Cambridge Books Online
http://ebooks.cambridge.org/
Elements of Reason
Cognition, Choice, and the Bounds of Rationality
Edited by Arthur Lupia, Mathew D. McCubbins, Samuel L. Popkin
Book DOI: http://dx.doi.org/10.1017/CBO9780511805813
Online ISBN: 9780511805813
Hardback ISBN: 9780521653299
Paperback ISBN: 9780521653329
Chapter
Beyond Rationality: Reason and the Study of Politics pp. 1-20
Chapter DOI: http://dx.doi.org/10.1017/CBO9780511805813.001
Cambridge University Press
Beyond Rationality: Reason and the
Study of Politics
ARTHUR LUPIA, MATHEW D. MCCUBBINS, AND SAMUEL L. POPKIN
A primary objective of social science is to explain why people do what
they do. One of the great difficulties inherent in crafting such explanations
is that we cannot observe the thoughts that precede a choice. As a
result, social scientific explanations of individual behavior must be based
on assumptions about the relationship between thinking and choosing assumptions
whose validity is not obvious.
In this volume, 18 scholars from a broad range of social scientific perspectives
join together in an effort to craft better explanations of political
behavior. Individually, each contributor argues that better
explanations will come from paying closer attention to the relationship
between thinking and choosing.
Collectively, our goal is to transform debates about the limits of
rationality into more effective explanations of why people do what
they do. Our attempt at such a transformation begins in this chapter,
where we develop an alternative approach to the study of politics.
Our approach combines a single, empirically sensible definition of
rationality with an aggressive pursuit of how people seek and process
information.
Our definition of rationality is motivated by the belief that scarcity
is ubiquitous in political contexts. There are, for example, more ways
to spend public funds than there are funds to spend, more ideas
about what a society should do than there are opportunities for society
to act, and so on. Time and energy are also scarce. As a result, people
lack the time and energy to pursue all possible opportunities. Because
scarcity is ubiquitous in political contexts, political actors must make
choices.
To explain why people make certain choices, it is necessary to understand
that choice is the product of reason, where reason is the human
process of seeking, processing, and drawing inferences from information.
People reason about how the consequences of various actions relate to
Downloaded from Cambridge Books Online by IP 147.251.110.115 on Tue Mar 08 09:25:31 GMT 2016.
http://dx.doi.org/10.1017/CBO9780511805813.001
Cambridge Books Online © Cambridge University Press, 2016
Lupia, McCubbins, and Popkin
the satisfaction of their desires. This relationship between reason and
choice is as true for the selfish or wealth-maximizing actors common to
economic models as it is for the altruist who chooses to allocate scarce
resources for the benefit of others. It is as true for people who think of
themselves primarily in terms of cultural or group identities as it is for
rugged individualists. Scarcity forces all people to make choices, and
these choices are the product of reason.
Since reason is the antecedent of choice, limits in our understanding
of reason shackle our ability to explain what people do. We argue that
these shackles can be lifted. Indeed, the point of the volume is to identify
elements of reason, the systematic components of the process by
which humans seek and use information. Our main premise is that, whatever
their motivations, people seek and process information about politics
in predictable ways. The purpose of Elements of Reason is to clarify
what these predictions might be.
The premise that people seek and process information about
politics in predictable ways permeates the entire book. Each of our
contributors defines an element of reason that should be incorporated
into the study of politics. Collectively, these elements are supported
by empirical discoveries from a broad range of perspectives. Cognitive
scientists, economists, political scientists, and psychologists all contribute
elements of reason to this volume. While these essays - described
in greater detail at the end of this chapter - are not a comprehensive
overview of the links between the social sciences and the study
of the mind, each clarifies an important aspect of the relationship
between thinking and choosing. They show that reason determines how
the environment, the basic needs of the human organism, and past experiences
combine to determine people's desires, and ultimately, their
choices.
Our goal is to create a new understanding of how people reason. By
applying these insights to the study of politics, we can better explain why
people do what they do.
RATIONALITY AND THE IMPORTANCE OF LISTENING
TO REASON
Many scholarly approaches to the study of politics rely on controversial
or unspecified assumptions about the relationship between thinking
and choosing. Political systems theories, group theories, elite
theories, and prospect theories, for example, rely on vague or narrowly
tailored assumptions about how people think about what they
choose.
One well-known and particularly controversial approach to the study
2
Downloaded from Cambridge Books Online by IP 147.251.110.115 on Tue Mar 08 09:25:31 GMT 2016.
http://dx.doi.org/10.1017/CBO9780511805813.001
Cambridge Books Online © Cambridge University Press, 2016
Beyond Rationality
of politics operates on the premise that people are rational.1
Unlike some
other approaches to the study of politics, rational choice theories are
seldom vague in their assumptions about how people think. One of the
reasons that rational choice theories are so controversial, however, is that
some of the best-known applications are perceived to rely on unrealistic
assumptions. Complicating matters further is the fact that the term rationality
means different things to different people. As a result, when people
argue about the role of rationality in the study of politics, they are often
arguing about very different concepts.
In this section, we use the controversy about rational choice theories
to demonstrate that more careful attention to reason can help
a broad range of scholars - regardless of whether they glorify or vilify
rational choice theories - to craft better explanations of why people do
what they do.
Rationality's Dueling Definitions
Here the problem is first to see just what is meant by rationality. The term,
as a recent writer noted, "has enjoyed a long history which has bequeathed
to it a legacy of ambiguity and confusion. .. . Any man may be excused
when he is puzzled by the question how he ought to use the word and particularly
how he ought to use it in relation to human conduct and politics."
Several meanings can be differentiated.
- Berelson, Lasarsfeld, and McPhee
1954: 310
Rationality is a concept so central to human endeavor, so important, and
yet so empirically elusive that it cannot help but be subject to a broad
range of interpretations when used in everyday discourse. In scientific
discourse, however, it is desirable for words to have precise meanings.
The validity and soundness of scientific claims about causality, correlation,
and truth depend critically on such precision. So, while it is not surprising
that words such as rationality have multiple meanings in everyday
discourse, when it comes to achieving scientific objectives, multiple
meanings cause problems.
1
The role of reason in rational choice theories is particularly relevant to the study
of politics. Many of the best-known and most-cited scholarly tomes of the last two
decades - including works on voting (Cain, Ferejohn, and Fiorina 1987), the U.S.
Congress (Cox and McCubbins 1993; Kiewiet and McCubbins 1991; Krehbiel 1991;
Rohde 1991; Shepsle 1978,1979; Shepsle and Weingast 1981,1995), political parties
(Aldrich 1995), elections (Cox 1997; Enelow and Hinich 1985; Fiorina 1981; Hinich
and Munger 1994; Popkin 1991), interest groups (Olson 1965; Ostrom 1990), international
relations (Bueno de Mesquita and Lalman 1992; Powell 1990), and comparative
politics (Bates 1989; Laver and Schofield 1990; Laver and Shepsle 1996) are
influenced by rational choice theories.
Downloaded from Cambridge Books Online by IP 147.251.110.115 on Tue Mar 08 09:25:31 GMT 2016.
http://dx.doi.org/10.1017/CBO9780511805813.001
Cambridge Books Online © Cambridge University Press, 2016
Lupia, McCubbins, and Popkin
The Oxford English Dictionary (OED) is arguably the most authoritative
and general source for determining the print-based origins of
English words. The 1989 OED contains the following definitions of the
word rationality.
1. The quality of possessing reason; the power of being able to exercise
one's reason.
2. The fact of being based on, or agreeable to, reason. A rational or reasonable
view, practice, etc.
3. The tendency to regard everything from a purely rational point of
view.
This definition, while not multiplicitous, is not very informative. In
each case, rationality is defined in terms of the words rational or reason.
Consider OED's definitions of the term rational.
1. Having the faculty of reasoning; endowed with reason.2
2. Exercising (or able to exercise) one's reason in proper manner; having
sound judgment; sensible, sane.
3. Of, pertaining, or relating to reason.
4. Agreeable to reason; reasonable, sensible; not foolish, absurd, or
extravagant.
5. Descriptive of methods of analysis and planning that make use of calculation
to bring about a projected result, esp. in economic or social
organization.
While we will discuss reason in greater detail below, the OED defines
it3
as
2
Older uses of rationality in the OED are quite clear about rationality's dependence
on reason. Consider, for example, usage from the year 1398: "the soule racional, in
that he vsyth contemplacion, he hyghte speculativus"; from the year 1547: "the
racionall sences consisteth in reason, the whiche doth make a man or woman a reasonable
beaste"; and from the year 1615: "We determine that the Braine is the Pallace
of the Rationall Soule." Note that these uses, unlike contemporary definitions, predate
the advent of scientific revolution and originate from Europe - England, in particular.
As a result, they were offered in historical contexts where men could employ reason,
but it was sacrilege to assert that anyone but God could be omniscient. Hence, at the
time of these uses of the term rationality, rationality and omniscience were clearly distinct
concepts. A problem with many modern uses of the term rationality, by critics
and proponents of rational choice theorists alike, is the belief that rationality implies
omniscience (i.e., the belief that someone making a mistake is sufficient to judge them
irrational). The rationality definition we advocate restores the separation between the
two concepts. The confounding of omniscience and rationality has, in our opinion,
severely impeded scientific progress, particularly in the last half century.
3
We recite here the relevant definitions of the word, omitting irrelevant definitions,
such as "A statement of some fact (real or alleged) employed as an argument to justify
or condemn some act, prove or disprove some assertion, idea, or belief."
Downloaded from Cambridge Books Online by IP 147.251.110.115 on Tue Mar 08 09:25:31 GMT 2016.
http://dx.doi.org/10.1017/CBO9780511805813.001
Cambridge Books Online © Cambridge University Press, 2016
Beyond Rationality
1. That intellectual power or faculty (usually regarded as characteristic
of mankind but sometimes also attributed in a certain degree to the
lower animals) which is ordinarily employed in adapting thought or
action to some end; the guiding principle of the human mind in the
process of thinking.
2. The ordinary thinking faculty of the human mind in a sound condition;
sanity.
Many of the controversies concerning the social scientific usage of
rationality are apparent in the definition of the word rational. Consider,
for example, the OED's characterizations of what rational behavior must
not be. It must not be "foolish." It must not be "extravagant." In addition,
it must be "sane." Yet these terms, each descriptive of behavior,
complicate more than they clarify. Who is to judge what constitutes foolishness
or extravagance? Even the meaning of sane behavior is subject
to broad interpretation (e.g., Foucault 1965).
A consensus on the constitution of a scientifically useful rationality
definition would ameliorate the problems caused by multiplicity
of meaning. A consensus that rationality is wealth-maximizing
behavior, or that a behavior is rational only if it is the same one that
an omniscient calculator would choose, would help determine whether
rationality is a useful basis for social scientific explanation. But no
consensus has emerged, even within economics - the discipline with
which rational choice explanations are most commonly associated. The
present condition of the term rationality is that it has multiple personalities.
Consider, for example, these contemporary definitions of the
term:
1. "A decision is only rational if it is supported by the best reasons and
achieves the best possible outcome in terms of all the goals."4
2. "Rational choice theory holds that choices among relevant goods
involve comparing all goods against each other to make correct
choices dictated by preference schedules."5
3. "The second basic assumption of positive economics is that
people are rational in the sense that they have an objective and
pursue it in a reasonably consistent fashion. When considering
persons, economists assume that the objective being pursued is utility
maximization; that is, people are assumed to strive toward the goal
4
From Herbert A. Simon, "Rationality in Political Behavior," in Political Psychology
16 (1995): 48. Simon provides this definition as modern neoclassical economics'
standard of rationality.
5
From Robert E. Lane, "What Rational Choice Explains," in Critical Review 9
(1995): 110.
Downloaded from Cambridge Books Online by IP 147.251.110.115 on Tue Mar 08 09:25:31 GMT 2016.
http://dx.doi.org/10.1017/CBO9780511805813.001
Cambridge Books Online © Cambridge University Press, 2016
Lupia, McCubbins, and Popkin
of making themselves as happy as they can (given their limited
resources)."6
4. "We have assumed, in particular, that consumers always strive to
obtain the greatest possible satisfaction from spending their incomes,
and that businessmen always try to make the biggest profits they
can."7
5. "Beyond requiring that the utility function is consistent, no specific
content is specified for it. As far as economics is concerned, a utility
function can assign as much or more utility to giving away goods as
to consuming them; it can award as much utility (to me) for an
increase in the living standard of an Indian peasant as it does for the
same increase in my own living standard. Rationality is orthogonal
to selfishness [emphasis added]."8
6. "A 'rational agent' in such a model is simply one who obeys certain
axioms, and that is the end of it."9
7. "A key economic assumption is that individuals, in making choices,
rationally select alternatives they perceive to be in their best interests.
. . . This reliance ori rational self-interest should not be viewed as
blind materialism, pure selfishness, or greed. We all know people
whose favorite radio station is WIIFM (What's In It For Me?), but
for most of us self-interest often includes the welfare of our family,
our friends, and perhaps the poor of the world."10
8. "[RJational choice involves three optimizing operations. The action
that is chosen must be optimal, given the desires and beliefs of the
agent. The beliefs must be optimal, given the information available
to the agent. The amount of resources allocated to the acquisition of
information must be optimal..." (Elster 1999: 285).
9. "Virtually all human behavior is rational. People usually have reasons
for what they do, and if asked, can opine what these reasons are"
(Simon 1995: 48).
From an intellectual distance, it is possible to agree with all of
these definitions - or at least to be offended by none of them. But
for the purpose of scientific inquiry, the conflict inherent in these defi-
6
From Modern Labor Economics: Theory and Public Policy, by Ronald G.
Ehrenberg and Robert S. Smith, 6th ed. (Reading, MA: Addison-Wesley, 1997), p. 4.
7
From A Textbook of Economic Theory, by Alfred W. Stonier and Douglas C.
Hague, 4th ed. (New York: John Wiley & Sons, 1973), p. 657.
8
Simon 1995: 48.
9
From Rationality, Allocation, and Reproduction, by Vivian Walsh (Oxford:
Clarendon Press, 1996), p. 3.
10
From Macroeconomics: A Contemporary Introduction, by William A. McEachern,
3rd ed. (Cincinnati: College Division, South-Western Publishing Co., 1994), pp.
6-7.
Downloaded from Cambridge Books Online by IP 147.251.110.115 on Tue Mar 08 09:25:31 GMT 2016.
http://dx.doi.org/10.1017/CBO9780511805813.001
Cambridge Books Online © Cambridge University Press, 2016
Beyond Rationality
nitions is detrimental. For example, someone who wants to explain
political behavior may have to grapple with a question such as this:
What will people do if forced to choose between obtaining the greatest
possible satisfaction from spending their incomes, attempting to
make the biggest profits they can, or considering the welfare of the poor
of the world? Conventional notions of rationality give inconsistent
guidance.11
If we can distill these many definitions of rationality into one that is
sensible empirically and widely applicable, we can avoid much of the
confusion currently associated with the concept of rationality and, as a
result, craft better explanations of why people do what they do. We will
now argue for such a definition. The basis of our argument is that there
is at least one issue on which these many definitions of rationality agree.
The issue is that people have reasons for the choices they make. That is,
regardless of people's genetics or socialization, if they are able to make
choices, then reasons will precede these choices. Therefore, we conclude
that a rational choice is one that is based on reasons, irrespective of what
these reasons may be.
Given the definition we offer, it may be the case that all of the choices
we observe are rational. Of course, rational choice theorists have made
this claim before, much to the delight of the approaches' critics.12
In this
case, however, such derision misses a fundamental point of the social scientific
enterprise. If our collective scientific goal is to explain why people
do what they do, then our task is to understand the reasons for the
11
Ordeshook 1986 also articulated this problem in the introduction to his book
Game Theory and Political Theory:
The word "rational" is commonly used to summarize our assumptions about
these choices. The meaning of this word has been the subject of a lively debate
during the past 20 years and has given rise to considerable misunderstanding
among the antagonists. Instead of entering this debate, we note that purposeful
choice does not necessarily mean that people carefully and consciously list
their alternative actions, map all the relevant or possible consequences of each
act, estimate the probability of each consequence, and define precisely their
preferences across all consequences. Thus, we cannot ignore habit, instinct, and
the use of simple cues and heuristics to uncomplicate complex decisions.
Indeed, one of the formal theorists' most important activities is to understand
why various heuristics are reasonable responses to complex environments and
to the costs of alternative modes of making decisions. The presumption of purposeful
choice implies simply that, after taking account of people's perceptions,
values, and beliefs, we can model their decisions by asserting that they act as
if they make such calculations.
12
The critics' delight stems from the idea that if we can conjure a rational explanation
for all observed behaviors, then rational choice theories are post hoc and
explain nothing. In our claim that all behavior is rational, our intent is to draw
maximum attention to the fact that reason precedes choice and to the implication
that explaining many choices is aided by understanding human reason.
Downloaded from Cambridge Books Online by IP 147.251.110.115 on Tue Mar 08 09:25:31 GMT 2016.
http://dx.doi.org/10.1017/CBO9780511805813.001
Cambridge Books Online © Cambridge University Press, 2016
Lupia, McCubbins, and Popkin
choices they make. Whether we agree with these reasons or not, whether
these reasons make sense to us or not, and whether we use the term rationality
to describe the process by which these reasons are formed or not
is irrelevant.13
If we desire improved explanations of why people do what
they do, then we must introduce greater clarity about the properties of
human reason into our analyses.
STANDARD ECONOMIC RATIONALITY, BOUNDED
RATIONALITY, AND ELEMENTS OF REASON
In the search for improved explanations of human behavior, scholars
have offered many distinct definitions of rationality. The continuing
battle over the meaning of rationality continues to produce many new
and important insights (see, e.g., Rubenstein 1998 for a review). To
clarify our plan for incorporating reason into the study of politics,
however, we restrict our focus to two of the most common definitions
of rationality. Both are often cited in debates about rational choice theory
and distinguish themselves from the extant population of definitions
because they are so well known.
The definitions are standard economic rationality (aka homo economicus,
substantive rationality) and bounded rationality. Each definition
is based on a distinct assumption about how people reason. In what
follows, we will compare and contrast these two focal definitions to show
why progress in the attempt to explain social behavior requires more
advanced concepts of reason than most social scientists currently employ.
Standard Economic Rationality
The standard economic definition of rationality (also commonly
employed in rational choice theories of politics) equates rational actors
with omniscient calculators.14
Consider, for example, the rationality definition
in Kreps's (1990a) microeconomic theory text. Kreps's rational
actor is omniscient with respect to the domain of choice. By his definition,
rational actors know the consequences of all the actions available
to them and choose the action whose consequences provide the highest
benefit. Kreps defines a player as "not fully rational" when "she doesn't
quite understand the full implications of her actions" (480).
13
With respect to this goal, normative judgments about the quality of people's
reasons, while appropriate in philosophical discourse or casual political conversation,
are no substitute for a logically valid and empirically sound social scientific explanation
of political or economic behavior.
14
While there exist economic definitions of rationality that do not fit this description
exactly, it is a reasonable description of many economic treatments.
Downloaded from Cambridge Books Online by IP 147.251.110.115 on Tue Mar 08 09:25:31 GMT 2016.
http://dx.doi.org/10.1017/CBO9780511805813.001
Cambridge Books Online © Cambridge University Press, 2016
Beyond Rationality
While this definition of rationality is not completely realistic, it has
been very useful in application. The unique success of contemporary
microeconomic theory as an explanatory endeavor is a great testament
to the standard definition's effectiveness. This success should also be no
surprise, for in many of the cases that economists examine, interested
audiences do not find controversial the assumption that people act "as
if" they were omniscient calculators.
Standard economic rationality is not as widely accepted in political
science. Many political scientists know that when you ask the typical
political decision maker, be it a voter, a legislator, or a juror, a detailed
question about a political matter, he or she often offers incorrect
responses. Therefore, it is only natural that many political scientists are
skeptical of explanations of political behavior that rely on the assumption
that political actors act "as if" they are omniscient calculators. This
characteristic of political decision making has led to heated debates in
the discipline about the usefulness of rationality-based theories and proposed
alternatives to standard economic rationality (see, e.g., Critical
Review 9, no. 102, 1995).
Bounded Rationality
Herbert Simon argued that unlike homo economicus, people are not
omniscient calculators - they do not do all of the calculations all of
the time. Simon affected a generation of social scientists by introducing
bounded rationality - a framework for explaining behavior that recognizes
that human cognition is limited. His insight, so controversial at first
and so obvious in retrospect, challenged scientists to consider a very different
view of economic and political behavior.
The concept of bounded rationality found a receptive audience in
political science. It is clear that political decision makers have limited
ability to perceive and calculate. However, while bounded rationality
seems like the perfect alternative to standard economic rationality, the
literature growing out of this concept has had two serious shortcomings.
The first shortcoming is that bounded rationality is often incorrectly
presumed to imply nonmaximizing or non-self-interested behavior. This
presumption leads to a false polarization between bounded rationality
and other rationality definitions. As Jensen and Meckling (1976: 307)
observe:
Simon's work has often been misinterpreted as a denial of maximizing behavior,
and misused, especially in the marketing and behavioral science literature. His
later use of the term 'satisficing' (Simon 1959) has undoubtedly contributed to
this confusion because it suggests rejection of maximizing behavior rather than
maximization subject to costs of information and decision making.
9
Downloaded from Cambridge Books Online by IP 147.251.110.115 on Tue Mar 08 09:25:31 GMT 2016.
http://dx.doi.org/10.1017/CBO9780511805813.001
Cambridge Books Online © Cambridge University Press, 2016
Lupia, McCubbins, and Popkin
Indeed, Simon (1995: 45) later admitted that "virtually all human behavior
is rational. People usually have reasons for what they do."15
A second problem with the concept of bounded rationality as it is
often interpreted is that it is underdefined. Simon forced scholars to recognize
that people are limited in their ability to seek and process information.
This was brilliant and valuable. Yet beyond that, Simon's
bounded rationality offers little systematic guidance as to where the
bounds of rationality are. Simon provides his most specific clues about
these bounds when he discusses irrationality. Consider the following
passage (1985: 297):
We may deem behavior irrational because, although it serves some particular
impulse, it is inconsistent with other goals that we deem more important. We
may deem it irrational because the actor is proceeding on incorrect facts or ignoring
whole areas of relevant fact. We may deem it irrational because the actor has
not drawn the correct conclusions from the facts. We may deem it irrational
because the actor has failed to consider important alternative courses of action.
If the action involves the future, as most action does, we may deem it irrational
because we don't think the actor uses the best methods for forming expectations
or for adapting to uncertainty. All of these forms of irrationality play important
roles in the lives of every one of us, but I think it is misleading to call them irrationality.
They are better viewed as forms of bounded rationality.
Of course, this directive is problematic in that it does not lend itself
to better explanations of why people do what they do. For example, how
would a scholar seeking to render a positive explanation (rather than a
normative description) of political behavior determine "areas of relevant
fact" or "important alternative courses of action." Is a voter who bases
her choice on the endorsement of a political party or an interest group
(such as the Sierra Club or the NRA) less "rational" than she would have
been had she based her vote on a careful consideration of 100 candidate
issue positions? If the answer to this question seems to be an obvious
yes, then consider the fact that she may draw the same conclusion and
have the same degree of certainty about the consequence of her choice
15
The term satisficing, when employed post hoc, offers no clarity over and above
that provided by standard definitions of rationality. This is why the many satisficingbased
explanations of behavior that are unaccompanied by insights on the trade-offs
and reasons that precede a decision to satisfice, or without evidence on the cognitive
mechanisms that would activate a satisficing algorithm, have had no discernable
impact in the fields of political science and economics. Indeed, we have encountered
many scholars who interpret the possibility of satisficing as a license to throw away
all of rational choice theory, including its focus on motivations and its logical rigor.
This is nonsense. The idea that humans are limited cognitive entities, far from relieving
scholars from the burden of rigorous analysis, requires us to apply the most reliable
scientific methods we can muster to understanding the complex contours of
human reason.
10
Downloaded from Cambridge Books Online by IP 147.251.110.115 on Tue Mar 08 09:25:31 GMT 2016.
http://dx.doi.org/10.1017/CBO9780511805813.001
Cambridge Books Online © Cambridge University Press, 2016
Beyond Rationality
in both cases. What then would distinguish rational choice from irrational
choice? More importantly, how would such a distinction, if it
could be made, help scholars explain why people do what they do? The
simple answer is that such a distinction would not be helpful. Of course,
Simon realizes this. He concludes that it is impossible to generalize about
the bounds of rationality as the following passage (1985: 297) indicates:
"To understand and predict human behavior, we have to deal with the
realities of human rationality; that is bounded rationality. There is
nothing obvious about these boundaries; there is no way to predict, a
priori, where they lie."
Why We Seek Elements of Reason
If our goal is to explain why people do what they do, then both the
standard economic treatment of rationality and the common interpretation
of bounded rationality are of limited applicability. Standard economic
rationality assumes that rationality is merely the pursuit of
objectives, given perfect information and unlimited informationprocessing
capacity. But, in reality, omniscience is impossible. People
never have all of the facts; further, even if an individual did have all of
the facts, then cognitive limitations would often render him or her unable
to perform the calculations necessary to deduce an optimal choice. So,
from the perspective of explaining many political behaviors, standard
economic rationality is either an impossible fantasy or a stylization, but
not a realistic option.
The concept of bounded rationality tells us this much, but nothing
more. Worse yet, the initial attractiveness of bounded rationality has led
many observers to confound the fact that people are not omniscient calculators
with the notion that their decisions are not the product of reason
(e.g., analysts labeling as "irrational" those actions based on reasons they
either disagree with or do not understand). This confusion has not only
retarded debates about the applicability of rationality; it has hampered
our discipline's ability to get on with the business of explaining why
people do what they do. If a primary objective is to explain why people
do what they do, we accomplish it more effectively by understanding
how people reason. It is time to move beyond the heated, but fruitless,
debate about whether people are rational and toward a scientific search
for the elements of reason.
It is at this juncture that our efforts in Elements of Reason part ways
with many extant rational choice theories and Simon's view of bounded
rationality. We argue that human reason has systematic attributes and
that social scientists should learn these properties and incorporate them
into their explanations. We advocate an approach to the study of poli-
11
Downloaded from Cambridge Books Online by IP 147.251.110.115 on Tue Mar 08 09:25:31 GMT 2016.
http://dx.doi.org/10.1017/CBO9780511805813.001
Cambridge Books Online © Cambridge University Press, 2016
Lupia, McCubbins, and Popkin
tics that combines the premise that people have reasons for what they
do with the premise that our treatment of how people reason should be
informed by modern scholarship about how cognition and affect affect
information processing.
PLAN OF THE BOOK: BUILDING SIMON'S BRIDGE
Simon urged social scientists to think about the cognitive limits that
underlie human behavior and to pursue what individual and collective
actions are possible given those limits. At the same time, he proposed a
bridge between the sciences of the mind and social sciences such as economics
and political science. But in the 1950s and 1960s, the era in
which Simon first advocated such a bridge, the tools for building it were
not available.
At the time of Simon's initial challenge to standard economic rationality,
the cognitive and neural sciences were in their infancy. Throughout
the 1950s and 1960s, for example, psychology was still emerging
from an era dominated by behaviorism and the belief that behavior can
be explained without peering into the black box of the brain (Delgato
and Midgley 1992). And cognitive science, as we know it today, did not
exist. As a result, experimental evidence and theorizing on the contours
of human reason were quite sparse.
In the years since Simon's initial contributions, the science of the mind
has undergone great change. Advances in technology have allowed scholars
ranging from psychology to biology to observe brain-behavior
correspondences more accurately (e.g., Damasio 1994). Advances in
computer technology have allowed for estimation and simulation procedures,
such as neural networks, that allow evaluation of complex
hypotheses about reasoning and behavior (e.g., see Churchland and
Sejnowski 1991, Clark 1997, and Crick 1994 for reviews of progress in
this field; see McCauley 1997 for current controversies). Moreover,
advances in the technology of game theory, such as Harsanyi's (1967,
1968a, b) innovations regarding games with incomplete information,
provide a window through which clear links between rationality and
reason can be seen. Put another way, the scientific advances of the last
four decades give us an opportunity that Simon did not have when he
forged the concept of bounded rationality - the opportunity to build
Simon's bridge.
In the chapters that follow, we take the next step in building that
bridge and toward offering more effective explanations of why people
do what they do. Like Simon, we proceed with the belief that an exhaustive
brain-behavior mapping is an unrealistic goal. However, such a
mapping is unnecessary for improving our ability to explain why people
12
Downloaded from Cambridge Books Online by IP 147.251.110.115 on Tue Mar 08 09:25:31 GMT 2016.
http://dx.doi.org/10.1017/CBO9780511805813.001
Cambridge Books Online © Cambridge University Press, 2016
Beyond Rationality
do what they do. Between a brain-behavior mapping and such widely
used concepts as standard economic rationality and bounded rationality
are other alternatives - explanations of behavior that incorporate wellestablished
elements of reason and provide more precise inferences about
why people do what they do. The chapters that follow represent the
responses of scholars from different intellectual perspectives to the following
challenge: help broad, scientific audiences understand elements
of reason.
The chapters of the book are divided into two parts. The chapters in
Part I describe external elements of reason. External elements of reason
are incentive-altering forces outside the body, such as social norms and
political institutions, that affect the collection and processing of information.
In Part II, the focus switches to internal elements of reason. Internal
elements of reason are brain-based factors inside the skin, such as
affective states and prior knowledge, that have systematic effects on how
individuals seek and process political information.
All chapters are unified in at least two respects. First, each chapter
introduces a distinct element of reason. Second, each chapter challenges
conventional wisdom. Common themes, such as "voters lack consistent
attitudes" or "heuristics save the day," are eschewed in favor of more
instructive explorations of the contours of human reason. As a result,
each chapter points to a potentially important new direction in the study
of political and economic behavior. Moreover, each chapter, in the way
that it addresses the question at hand, reinforces our contention that
advances in the social scientific explanation of reasoning and choice will
depend less on whether people know what analysts thought they should
have known and more on scientists' abilities to understand how people
process the information they have.
Part I: External Elements of Reason
In the first three chapters of Part I, political institutions are the surprising
source of elements of reason. We say surprising because the focus of
the modern study of institutions has been on the way that institutions
aggregate individual preferences into collective outcomes and on the way
that institutions affect individuals' strategic considerations. In the chapters
by Denzau and North, Lupia and McCubbins, and Sniderman, by
contrast, the authors argue that institutions also affect what people do
by structuring their choices and their beliefs.
In "Shared Mental Models: Ideologies and Institutions," Arthur
Denzau and Douglass North argue that the "mental models that the
mind creates and the institutions that individuals create are both essential
to the way human beings structure their environment in their inter-
13
Downloaded from Cambridge Books Online by IP 147.251.110.115 on Tue Mar 08 09:25:31 GMT 2016.
http://dx.doi.org/10.1017/CBO9780511805813.001
Cambridge Books Online © Cambridge University Press, 2016
Lupia, McCubbins, and Popkin
actions with it." They motivate their arguments for greater attention to
mental models by carefully differentiating cases in which the standard
economic definition of rationality has succeeded and failed. They then
suggest that replacing "the black box of the 'rationality' assumption used
in economics and rational choice models" is the key to better explanations
of many social behaviors.
The replacement they suggest is a theory of choice that recognizes the
central role of shared mental models. The idea is that many social interactions,
including most of those that we would recognize as political,
require collective action. They continue that the evolution of collective
endeavors will be shaped by the mental models (e.g., conceptualizations
of what causes what in the world) that people share. They conclude that
the types of uncertainty that social actors face and the types of institutional
environments in which they interact constrain the set of mental
models upon which collective action can be based. Denzau and North
thus provide substantive insight into the determinants of economic and
political development, as well as a methodological guide for incorporating
cognitive limits into formal theories of choice.
Arthur Lupia and Mathew D. McCubbins's "The Institutional Foundations
of Political Competence" continues the demonstration of how
institutions affect reasoning. Their chapter focuses on communication.
They begin with the premise that many political decision makers cannot
learn what they need to know from direct experience (e.g., most people
cannot experience the consequences of passing NAFTA or social security
reform until after they are passed). As a result, many political decision
makers must learn what they can from the testimony of others
(voters from politicians and the media, juries from witnesses, legislators
from bureaucrats, and so on). This reality implies that the types of decisions
that people make depend on whom they choose to believe.
Lupia and McCubbins show how political institutions affect "who
believes whom" by identifying conditions under which political institutions
clarify others' incentives. They then draw on case studies, experiments,
and formal modeling results to show when and how institutions
allow people to make more effective decisions about whom to believe.
Like the approach of Denzau and North, theirs blends the logical rigor
of formal theory with insights about learning from the cognitive sciences.
Lupia and McCubbins's research clarifies and amends extant explanations
of when and how voters, jurors, and legislators learn what they
need to know.
In "Taking Sides: A Fixed Choice Theory of Political Reasoning," Paul
M. Sniderman argues that institutions affect reasoning by fixing certain
choices. His argument is quite innovative in that it builds an institutioncentered
argument about politics from a social psychological perspective
14
Downloaded from Cambridge Books Online by IP 147.251.110.115 on Tue Mar 08 09:25:31 GMT 2016.
http://dx.doi.org/10.1017/CBO9780511805813.001
Cambridge Books Online © Cambridge University Press, 2016
Beyond Rationality
- the usual projection from that perspective is an attempt to explain
political behavior by focusing exclusively on citizen characteristics. Passages,
such as the following, reveal the extent to which Sniderman's argument
departs from the social psychological research tradition from which
he hails:
Initially, we asked how citizens effectively can simplify political choices so as to
make them coherently. Putting the question this way led us, like virtually everyone
else, to start the explanatory process by focusing on the characteristics of
citizens. How much attention do they pay to politics? What do they know about
it? What political ideas do they hold and how are they organized? Answer these
questions, and we should be in a position to figure out how citizens make political
choices. Or so it seemed then. Now, I am persuaded, we had the order of
things wrong.
Throughout the chapter, Sniderman builds the foundations of a
theoretical framework in which citizens "do not operate as decision
makers in isolation from political institutions. If they are in a position
to overcome their informational shortfalls by taking advantage of
judgmental shortcuts, it is because public choices have been organized
by political institutions in ways that lend themselves to these shortcuts."
His argument clarifies the path to a way of thinking where psychological
considerations of reason and the institutional structure of choice
combine to produce more effective explanations of why people do what
they do.
In the fourth through sixth chapters of Part I, public pressures generate
external elements of reason. In the chapter by Frohlich and Oppenheimer,
social norms are the public pressures that induce individuals to
give up material gain for the benefit of anonymous strangers. For Iyengar
and Valentino, campaign advertisements are the form of public pressure
in question. In both chapters, clever experiments reveal systematic
factors that determine how these public pressures affect reasoning and
choice. In the chapter by Rahn, the source of pressure is public mood.
Rahn uses data from several surveys to show how certain very public
events (e.g., a war) lead entire populations to interpret new pieces of
information in distinct ways.
The elements of reason in Norman Frohlich and Joe Oppenheimer's
"How People Reason about Ethics" are social norms that induce selfinterested
individuals to engage in other-regarding behavior. Their
chapter begins by alerting us to the fact that motives for altrustic behavior
are hard to disentangle - selfish behavior and other-regarding behavior
are often observationally equivalent. For example, if Joe helps Norm,
it is hard to know whether Joe does it primarily for his own sake or primarily
for the sake of Norm.
15
Downloaded from Cambridge Books Online by IP 147.251.110.115 on Tue Mar 08 09:25:31 GMT 2016.
http://dx.doi.org/10.1017/CBO9780511805813.001
Cambridge Books Online © Cambridge University Press, 2016
Lupia, McCubbins, and Popkin
Because selfish and other-regarding behaviors are often difficult to disentangle,
it can be hard to evaluate claims about the frequency or causes
of other-regarding behavior. Frohlich and Oppenheimer have overcome
some of these difficulties with a series of clever experiments. In their
experiments, subjects have an opportunity to leave money for others.
The donations are totally anonymous - the experiment is designed so
that neither the experimenter nor the recipient can determine the source
of any particular contribution. Their experiments reveal regular patterns
in other-regarding behavior. In particular, they find that altruism is mediated
by subjects' considerations of the unseen recipients' moral worthiness
and by the personal costs of engaging in other-regarding behavior.
Put another way, altruism in these experiments follows some simple,
perhaps even economic, rules.
In Shanto Iyengar and Nicholas Valentino's "Who Says What: Source
Credibility as a Mediator of Campaign Advertising," the source of public
pressure is campaign advertisements, and the element of reason in question
is the role of credibility in cue-based processing. A truth of modern
politics is that some of the most important political communications are
sent and received through the media. A particularly interesting class of
such communications is the political advertisement - 30 to 60 seconds
of carefully constructed prose and imagery. Although many ads are powerful,
political advertisements vary in how effectively they build or reinforce
support for the cause at hand.
Iyengar and Valentino offer a series of experiments that explain some
of this variation. They focus on the relationship between a candidate's
ideological background and his or her claims. Whereas successful candidates
do not typically choose campaign themes on which they are not
credible, the candidates in Iyengar and Valentino's experiments sometimes
do. That is, they show that Democratic candidates suffer after
airing advertisements on traditionally Republican issue areas and vice
versa. The results are strong and emphasize how credibility is an essential
part of effective campaigning in the television age.
Part I's final chapter is "Affect as Information: The Role of Public
Mood in Political Reasoning." In it, Wendy Rahn challenges the common
view that emotion only gets in the way of some proper or pure form of
reason. She argues instead that certain types of emotional responses may
actually help people process information more effectively. The element
of reason upon which she focuses is public mood. Public mood is a type
of time-dependent nationalism - a feeling that is good when your country
wins a war or a major sporting event, a feeling that is bad following a
national tragedy. Rahn then conducts several analyses in which she
reveals a correspondence between variations in public mood and variations
in opinion and behavior. Rahn also provides evidence that public
16
Downloaded from Cambridge Books Online by IP 147.251.110.115 on Tue Mar 08 09:25:31 GMT 2016.
http://dx.doi.org/10.1017/CBO9780511805813.001
Cambridge Books Online © Cambridge University Press, 2016
Beyond Rationality
mood acts as a filter through which other important types of information
are received and interpreted. Her research suggests that public mood
can have a very large influence on political judgment, and that such an
influence, rather than interfering with cognition, can actually aid it.
Part II: Internal Elements of Reason
In Part II, the focus is on elements of reason that are housed within the
skin. The first three chapters of Part II are attempts to construct general
theoretical frameworks for understanding how people process
information.
In James Kuklinski and Paul Quirk's "Reconsidering the Rational
Public: Cognition, Heuristics, and Mass Opinion," the elements of
reason in question are heuristics - common judgmental shortcuts that
people use to draw complicated inferences from simple environmental
cues. Kuklinski and Quirk's investigation starts at the point where psychological
and political science treatments of heuristics part from one
another. Kuklinski and Quirk remind us that psychologists view heuristic
decision making as leading to inferior decisions. Political scientists,
by contrast, have come to view heuristic decision making as therapeutic
- leading people to make the same decisions that they would have made
if better informed. This division suggests a middle ground - a set of conditions
under which heuristics aid decision making and a set of conditions
under which heuristics lead people to make different decisions than
they would have made had they possessed greater amounts of political
information.
Kuklinski and Quirk then set out to clarify the role of heuristics in
political reason by stating and testing hypotheses about the benefits and
drawbacks of heuristics usage. They review and compare results from a
series of innovative experiments about heuristics that are distinctly political.
These experiments document the fallacies that heuristic decision
making can engender in important policy contexts. Together, their argument
and experiments provide valuable clarification of the surprising
ways in which common heuristics affect political choices.
In Milton Lodge and Charles Taber's "Three Steps toward a Theory
of Motivated Political Reasoning," our attention is directed away from
heuristics and toward an element of reason known as affective charge.
In this chapter, as in previous ones, Lodge and Taber use theory and
experiments to clarify how a common aspect of human cognition affects
political choices.
Lodge and Taber's efforts extend exciting lines of research in psychology
and cognitive science concerning the deep connections between
cognition and affect. Whereas cognition and affect were once viewed as
17
Downloaded from Cambridge Books Online by IP 147.251.110.115 on Tue Mar 08 09:25:31 GMT 2016.
http://dx.doi.org/10.1017/CBO9780511805813.001
Cambridge Books Online © Cambridge University Press, 2016
Lupia, McCubbins, and Popkin
polar opposites in the context of understanding reason, an increasingly
common view is that each is necessary for the proper functioning of the
other. Lodge and Taber's research demonstrates this point in political
contexts. They show how three factors ("hot cognition" - the hypothesis
that people store feelings as well as facts about past events; "on-line
processing" - the hypothesis that the interpretation of new information
is affected by information or feelings that they already have; and a
person's affective state at the time of processing) combine to lead people
to interpret new pieces of political information in nonobvious ways.
They show how feelings become information and are treated in the reasoning
process as if they were objective facts.
Lodge and Taber's findings reinforce and extend the implications of
the relationship between affect and reasoning suggested in the chapter
by Wendy Rahn. Like Rahn's findings, those of Lodge and Taber suggest
that affect and cognition are inextricably linked. Their unique contribution,
however, is to offer us direct and controlled experimental observations
of this link. In particular, we see affect leading people to have
identifiable biases in the types of political information to which they will
attend.
In Part IPs third chapter, Samuel L. Popkin and Michael Dimock
examine how prior knowledge affects the processing of new information.
In "Knowledge, Trust, and International Reasoning," Popkin and
Dimock show how the degree to which people's trust in government and
in others corresponds to their interpretations of foreign affairs issues.
They use surveys from the American National Election Study and the
Times-Mirror polling organization to show that people who neither
understand nor trust their own government are also suspicious of foreigners,
apprehensive about international trade, and isolationist. These
surveys form the basis for their ultimate, and more general, conclusion
- that what people know about the political process and about political
institutions have systematic effects on how they reason about new events
in the political world. They conclude that prior knowledge of key political
concepts provides people with a life raft and a compass with which
they can survive and better navigate the turbulent and often complex
seas of political debate.
The closing chapters of Part II draw our attention to the limits on the
types of associations that political decision makers can and do draw.
Knowing these limits is critical to many questions in political science.
For example, understanding the conditions under which some political
actors can get others to listen to and believe that an observable social
phenomenon is associated with an unobservable political process is the
key to winning elections and public policy debates. To supply us with
elements of reason relevant to this question, we recruited a scholar with
18
Downloaded from Cambridge Books Online by IP 147.251.110.115 on Tue Mar 08 09:25:31 GMT 2016.
http://dx.doi.org/10.1017/CBO9780511805813.001
Cambridge Books Online © Cambridge University Press, 2016
Beyond Rationality
a background in social psychology, Philip E. Tetlock, and a scholar with
a background in cognitive science, Mark Turner.
The starting point of Philip E. Tetlock's "Coping with Trade-Offs: Psychological
Constraints and Political Implications" is that political interactions
involve trade-offs. Economists typically treat people as able to
engage in trade-off reasoning. Tetlock disagrees, arguing that "people are
reluctant decision makers who do their damnedest to minimize cognitive
effort, emotional dissonance, and moral ansgt by denying that important
values conflict." So when faced with little motivation to make trade-offs,
people refuse to do so. However, when they are motivated, Tetlock concludes
that "[p]eople are best thought of not as cognitive misers but as
cognitive managers who deploy mental resources strategically as a function
of the perceived importance and tractability of the problem."
Tetlock's chapter provides important clues about why people retain
certain "core values," even in the face of evidence that the rationale for
holding such values implies painful trade-offs.
The penultimate chapter of Part II is written by Mark Turner. Turner
is a central figure in one of the most fundamental debates in cognitive
science - the dynamics of concept formation and change. This debate
matters to politics because the outcomes of familiar processes, such as
persuasion, agenda setting, framing, and priming, depend on the conditions
under which concepts do and do not change.
Where the standard view of cognitive operation once represented the
mind as an inflexible serial calculator, Mark Turner's "Backstage Cognition
in Reason and Choice" draws extensively from the modern view
of the mind as a highly adaptable organ capable of drawing effective and
complex inferences from relatively simple environmental stimuli. His
argument reveals the often-obscure dynamics that differentiate persuasive
claims from nonpersuasive ones. He considers, for example, the
counterfactual claim that Hitler's influence would have been different if
Churchill had been prime minister of England in 1938 instead of Chamberlain.
He then shows us what an audience must know for such a claim
to be effective in debate, as well as the situations in which analogous
claims would be totally ineffective. While students of politics are quick
to explain the effectiveness of rhetorical strategies by focusing exclusively
on effective strategies, Turner's method of inference provides the foundation
for a more systematic and useful approach.
CONCLUSION
The future success of political science depends on our discipline's ability
to explain why people do what they do in political contexts. The chapters
in Elements of Reason are the result of an intellectual challenge that
19
Downloaded from Cambridge Books Online by IP 147.251.110.115 on Tue Mar 08 09:25:31 GMT 2016.
http://dx.doi.org/10.1017/CBO9780511805813.001
Cambridge Books Online © Cambridge University Press, 2016
Lupia, McCubbins, and Popkin
we offered to scholars working in a number of academic disciplines. We
challenged each contributor to help us understand and expand on why
people do what they do. The product, though not a tightly bound theory
of human behavior, clarifies the connections among exciting new ideas
in several scientific disciplines.
Simon's bridge is worth building, but it will take a long time to build.
Working with the contributors to Elements of Reason has given each of
us, the editors, a clearer idea of how to contribute to its construction.
We hope that it does the same for you.
20
Downloaded from Cambridge Books Online by IP 147.251.110.115 on Tue Mar 08 09:25:31 GMT 2016.
http://dx.doi.org/10.1017/CBO9780511805813.001
Cambridge Books Online © Cambridge University Press, 2016