Thinking and Feeling
A Presentation by Professor Paul Romer
Economics and Graduate School of Business,
Stanford University
Friday, March 5, 1999 at 12:30pm - 2:00pm
Anthropology, HSSB 2001A
Abstract
When economists talk about human behavior, they have tended to alternate
between two extremes. Traditionalists in economics and their followers
in fields such as political science specify a utility function and a maximization
operator and make no claims about what goes on inside the human head. When
pressed, a traditionalist will fall back to the standard "as if" position:
"All we claim is that people behave as if they maximize utility." At the
other extreme, behavioral economists, psychologists, anthropologists and
other scientists study the particulars of behavior. Until recently, scientists
working at these two extremes have traced the divergence to fundamentally
different conceptions of human nature. This essay argues to the contrary
that the tension between these approaches springs from a methodological
gap that can easily be resolved and that is already in the process of being
filled. To communicate effectively and productively with each other and
to achieve the kind of interaction between theory and evidence that is
essential for scientific progress, economists need to develop an intermediate
layer of theory that lies between these two domains. This new level of
theory involves the construction of models that look inside the human head
and distinguishes between thinking and feeling as two major classes of
mechanisms guiding choice.
How, in practice, can economists and other students of human behavior
distinguish between thoughts and feelings? The traditional approach in
economics gave the concept of utility maximization empirical content by
imposing specific restrictions. It argued that the utility function should
have as arguments only current consumption activities that are under the
control of the person in question. The three examples I will discuss show
that feelings are functions of both current variables and variables that
were realized in the past. They are also determined at least in part by
variables over which the subject may have no control. A growing body of
evidence shows that if we interpret the traditional utility function as
a one-dimensional summary of a person's feelings, these traditional restrictions
on the arguments in the utility function do not hold.
Once we admit these results, many of the superficial differences between
feelings and thoughts disappear. In particular, the mental systems that
generate feelings look like systems that collect information, store it,
make inferences, and change behavior in ways that we usually associate
only with the metal modules that produce conscious thought. However, two
suggestive differences between thoughts and feelings remain. The first
is that the systems that generate feelings can be less sensitive to --
in some cases totally insensitive to -- information that from a design
point of view should be relevant but that is encoded linguistically or
symbolically. The second difference is implicit in the use of the term
welfare. Humans agree that feelings like pain and shame are unpleasant
or bad and that feelings like accomplishment and solidarity with a group
are pleasant or good. If we interpret feelings in the broadest possible
sense, human welfare is determined by the feelings that a person experiences.
Thoughts have instrumental value for people, but feelings have intrinsic
value. Economics will not lose all of its scientific content if we admit
that people actually have feelings.
Background
The Economist identified Paul Romer's Ph.D. thesis in 1983 as
laying the foundation of "new growth theory" in economics. Up until then,
most work in macroeconomics focused on government policies that would encourage
capital accumulation or fine-tune aggregate demand with adjustments to
monetary and fiscal policy. This neo-classical approach treated scientific
discovery, technological change, innovation, and productivity growth as
peripheral concerns in national economic policy. Romer's new growth theory
moved these factors back to the center of macroeconomic analysis, made
them endogenous rather than exogenous to models of growth, and used their
interrelationships to answer the puzzle: What makes possible and sustains
economic growth in a physical world characterized by diminishing returns
and scarcity? For a developing country, the most important government policies
may be those that determine the rate of technology transfer from the rest
of the world. For an advanced economy, the most important policies may
be the ones that influence the rate of technological innovation in the
private sector. These theories have become influential in the understanding
of otherwise puzzling features of long-term national and global economic
development. Last year he was identified as one of America's 25 most influential
people by Time magazine.
As an undergraduate at the University of Chicago, Paul Romer studied
mathematics and physics. He began graduate work in economics at MIT, then
transferred back to the University of Chicago where he received his Ph.D.
in 1983. He went on to become a faculty member in the Department of Economics
at the University of Rochester, the University of Chicago, and the University
of California, Berkeley. Since July 1996, he has been a Professor of Economics
at Stanford and is now the Ralph Landau Senior Fellow of the Center for
Economic Policy Research, and has been a Fellow at the Center for Advanced
Study in the Behavioral Sciences, and a member of the Executive Committee
of the American Economic Association.
Top of page
