
affected by the limitations of the human cognitive apparatus (Hogarth 1994), by human risk
attitudes (Starmer 2000), and human social preferences (see, e.g., Henrich et al. 2004). This
focus overlaps to some extent with that of evolutionary psychology. Both approaches are
interested in certain features of human cognition and choice, explaining the pervasiveness of
decision heuristics, biases, and framing effects that fall short of Olympic rationality (see
Gigerenzer and Goldstein 1996). Unlike in behavioral economics, the very rationale of
evolutionary psychology is, of course, to derive these features from constraints of the human
brain as natural selection can be conjectured to have shaped them.
3
A topic that is grossly ignored so far by behavioral economics -- and to some extent
also by evolutionary psychology -- relates the motivational aspects of behavior (with the
notable exception of Kahneman et al. 1997 and Lowenstein 2000). The driving forces behind
the choices that are made are not addressed. Regarding the questions of what motivates
economic behavior, where the motivation come from, and how it changes over time,
behavioral economics is as ignorant as the canonical utility maximization paradigm
(Lowenstein 2004). As will be claimed in this article, it is precisely here that a reconstruction
of the influence of the human genetic endowment with its likely evolutionary foundations on
economic behavior can make a great difference in terms of explanatory power. There is no
need in such a reconstruction to confine the focus to the architecture of the human mind, as
in evolutionary psychology. The analysis can be extended to what it is that the agents
choose rather than only exploring how they choose.
In terms of economic theory, the motivational question refers to the content of
individual preferences. In canonical microeconomics the content is traditionally specified in
an ad hoc fashion (if at all). By accommodating for evolutionary arguments within the utility
maximization paradigm, recent attempts have been made to try to rationalize “why nature
[would] give individuals utility functions” (Robson 2001a; see also Robson 2001b) -- whatever
the variables in the utility functions stand for. In contrast, adopting an evolutionary
perspective is understood here as just the opposite strategy. What needs to be explained in
the first place are the arguments of the utility functions (if one wishes to work with utility
functions) or, more generally, the motivations underlying the observed behavior. The
theoretical expectation is that understanding how the human genetic endowment has
evolved may help to explain innate motivational dispositions and innate learning mechanisms
that modify action motivation over time. In view of their phylogenetic origins, it would not be
surprising if innate motivational dispositions turn out to be far from being functional or optimal
in some sense -- cases of behavioral mismatch being quite frequent.
In line with these considerations the present article proceeds as follows. The next
section discusses in more detail the approach of behavioral economics. Similarities with
evolutionary psychology are highlighted and the evolutionary origins of behavior are
explained (which behavioral economics, unlike evolutionary psychology, neglects). The
following, section 3, turns to the motivational underpinnings of economic behavior and
elaborates on how the evolutionary perspective helps to specify the content of the agents’
preferences and their change over time. In the spirit of the utilitarian program and its affinity
to moral philosophy, the fourth section then turns to a discussion of the possible normative
implications of the suggested evolutionary perspective on economic behavior. The final
section offers conclusions.
Innate Constraints on Behavior and the Agenda of Behavioral Economics
Elementary motivational dispositions and adaptation patterns like operant conditioning and
conditioned reinforcement are important parts of the innate behavioral repertoire of animals
4
(Dugatkin 2003, Ch. 4). As Vaughan and Herrnstein (1987) explain, with their direct or
indirect effect on reproductive success, they are likely to have been shaped by natural
selection in a way that enhances the individual fitness of the organisms carrying the
corresponding “behavior genes.” In sociobiology, this hypothesis is extended to animal
behavior in social interactions (Trivers 1985). Prominent examples are rearing offspring, the
joint hunting of prey, food sharing, support of mating and breeding activities of other animals,
and--most puzzling--self-sacrifices that increase the survival chances of others at the
expense of one’s own. Whether these and other social forms of behavior need to be
explained in terms of the concept of “inclusive fitness” (rather than individual genetic fitness,
see Hamilton 1964) or in terms of group selection theory (see Sober and Wilson 1998) is still
under scrutiny (see Nowak et al. 2010 for a recent assessment).
To a certain extent, human social behavior seems to be shaped by innate influences
too. The question of whether the sociobiological approach can be extended to explaining
human social behavior is, however, controversial (cf. Caplan 1978). Particularly in the
context of early (and of still living, indigenous) human societies, the problems of coordination
of joint activities, mutual support, reciprocity, and “altruism” seem to present themselves
similarly as they do in higher animal societies. Competition for the scarce resources of food,
habitat space, access to mating partners, etc. is a general condition for survival and
reproduction. Yet, even in primitive societies human social behavior is not reducible to
genetically coded forms. There are also culturally conditioned and cognitively created forms
of behavior which may, in some stages of human history, have co-evolved with the genes
(Boyd and Richerson 1985). Culturally acquired knowledge and problem-solving strategies
certainly provide the key for understanding why contemporary human societies are capable
of mastering their environment so successfully as to decrease dramatically the selection
pressure on their social and economic behavior.1
As a consequence, natural selection pressure no longer reshapes and constrains the
variety of culturally conditioned and idiosyncratic human behaviors that lack adaptive value in
terms of reproductive success. The question then is whether there is any regularity in how
this increasing variety continues to evolve, and what determines which behavior is likely to
be observable under what conditions. To answer this it is necessary to distinguish more
thoroughly between various strata of human behavior, from basic, instinctive responses via
innate, non-cognitive learning mechanisms further to cognitively reflected, insightful,
intentional choices (see Lowenstein 2000). It will be claimed that observable behavior at all
these strata is “produced”--to use an economic metaphor--by a hard-wired physiological and
mental apparatus that has emerged from natural selection. Its constraints are still binding for
1
Evidence for a decreasing selection pressure is provided by the fading correlation between
the amount of resources commanded on the one hand and reproductive success on the
other. As Maddison (2001, Ch. 1) shows in a cross-country comparison, the more per capita
income in real terms increased from 1820 to 1998, the more both birth rates and population
growth went down. In pre-industrial societies, in contrast, there was still evidence for a
positive correlation; see Chagnon and Irons (1979) for a historical study.
5
the human behavioral repertoire, but more or less so at the different strata of behavior.
Of the various strata of human behavior, conscious deliberation and decision making
is likely to be the one that, in terms of phylogenetic time scales, evolved most recently.
Evolution has endowed humans with a unique intelligence. Nonetheless, this intelligence can
be argued to show traces of the survival necessities under which it evolved – i.e., genetic
influences -- as evolutionary psychology set out to show (Buss 2003). The mental
“apparatus” that has emerged from natural selection implies specific and more general
constraints. Some of them are well known to economists as limitations in memory,
information processing, 2 and judgment-exactly the issues central to behavioral economics.
The focal point here is on the features of the human cognitive system.
In human perception a limited number of sensory stimuli can spontaneously be
processed in parallel and be recognized.3 The brain commands a number of parallel
processing systems for the various sensory perceptions, the motor system, and cognition. In
each of them attention must be allocated to competing processing demands when, as usual,
stimuli are offered in abundance to the sensory system. For the cognitive capacity this
implies a bottleneck. Spontaneous selective attention processes must filter out a subset of
the information coming in at any given point in time. Only this subset is processed further.
What pieces of incoming information grab attention depends on the one hand on their
frequency and relative strength. Both are defined in terms of adaptation levels dependent on
the previously experienced stimuli Helson (1964). Thus within normal intensity limits it is not
the absolute strength of the stimulus which is decisive for perception, but rather a sufficiently
strong change in the stimulus. 4 On the other hand, attention is allocated according to
whether incoming information contains cognitive cues for which there exists an associative
basis with knowledge already existing in memory so that a meaning can be attributed.
Perceived problems which cannot adequately be associated with information on
problem solutions previously stored in memory trigger cognitive effort in sense-making and
constructing solutions. These spontaneously occurring cognitive activities draw on a rather
limited capacity for immediate information processing (sometimes called the working
memory) that does not allow for the representation of elaborate decision trees or chains of
inference. For this reason, the spontaneous cognitive construction of problem solutions
results in rules of thumb and other decision heuristics. As a consequence, the performance
in problem situations not sufficiently subjected to experience or deliberate training can be
rather poor and fall short of the normative standards of rational decision making.
2
These limitations were center stage in the early bounded rationality debate (Simon 1978).
For the following see, e.g., Kellogg (1995, Ch. 3) or Anderson (2000, Ch. 3, 6, and 7).
4
The adaptive value of such a selective information processing seems to lie in the fact that,
with the brain=s limited information processing capacity, information overflow could paralyze
3
action. Selective attentiveness to the stimulus discrepancies is an effective means of
concentrating attention to the cases of environmental changes, but otherwise keeping the
apperceptive capacity free for other purposes.
6
Because of the narrow information-processing constraints, decision heuristics
frequently partition the decision problems and work off the parts sequentially. Early
experiments related to behavioral decision theory have already shown that this procedure
results in inconsistencies as the complexity of the problem or the time pressure to decide
increases (Slovic et al. 1977). An example is the “elimination by aspect “heuristic (Tversky
1972). It is often observable when decisions have to be taken on alternatives with many
characteristics. Consider the problem of choosing a home in which price, construction quality
and materials, furnishings, site, age, resale value, etc. can all be assumed to play a role.
Decision makers then tend to sort the alternative choices they have according to whether or
not they possess highly valued characteristics. This amounts to making comparisons first
within and then between subsets of alternatives. When, as a consequence, the preferences
for the alternatives are no longer perfectly transitive, the heuristic results in decision
inconsistencies.
The interpretation of what needs to be decided and the evaluation of alternatives is
frequently influenced by “framing effects” (Tversky and Kahneman 1981). This means that
actions and their consequences are assessed differently if presented in different contexts or
in different order.5 Similar constraints of the mental apparatus can be noted in the case of
chance events and judgments under uncertainty. Compared with rationally constructed
concepts of probability and statistical inference, particularly Bayesian probability estimates,
the intuitive assessment of such situations is usually biased if not altogether deficient. In an
evolutionary perspective this should not be very surprising as the rationally constructed
concepts are only a relatively young cultural achievement developed over the last 350 years
(see Hacking 1975).
In their seminal paper, Tversky and Kahneman (1974) listed three examples of
biased or deficient, intuitive judgmental heuristics related to probabilities. The first is
“judgment by representativeness.” Information provided by an observation should be judged
by its objective statistical trustworthiness. Yet, this is not the case if, as is usually
spontaneously done, the information is assessed according to whether it appears
representative for the beliefs an observer already holds. The second example of a bias is
“judgment by availability.” The frequency or probability of an event is judged according to
how well the event can be imagined or how easily one can remember similar or identical
events. As a consequence, risks or chances of less well-known events are systematically
undervalued. The final example is “anchoring and adjustment.” The estimation of relative
frequencies starts from a rough first approximation--the anchor--and is then adjusted by more
detailed considerations typically resulting in a final estimate systematically biased towards
the first approximation.
5
A special case is assessment not by absolute criteria, but relative ones. This is a frequently
observed phenomenon also underlying Kahneman and Tversky’s (1979) prospect theory
where it is paired with a value function that assesses estimated profits and losses
asymmetrically. Further examples of relative assessment in economic context can be found
in Thaler (1980).
7
These judgment biases have been found in experiments with psychologists,
experienced Las Vegas casino patrons, and stock market traders alike, indicating that even
“experts” are not necessarily able to overcome the constraints of a faulty intuition. These
experts frequently also showed “over-confidence,” i.e., unjustified trust in their ability to
estimate objective probabilities (see Slovic et al. 1977). For the conditions under which
decisions had to be made by the early humans, it can be conjectured that the immediacy of
intuitive judgments was more important for survival than the distortions caused by that kind of
judgment. However, unlike in evolutionary psychology (see Cosmides and Tooby 1996), in
behavioral economics the diagnosed regularities and limitations of the human cognitive
apparatus are not put in perspective with, or explained by, the way in which they may have
emerged from natural selection as characteristics of the human genetic endowment.
Obviously, actual decision-making behavior systematically deviates from the
normative standard of rationality cherished as the model of economic behavior by modern
economic theory. These deficiencies can develop disastrous consequences when decision
making has to deal with the risks of the large-scale technologies of modern industrial
economies. However, the way in which decisions are made is only one determinant of
observable economic behavior. Another determinant is the preferences which express the
(not necessarily cognitive) reasons for, or causes of, the choices that are made, be it in a
fully rational mode or not. For the explanation of the ordinary business of life this determinant
may be considered the more consequential one. Nonetheless, with few exceptions (e.g.,
Kahneman et al. 1997; Lowenstein 2000), it has so far been grossly neglected in behavioral
economics. Discussing this determinant leads to the question of what motivates economic
behavior. Dispelled from the agenda by the 19th century subjectivist revolution in economics,
the motivation-theoretic question – addressing the content of the decision makers'
preferences – is still also neglected in behavioral economics.6 But, as will be argued in the
next section, it is precisely by acknowledging the influences of the human genetic
endowment on economic behavior that much progress can be made with this question.
The Neglected Motivational Underpinnings of Economic Behavior
One of the major trends in establishing microeconomic theory in the 20th century has been
the progressive elimination of the motivational aspects of economic behavior. The early
Benthamite sensory utilitarianism had been based on an elaborate hedonistic theory of
action motivation. As is well known, Bentham (1789) had equated utility with the enjoyment
of pleasures and the avoidance of pains and had explained in great detail what pleasures
and pains motivate people to act. A complex theory like this did not fit the “felicitous calculus”
of utility maximization analogously to the dissipation (minimization) of potential energy with
which Jevons (1879) wanted to revolutionize economics. Consequently, efforts in the
6
Exceptions are the special cases of risk preferences and social preferences; see Camerer
and Lowenstein (2004).
8
following decades were directed at finding a proper functional representation of utility and a
corresponding measurement concept. The hedonistic connotations were successively given
up and the idea of a utility function as a representation of a happiness index replaced by the
idea of a utility function as a preference representation function.
While in the behavioral and human sciences research on the motivational
underpinnings of behavior was deepened and increasingly merged with a focus on the
genetic, evolutionary roots of action motivation, economists thus worked in the opposite
direction. “Purging out of objectionable, and sometimes unnecessary, connotations ... of the
Bentham ::: variety” (as Samuelson 1947, p. 90, approvingly stated), they eliminated all
motivational considerations. As a consequence, the ultimately resulting theory of revealed
preferences claims that a preference is revealed when an alternative x is chosen over an
alternative y. It is unable, however, to answer the simple question of what causes this to be
so.7
Recently, attempts have been made in behavioral economics to revive the
motivational underpinnings of sensory utilitarianism (Kahneman, Wakker, and Sarin 1997) by
putting back in place hedonistic motives for taking action: enjoyment of pleasure and
avoidance of pains. A revival of hedonistic, utilitarian hypotheses is, of course, not the only
possible way to reconstruct the motivation-theoretic foundations of a behavioral approach to
economics. Both pleasures and pains are themselves explicable in terms of physiological or
psychological processes that trigger such feelings and induce the agents to act (Rozin 1999).
These physiological and psychological processes belong to the domain of theories of needs
and drives as motivators/inhibitors of action.8 In terms of these theories it can be argued that,
if a need is deprived, the (temporary) reduction or removal of deprivation is classified by an
organism’s sensory system as a pleasurable experience (Cabanac 1979; Pulliam and
Dunford 1980). Conversely, rising deprivation of a need can be expected to cause
increasingly painful feelings.
As discussed elsewhere (Witt 2001), to explain what it is that motivates action it is
necessary to be more specific as to what the relevant needs are. Unlike in earlier attempts to
deal with needs in economics, this specification will be derived here by taking a behavioral
science approach.9 Let a certain action thus reduce or eliminate deprivation in a certain
7
For early criticisms raised against this research strategy see Georgescu-Roegen (1954a)
and Sen (1973).
8
Need-theoretic reasoning has a tradition reaching back to Aristotle. Several economists
have used it in the past for explaining the motivation underlying economic behavior, among
them Duesenberry (1949), Georgescu-Roegen (1954b), and Ironmonger (1972).
9
At any point in time a need is more or less deprived if it is not completely satisfied or
satiated. Pressing as deprived basic needs may be, they are not assumed to give rise to
anything like lexicographic preferences, i.e., a hierarchical order in which they are satisfied.
The present approach therefore also contrasts with Maslow’s (1987) influential hypothesis of
a hierarchy of needs, a hypothesis that could not be empirically confirmed; see Wahba and
9
need. If this event increases the rate with which that action is chosen in the future, the
satisfaction of the need is connected with (unconditioned) reinforcement. The notion of needs
will be restricted here to precisely those for which this connection holds. Need satisfaction
can then be identified with primary reinforcers in the theory of instrumental or operant
conditioning (Herrnstein 1990; Staddon and Cerutti 2003). Obviously, only a limited number
of physiological and psychological needs qualify for this category (dubbed “basic needs”)
here. Among them are the needs for air, water, sleep, food, body heat, shelter, pain relief,
physical activity, sex, affection, social recognition and status, sensory arousal, cognitive
consistency, and achievement (Millenson 1967, p. 386). Given their obvious reproductive
value in times of fierce selection pressure, these basic needs can be argued to be innate
and, indeed, they are commonly shared by humans (with the usual genetic variance).
The behavioral reduction of the need-theoretic explanation for why an agent should
be motivated to choose an alternative (or order several alternatives in a certain sequence)
has certain advantages. Within the framework of (sensory) utilitarianism, the explanation of
the motivation to act is cast in terms of a balance of pleasures and pains associated with the
alternative(s). Simplifying somewhat, pleasures and pains are assumed to be homogenized
into a single hedonic currency--the utility index--taken to express the relative strength of the
action motivation (see Shizgal 1999). Thus, a higher value of the hedonic currency suffices
as a proximate cause for the motivation underlying an observed choice of an action. What
determines the pleasure and pain feelings from which utility is derived does not have to be,
and usually is not, specified.
In contrast, the suggested need-theoretic explanation provides an ultimate cause for
the motivation underlying the choice of an action. It does so by identifying how the action
reduces deprivation with respect to some specific need(s)--related above to primary
reinforcers--and, thus, triggers a pleasurable experience (generates utility). Moreover,
besides giving deeper reasons for how utility is generated, the present interpretation also
suggests important dynamic extensions of the motivational underpinnings of behavioral
economics. These dynamic extensions are an implication of reinforcement theory claiming
that (non-genetic) behavior adaptation is governed by two different kinds of innate dynamics:
that of instrumental conditioning on the one side and that of conditioned reinforcement or
conditioning learning on the other (see Leslie 1996).
The adaptation dynamics implied by operant or instrumental conditioning
(“reinforcement learning”) basically converge to the “matching law“(Herrnstein 1997). This is
an empirical generalization derived from a large number of experiments testing behavior that
is not, or is only marginally, cognitively reflected and controlled. If there are several actions
feasible that serve one and the same need, an organism learns by instrumental conditioning
how to adjust the relative frequency of actions chosen to the relative size of the rewards
obtained by the actions.10 The matching law then postulates that the relative frequency with
Bridwell (2002).
10
In the need-theoretic interpretation, actions and rewards are related by their capacity to
reduce deprivation of a basic need. For example, eating, or more specifically, calorie intake,
10
which the alternative actions are chosen roughly equals the relative size of the average
reward they yield (unless one action dominates all others in terms of average reward
whatever the frequency of choices is).
It is important to note, however, that the consequences of reinforcement typically
differ between laboratory experiments and real-life situations. First, while in an experiment
both the rate of reinforcement and the level of deprivation are experimental control
parameters, in economic reality they represent trend variables that are correlated with the
growth trend in disposable per capita income. Second, while experiments are usually
conducted with one reinforcer only--usually some form of food--behavior outside the
laboratory is subject to all, more or less, deprived needs at the same time. This is a more
complicated case. As a consequence, at any point in time the relative strength of the
motivation to act to reduce deprivation of any particular need depends on its degree of
deprivation relative to the degree of deprivation of all other needs.
To put it differently, on the one hand the agents adjust over their conditioning history
to the reward structure of each single need by reinforcement learning. On the other hand,
they learn to adjust to the relative ease with which reinforcement can be obtained in their
environment across their needs. Relatively strongly positively reinforced needs (i.e., needs
that can be satisfied relatively easier with feasible actions) are more frequently pursued, less
strongly or even negatively reinforced needs less so. People thus develop an individual
approach to where they seek, more or less intensely, the rewarding experience of need
satisfaction and how--some become gluttons, some party lions, some sex obsessed, some
workaholics, and so on, and many a little bit of everything. In the utilitarian language the
emergence of such individual specialization patterns is part of the formation of individual
preferences (a process that may be strongly supported by cognitively controlled self-efficacy;
see below).
As always, individual (but usually culturally contingent) specialization induces a
significant inter-individual variance in observable behavior. Nonetheless, it can be claimed
that, in the mean, behavior displays a general tendency when the overall capacity of
satisfying needs increases with a steadily rising income. This tendency is caused by the
presence or absence of homoeostatic features in need satisfaction. As a consequence, basic
needs differ with respect to their deprivation-satiation patterns in a way that is similar across
all humans. For some basic needs, deprivation can, in principle, be reduced temporarily to
zero. Examples of needs that can be satiated quite easily are the needs for food and
something to drink. But there are also basic needs whose average deprivation is for different
reasons difficult, if not impossible, to reduce to zero. Typically, these are needs whose
satiation level is defined in relative terms. Examples are the need for arousal, i.e., sensory or
cognitive stimulation, and the need for social recognition. The consequences of their factual
insatiability have been the subject of works by Scitovsky (1981) on arousal and Hirsch (1978)
triggers the rewarding feeling of reducing hunger. Adaptations under reinforcement learning
are also influenced by the immediacy and contingency of the reward, two variables not to be
discussed here; see Leslie (1996) for details.
11
and Frank (1999) on social recognition.
The adaptation dynamics implied by conditioned reinforcement or conditioning
learning are quite different. They result from the fact that an organism tends to learn to
associate stimuli that trigger an action leading to reward (a pleasurable experience) and
neutral stimuli (triggering neither a pleasurable nor an aversive experience), if these two
kinds of stimuli coincide repeatedly. Once such an association is established, the originally
neutral action triggers a rewarding experience via the learned association. A conditioned
(secondary or learned) reinforcer is established.11 This effect works even if the previously
coinciding primary reinforcement is dropped, but the strength of a conditioned reinforcer
fades away if the association is not at least occasionally corroborated. One can speak here
of the emergence of “learned wants” whose satisfaction triggers pleasurable feelings. In the
utilitarian language the acquisition of such learned wants is a different part of the formation
and change of the preferences specific to an individual agent.
Unlike the widely interpersonally shared basic needs, the emerging structure of
learned wants is of a highly idiosyncratic nature. It would make little sense, therefore, to
produce a list of learned reinforcers comparable to that of the limited number of innate ones.
A good part of the observable inter-individual variance in human preferences can be
explained by the structure of learned wants. (An important other part is due to the individual
specialization patterns resulting from adaptations under instrumental conditioning in terms of
where to seek more or less intensely for need satisfaction.) However, the particular cultural
environment in which conditioning takes place and influences the associations that the
agents happen to learn can induce a certain cultural bias in the individually learned wants.
For this reason, agents in similarly socialized groups or in similar cultural environments may
show less variety in their learned wants than agents from different cultural backgrounds.
The motivational forces and their changes over time do, of course, not only play a
role for behavior that is not cognitively reflected or controlled. In the economic domain,
conscious cognitive deliberation often intervenes more or less intensely into the motivational
mechanisms of instrumental conditioning and conditioning learning. Hence, a theory of
motivation would be incomplete without hypotheses that account for cognitive influences. By
cognitive construction of means-ends relationships, actions are assessed with respect to
their instrumental value for attaining need satisfaction--with all the decision-making biases
discussed in the previous section. With regard to the present motivation-theoretic context,
11
An example is the following (see Witt 2001). Imagine repeatedly taking a good meal when
hungry in a special environment characterized by scenic architecture, furniture, tableware,
table music, etc. Assume that such a special environment is initially a neutral experience. If
so, the association that is learned between good eating and the special features of the
environment then has the effect that experiencing scenic architecture, furniture, tableware,
table music, etc. may become a rewarding experience in its own right--a conditioned
reinforcer. As a motivating force of its own it can induce actions aiming at the experience of
the mentioned features, even if no longer coinciding with eating activities.
12
the consequence of cognitive intervention is that it can selectively change observable
behavior compared to what reinforcement contingencies would predict. The perception of
instrumental relationships can consciously be manipulated and the reinforcement (the actual
satisfaction of some need) thus can be postponed as an instance of self-efficacy (Bandura
1986).12
An important feature of cognitive interventions in making choices is their highly
selective nature. This is due to the already discussed limitations of the human information
processing capacity and the selectivity of attention processes. This fact has implications also
at the motivational level because it results in a dynamic interaction of cognitive and
motivational processes. As was mentioned in the previous section, scarce attention capacity
is chiefly attracted to incoming information that contains cues for which there already exist an
associative basis in long-term memory so that it is easy to attribute meaning to that
information (Kellogg 1995, Ch. 3; Anderson 2000, Ch. 3, 4). This means that at any point in
time the incremental change of individual action knowledge through newly processed
information ultimately hinges on already existing knowledge. However, the cognitive cues
and the associated memory content differ in how much attention they are able to attract,
depending on the affective value of the particular meaning that is associated. The affective
value, in turn, reflects the strength of previous rewarding (pleasurable) or aversive (painful)
experiences that are memorized in association with the information.
If individual preferences also reflect the affective value of action information, it follows
that the new knowledge we acquire is not only contingent on our already existing knowledge;
it also hinges on the current state of our preferences that influence the selective allocation of
attention. This means that the interactions between the current state of our knowledge and
the current state of our preferences feed back on the further shaping of perceptions and
preferences. Extending the attention given to an action possibility at the expense of others
allows us to recognize details in that possibility which otherwise are likely to have gone
unnoticed (refinement effect in perceptions and preferences). Put differently, with more
refined perceptions of action possibilities further specialization in reinforcement sampling
under instrumental conditioning becomes feasible.
Normative Connotations of the Evolutionary Perspective
Both behavioral economics and evolutionary psychology have been shown to reflect mainly
on the characteristics and limitations of human perceptions and decision making. It has been
argued that, for a comprehensive evolutionary perspective on economic behavior, these
reflections need to be extended by a thorough reconstruction of the motivational
underpinnings of human behavior. In a sense, such a reconstruction indeed leads “back to
12
Cognitive deliberation can be guided by its own motivational forces, i.e., needs such as
the one for consistency of one’s self-image (Dunning 2007) or the need for high self-esteem
(Gollwitzer and Kirchhof 1998).
13
Bentham” (Kahneman et al. 1997). But the motivational side can, of course, be discussed
from many perspectives – besides from the utilitarian (hedonistic) point of view, also, e.g.,
from a need-theoretic or from a behavioral science point of view. (All three approaches have
been shown to correspond with one another in several respects.) By shifting the focus to the
motivational side of human behavior one may feel obliged to get back, too, to the moral
concerns that have always been associated with human motivations, not only, but surely
center stage, in the utilitarian tradition (Bentham [1789] 1948; Mill [1863] 1998). A short look
at possible normative implications of the approach suggested here may therefore be
warranted.13 It gives an opportunity, in particular, to briefly address a problem that presently
appears as a white spot on the map of evolutionary economics: the problem of the evolution
of individual well-being and societal welfare.
The straightforward connection is made by the fact that what is left today of the
utilitarians’ original interest in the moral legitimacy of behavior and its underlying motivations
appears in the very abstract disguise of welfare theory. Under the preference subjectivism
that is characteristic of welfare economics one does not need to know much about the
motivation underlying a chosen action. All that has to be ensured is that all parties involved
have made voluntary choices under freedom of contract and in the absence of externalities
that cannot be resolved through private negotiations. Under such conditions, it is usually
inferred that actions and transactions reflect what is feasible and satisfies the preferences of
the involved agents and, in this sense, is “good.” Yet, this argument is based on the
assumption of consistent and unchanging preferences. This is a strong and, in the light of the
motivation-theoretic considerations in the previous section, counter-factual assumption.14
To overcome it the black box of subjective preferences must be opened. It is usually
admitted then that individual preferences are partly inconsistent (as has been recognized by
behavioral economics; see, e.g., Sugden 2004). In the light of the considerations in the
previous section one can, however, go further. If preferences are dependent on innate
needs, culturally learned wants, and cognitively constructed motives, several questions can
be raised regarding welfare assessments. At what stage of preference learning should the
13
There is a considerable amount of works in biology on evolutionary ethics (see, e.g., Mohr
1984 and Wilson 2002) which take an explanatory or positive approach to human moral
values, i.e., inquires into how human morality may have evolved. The subsequent reflections
take a different--normative--approach. They dwell on whether and, if so, how evolved human
motivations are compatible with particular, hypothetically assumed, normative judgment
however one may have arrived at these judgments.
14
For a similar criticism see Gowdy and Mayumi (2001). The assumption is sometimes
replaced by assuming that individuals have different utility functions or different states of
preferences at different points in time, and that they can evaluate their different preference
orderings at different points in time on some common basis. This may, for instance, be a
meta-preference, i.e., an order over the preferences at different points in time; see, e.g., Sen
(1977) and Elster (1982).
14
state of preferences be taken as a measuring rod to assess whether there are any welfare
gains? Given that, with rising wealth, some of the innate needs can be satiated more easily
than others, do such differences in satiability matter for welfare assessments? With rising
wealth, more opportunities for developing further refinements in learned wants and
cognitively constructed motives arise. When they motivate further increases in spending,
does the self-perpetuating motivational spiral that seems to emerge here render welfare
altogether inapt as a measure of human progress?
No doubt, the last century’s dramatic growth of income and wealth in the developed
and many of the developing countries had an impact. Life improved tremendously compared
to what were the harsh living conditions over most of human history. The multiplication of per
capita disposable income has enabled not only the upper strata of society but also the
masses to conduct a “better life.” In the light of the motivational hypotheses suggested in this
paper it seems, however, that in welfare-theoretic terms a judgment like this cannot be made
independent of the level of income already reached. Once disposable income allows the
removal of deprivation in the pressing human needs, other motivational mechanisms take
over in guiding economic behavior. What is then enjoyed as pleasures--to use the utilitarian
diction--are often learned pleasures. And where the pleasures would, in principle, seem
satiable with the income level reached, cognitive motives may be learned to enjoy ways of
further income spending that avoid satiation.
Thus, at some point in development, what appears as "better life" by historical
standards becomes a standard of living that has been learned. And as with all learning there
is a peculiar asymmetry. Had there been no continued income increases, no opportunities to
experience all the new consumption possibilities would have occurred. People would not
know what ways of obtaining pleasant sensory perceptions they forego, because they would
not have learned to appreciate them. Once all the experiences have been made, though,
foregoing all the learned pleasures, e.g., because of a sustained decline in disposable
income, would be felt as harsh privation.
This is an abstract argument. Nonetheless, one may be concerned about its
implications. They seem to entail a rather relativistic assessment of the “good” in the notion
of a “good life.” Moreover, the very basis on which further improvements in the “goodness”@
(associated with further income increases) are diagnosed by welfare theory may start to
appear rather odd. One may ask, for instance, whether additional income spent on needs
that are basically satiated appears as legitimate as additional expenditure on highly deprived
needs. Is additional income spending to obtain rewarding experiences from learned wants
(conditioned reinforcers) as legitimate as additional spending to obtain reinforcement from
reducing deprivation in innate needs? How should refinements of the cognitive and the
sensory perceptions be assessed that are learned once income rises, given that they are
little valued as long as they are not yet learned, but their privation is felt as a loss after
learning? Obviously, very controversial value judgments are possible here, depending on
what aspects of civilization one has learned to appreciate and what aspects one has learned
to dislike.
However, the question becomes a morally pressing one if there are significant
externalities arising from high-income spending habits. Such externalities are indeed present
today in several forms: the degradation of the environment (e.g., the greenhouse effect) on
the one hand and the extreme interpersonal differences in disposable income on the other.
To put it in a provocative way, how is the interest of affluent consumers in seeking the kick of
the latest frenzies (which their high income has allowed them to learn to appreciate) or in
keeping ahead in status seeking by resource-intensive consumption to be morally assessed
against the interest of other people in saving them from starving hunger or living in a
degraded environment?
It is quite likely that our moral feeling tells us that there is a problem here. This may
be so because of the feelings of empathy and justice that seem to be a part of our inherited
sociality (Binmore 2006; Tomasello 2009). It is not surprising then that income redistribution
is frequently considered morally legitimate, or that the idea of enabling low-income countries
to catch up by development policies is propagated. However, precisely because there is no
absolute standard for defining what a “good life” means as long as there are further income
increases, the legitimacy of both income redistribution and development policies may again
depend on what income level has already been reached.
If the insights that can be gained by understanding the evolutionary foundations of
economic behavior are taken seriously, at some point in economic development welfare
judgments start to become judgments on a drift process. If it is true that large parts of income
spending are culturally learned forms of behavior providing forms of pleasure that one would
not have missed had they not been learned, moral feelings may suggest that an
unboundedly growing income is no necessary condition for a “good life” and the implicit
equation of income increases with “better life“ problematic. When putting oneself in the shoes
of people starving from hunger and of people living in a seriously degraded environment one
may feel inclined to reflect about the ethical argument that allows us to contribute to a
balance in how we spend our income and how much further income increases are indeed
valued. As man is a social animal, after all, the social discourse about what the right balance
is would perhaps help to stimulate a broad cognitive awareness of the moral connotations
the highly developed countries face with the way they spend their income.
Conclusions
An evolutionary perspective on human behavior has to account for the influences that the
agents’ genetic endowment has on their choices. Fixed in times of fierce selection pressure,
this endowment is presumably adapted to living conditions as the early humans faced them.
On this premise, evolutionary psychology explains various decision heuristics characteristic
of the human cognitive system violating modern rationality standards. Behavioral economics
shares the interest in various “decision anomalies,” i.e., in when and how economic decision
making deviates systematically from what appears to be rational choices. However, no
connection is made to any genetic influences.
16
In this article it has been argued that an evolutionary perspective on human behavior
needs to go beyond the question that both evolutionary psychology and behavioral
economics are concerned with, i.e., the question of how decisions are made. To fully account
for the influences that the agents’ genetic endowment has on their choices, focus needs to
be extended to the question of what decisions are made for what reasons or causes, i.e., to
the motivational underpinnings of economic behavior. Regarding the evolution of the human
economy as a whole, this may indeed be the more momentous question. In the language of
economics this means to inquire into the agents’ preferences and to explain how they relate
to the human genetic endowment and how they change over time.
As has been shown, the motivational side of human behavior can be discussed from
a utilitarian, a need-theoretic, and a reinforcement-theoretic point of view. The three
approaches turned out to lead to similar conclusions in several respects. It has been
suggested therefore to identify innate needs and learned wants with primary and secondary
reinforcers respectively. One can then be more specific as to what it is that motivates
behavior, or what generates utility, and in which way. On this basis one can go beyond the
current debate in behavioral economics focusing mainly on cognitive aspects of decision
making. Several implications, not least with respect to an evolutionary view on welfare
economic problems, have been highlighted.
Acknowledgments
I am grateful to Christophe Heintz for helpful comments on an earlier draft and to Kirby
Kempe for improving the readability of this paper.
References
Anderson JR (2000) Cognitive psychology and its implications, 5th edn. Worth, New York
Bandura A (1986) Social foundations of thought and action: a social cognitive theory.
Prentice-Hall, Englewood Cliffs
Bentham J ([1789] 1948) An introduction to the principles of morals and legislation. Hafner,
New York
Binmore K (2006) The origins of fair play. Papers on Economics and Evolution #0614. Max
Planck Institute of Economics, Jena
Boyd R, Richerson PJ (1985) Culture and the evolutionary process. University of Chicago
Press, Chicago
Buss DM (2003) Evolutionary psychology: the new science of the mind. Allyn and Bacon,
Boston
Cabanac M (1979) Sensory pleasure. Quarterly Review of Biology 54(1):1-29
Camerer CF, Loewenstein G (2004) Behavioral economics: past, presence, future. In:
Camerer CF, Loewenstein G, Rabin M (eds) Advances in behavioral economics.
17
Princeton University Press, Princeton, pp 3-51.
Camerer CF, Loewenstein G, Rabin M, (eds) (2004) Advances in behavioral economics.
Princeton University Press, Princeton
Caplan AL (ed) (1978) The sociobiology debate. Harper, New York
Chagnon NA, Irons W (eds) (1979) Evolutionary biology and human social behavior: an
anthropological perspective. Duxbury Press, North Scituate
Cosmides L, Tooby J (1996) Are humans good intuitive statisticians after all? Rethinking
some cConclusions from the literature on judgment and uncertainty. Cognition 58:173
Duesenberry JS (1949) Income, saving, and the theory of consumer behavior. Harvard
University Press, Cambridge, MA
Dugatkin LA (2003) Principles of animal behavior. Norton, New York
Dunning D (2007) Self-image motives and consumer behavior: how sacrosanct self-beliefs
sway preferences in the marketplace. J Consumer Psychology 17(4):237-249
Elster J (1982) Sour grapes: utilitarianism and the genesis of wants. In: Sen AK, Williams B
(eds) Utilitarianism and beyond. Cambridge University Press, Cambridge, pp 219-238
Frank RH (1999) Luxury fever: why money fails to satisfy in an era of excess. Free Press,
New York
Georgescu-Roegen N (1954a) Choice and revealed preference. Southern Economic Journal
21:119-130
Georgescu-Roegen N (1954b) Choice, expectations, and measurability. Quarterly J
Economics 68:503-534
Gigerenzer G, Goldstein DG (1996) Reasoning the fast and frugal way: models of bounded
rationality. Psychological Review 103:650-669
Gollwitzer PM, Kirchhof O (1998) The willful pursuit of identity. In: Heckhausen J, Dweck CS
(eds) Motivation and self-regulation across the life span. Cambridge University Press,
Cambridge, pp 389-423
Gowdy J, Mayumi K (2001) Reformulating the foundations of consumer choice theory and
environmental valuation. Ecological Economics 39:223-237
Grether DM, Plott CR (1979) Economic theory of choice and the preference reversal
phenomenon. American Economic Review 69:623-638
Hacking I (1975) The emergence of probability: a philosophical study of early ideas about
probability, induction and statistical inference. Cambridge University Press,
Cambridge
Hamilton WD (1964) The genetical evolution of social behavior I. J Theoretical Biology 7:1-16
Helson H (1964) Adaptation level theory. New York: Harper & Row
Henrich J, Boyd R, Bowles S, Camerer C, Fehr E, Gintis H (eds) Foundations of human
sociality. Oxford University Press, Oxford
Herrnstein RJ (1990) Behavior, reinforcement and utility. Psychological Science 1:217-224
Herrnstein RJ (1997) The matching law. Harvard University Press, Cambridge, MA
Hirsch F (1978) Social limits to growth. Harvard University Press, Cambridge, MA
Hogarth RM (1994) Judgment and choice: the psychology of decision. Wiley, Chichester
18
Ironmonger DS (1972) New commodities and consumer behavior. Cambridge University
Press, Cambridge
Jevons WS (1897) The theory of political economy, 2nd edn. MacMillan, London
Kahneman D (2003) Maps of bounded rationality: psychology for behavioral economics.
American Economic Review 93:1449-1475
Kahneman D, Tversky A (1979) Prospect theory: an analysis of decision under risk.
Econometrica 47:263-291
Kahneman D, Wakker PP, Sarin R (1997) Back to Bentham? Explorations of experienced
utility. Quarterly J Economics 112:375-405
Kellogg RT (1995) Cognitive psychology. Sage Publications, Thousand Oaks
Leslie JC (1996) Principles of behavioral analysis. Harwood Academic, Amsterdam
Lowenstein G (2000) Emotions in economic theory and economic behavior. American
Economic Review 90(2):426-432
Lowenstein G (2004) Out of Control: visceral influences on behavior. In: Camerer CF,
Lowenstein G, Rabin M (eds) Advances in behavioral economics. Princeton
University Press, Princeton, pp 689-723
Maddison A (2001) The world economy: a millenium perspective. OECD, Paris
March JG (1978) Bounded rationality, ambiguity, and the engineering of choice. Bell Journal
of Economics 9:587-608
March JG, Simon HA (1958) Organizations. Wiley, New York
Maslow A (1987) Motivation and personality, 3rd edn. Harper Collins, New York
Mill JS ([1863] 1998) Utilitarianism. Oxford University Press, Oxford
Millenson JR (1967) Principles of behavioral analysis. Macmillan, New York
Mohr H (1984) The ethics of science: compatible with the concept of evolutionary
epistemology? In: Wuketits FM (ed.) Concepts and approaches in evolutionary
epistemology. Reidel, Boston, pp 185-208
Nowak MA, Tarnita CE, Wilson EO (2010) The evolution of eusociality. Nature 466:10571062
Pulliam HR, Dunford C (1980) Programmed to learn: an essay on the evolution of culture.
Columbia University Press, New York
Robson AJ. (2001a) Why would nature give individuals utility functions? J Political Economy
109(4):900-914
Robson AJ (2001b) The biological basis of economic behavior. J Economic Literature 39:1133
Rozin P (1999) Preadaptation and the puzzles and properties of pleasure. In: Kahneman D,
Diener E, Schwarz N (eds.) Well-being: the foundations of hedonic psychology.
Russell Sage Foundation, New York, pp 109-133
Samuelson PA (1947) Foundations of economic analysis. Harvard University Press,
Cambridge, MA
Santos AC (2011) Behavioural and experimental economics: are they really transforming
economics? Cambridge Journal of Economics 35:705-728
Scitovsky T (1981) The desire for excitement. Kyklos 34:3-13
19
Sen AK (1973) Behavior and the concept of preference. Economica 40:241-259
Sen AK (1977) Rational fools: a critique of the behavioral foundations of economic theory.
Philosophy and Public Affairs 6:317-344
Sent E-M (2004) Behavioral economics: how psychology made its (limited) way back into
economics. History of Political Economy 36(4):735-760
Simon HA (1978) Rationality as process and as product of thought. American Economic
Review 68, Papers and Proceedings:1-16
Shizgal P (1999) On the neural computation of utility: implications from studies of brain
stimulation reward. In: Kahneman D, Diener E, Schwarz N (eds) Well-being: the
foundations of hedonic psychology. Russell Sage Foundation, New York, pp 500-524
Slovic P, Fischhoff B, Lichtenstein S (1977) Behavioral decision theory. Annual Review of
Psychology 28:1-39
Sober E, Wilson DS (1998) Unto others: the evolution and psychology of unselfish behavior.
Harvard University Press, Cambridge, MA
Staddon JER, Cerutti DT (2003) Operant conditioning. Annual Review of Psychology 54:
115-144
Starmer C (2000) Developments in non-expected utility theory: hunting for a descriptive
theory of choice under risk. J Economic Literature 38:332-382
Sugden R (2004) The opportunity criterion: consumer sovereignty without the assumption of
coherent preferences. American Economic Review 94:1014-1033
Thaler R (1980) Toward a positive theory of consumer choice. J Economic Behavior and
Organization 1:39-60
Tomasello M (2009) Why we cooperate. MIT Press, Cambridge, MA
Trivers R (1985) Social evolution. Benjamin/Cummings, Menlo Park
Tversky A (1972) Elimination by aspect: a theory of choice. Psychological Review 79:281299
Tversky A, Kahneman D (1974) Judgement under uncertainty: heuristics and biases.
Science 185:1124-1131
Tversky A, Kahneman D (1981) The framing of decisions and the psychology of choice.
Science 211:453-458
Vaughan W, Herrnstein RJ (1987) Stability, melioration, and natural selection. In: Green L,
Kagel JH (eds) Advances in behavioral economics, Vol. 1. Ablex, Norwood, pp 185215
Wahba M, Bridwell L (2002) Maslow reconsidered: a review of research on the need
hierarchy theory. In: Cooper C (ed) Fundamentals of organizational behavior, Vol. 1.
Sage, London, pp 42-67
Wilson DS (2002) Evolution, morality and human potential. In: Scher SJ, Rauscher F (eds)
Evolutionary psychology: alternative approaches. Kluwer, Boston, pp 55-70
Wilson EO (1998) Consilience: the unity of knowledge. Alfred Knopf, New York
Witt U (2001) Learning to consume: a theory of wants and the growth of demand. J
Evolutionary Economics 11: 23-36