Senior
Honors Theses in Economics: Past and Future
Behavioral Economics
Five of
the theses for which Ralph Byrns served as advisor in the past three years have
centered on behavioral issues. Cognitive psychologists and behavioral economists
question the general applicability of some behavioral assumptions that underpin
standard economic theory (SET). Their research indicates that anomalies for SET are pervasive
and that our choices are limited by bounded
rationality, bounded self-interest, and bounded willpower. Moreover, our
decisions often fail to reflect the systematic risk aversion that economists view
as common when choices are risky, and people seem to be risk lovers when confronted with
potential losses. [See Susan Fisk's short paper here for a brief overview of the
concept of bounded rationality.]
Aspects of behavioral economics are an increasingly
important theme in the research of Nobel Prize winners in Economics. These
Nobel winners include Herbert Simon (1978), Maurice Allais (1988), James
Heckman and Daniel McFadden (2000), George Akerlof, A. Michael Spence, and
Joseph Stiglitz (2001) and Daniel Kahneman and Vernon L.
Smith (2002). Notable works by Robert Frank (Cornell), Richard Thaler
(Chicago), Robert Shiller (Irrational Exuberance), and Steve Leavitt (Freakonomics) also have a
distinctly behavioral flavor. See also 14 recent Freakonomics articles written by Leavitt and Dubner
for the New York Times Magazine .
Recent UNC Theses Focused on Behavioral Economics
|
Cognitive Anomalies
and the Assumption of Economic Rationality Douglas R. Sue, UNC BA with Highest Honors, 2004 |
Douglas Sue
used the Kahneman-Tversky questionnaire (link) in an attempt to ascertain if economists are more
rational than most people, and if they are more
consistent in dealing with risk. Sue’s research
queried 100 UNC juniors and seniors. One-third were
majoring in liberal arts, one-third were majoring in mathematics or the
sciences, and one-third were economics majors.
Findings. Sue’s results indicate that
people who choose to be educated in economics tend to be more rationally
consistent and to treat risk more systematically than is true of
students who major in other social sciences (broadly construed) or in
mathematics or the sciences.
Awards: In addition to yielding Highest Honors, Doug’s work was awarded Best Senior
Thesis by the UNC Department of Economics, and he won a statewide thesis
competition at the Duke University Economics Symposium in 2004.
|
Pricing and Advertising Strategies and Their
Effect on Consumer Decisions Jennifer Wade, UNC BA with Highest Honors, 2004 |
Jenny
Wade secured data covering purchases in a grocery store to test whether the way
prices are “framed”
affect consumer behavior, contrary to standard economic theory [SET]. For example, if whether an item’s sale price is stated as 2/$4
versus $2 per unit affects consumer purchasing patterns, then consumers are not
acting in accord with SET.
Findings. Wade’s
study of more than 100,000 consumer purchases of nine items compared the
amounts people bought at regular prices [e.g., $3.19] versus purchases at sale
prices [e.g., 2/$4]. When prices were stated as, e.g.,
2/$4, customers bought in increments specified in the
stated price. In other words, even though 2/$4 translates to $2 per unit,
consumers bought 2 or 4 or 6 (etc.) units far more often than they bought 1 or
3 or 5 units. Wade’s results were quite robust, with
confidence intervals for seven items at the 99+%
range, and for the other two items, the confidence intervals exceeded 95
percent.
|
Economic Reasoning
and the Academic Majors of UNC Students Alexandra Samet, UNC BA with Highest Honors, 2005 |
Alexandra
Samet’s study focused on whether taking courses in economics
or business administration improve the logical consistency and made risk aversion a more consistent pattern. Samet extended the Sue results by having 50
undergraduates majoring in economics and 50 undergraduates majoring in business
administration answer the K-T
questionnaire. All of the
respondents in her study had credit for the Introduction to Economics course
Findings. Samet’s research strongly
suggests that taking more advanced courses in economics or
business does not improve the rational consistency of people, nor does
it change their treatment of risk. Moreover, Samet’s results confirmed the results of Sue to the effect that
economists are “more rational” and “more consistent in their treatment of risk”
than are other people, and found that business majors not also specializing in
economics were no more likely to be rationally consistent than were liberal
arts or science and math majors, nor were they any more consistent in their
treatment of risk.
Inasmuch
as all of the students studied by Samet had credit for principles of economics
courses, further research needs to be undertaken to
ascertain if rational consistency and treatment of risk are affected by exposure
to introductory economics courses.
|
Behavioral Analysis of Commercial Bank
Mergers: Are Banks Really Profit Maximizers? Thomas Hodges,
UNC BA with Highest Honors, 2005 |
Tom Hodge’s study addressed whether the restructuring (branch
consolidation) that follows bank mergers maximizes profit and the present
values of stockholders in the surviving bank. Specifically, when
banks with overlapping territories merge and numerous branches of the two
original banks are geographically close to each other, does the
acquiring bank tend to close branches of the acquired bank instead of the
acquirer bank more than standard profit-maximizing assumptions would predict?
The unavailability of any comprehensive database identifying the branch
consolidations following mergers resulted in this research being a case study
of the 2004 acquisition of SouthTrust by Wachovia, and the consequent 173
branch consolidations.
Findings. Hodges found that, after adjusting for the distances
between branches, size, and other proxies for expected profitability (e.g.,
deposits and loan activity), the branches of the acquired bank were twelve (12)
times more likely to be closed than the branches of
the acquiring bank. This result seems much less consistent with profit
maximizing decisions than with how powerfully top managers of acquiring banks
conformed to Adam Smith’s principle of the relative
influence of “custom” (i.e., proximity, or familiarity, or affection) when
compared to “moral sentiments” (e.g., utility or profit maximization).
|
Male to Female
Ratios and Female Attractiveness Susan Fisk,
UNC BA with Highest Honors, 2005 |
Data |
Susan Fisk’s study addressed whether women compete more
vigorously for the attention of males when the ratio of women to men in a
population increases. She had a panel of more than 120 UNC students judge, on a
scale of 1 to 10, pictures posted on Facebook by more than 1000 Caucasian female students at more than
20 colleges in the Southern United States. Her major hypothesis was that
women would post more attractive pictures of themselves
if fewer males were available per female on these college campuses. Other
hypotheses considered whether the backgrounds of the “judges” resulted in
systematic differences in the evaluations of these pictures. This link is to a
paper Susan wrote that summarizes some of the literature on the economics of
beauty, considered behaviorally.
Findings. Susan’s basic hypothesis was not supported by the data. However, three statistically significant
results emerged from her analysis of how the backgrounds of “judges” (i.e.,
gender, race, and geographic place of residence prior to enrolling at UNC)
affected their evaluations of female attractiveness. First, women were more
generous than males, on average, in the “attractiveness” scores that UNC
students assigned to the pictures posted on Facebook.
Second, the attractiveness scores of Caucasian and non-Caucasian judges were
not distinguishably different. Third, geographic region of origin of the judges
(i.e., states close to North Carolina versus other states or nations of origin)
had no influence on the attractiveness scores. That is, there seems to be no
regional differences in perceptions of attractiveness.
Awards: In addition to graduating with Highest Honors, Susan’s work was awarded Best
Senior Thesis by the UNC Department of Economics, and she placed second
in the 2006 statewide thesis competition at the Duke University Economics
Symposium. These accomplishments were
remarkable given that her major hypothesis was rejected,
and stand as evidence of the strength of Susan’s
writing, and the rigorous methodology used in her study.
Possible Theses Topics for 2007
Risk, Rationality, and Familiarity with Basic
Economic Theory
This thesis
would extend the works of Sue [2004] and Samet
[2005] to address the issue of whether UNC students arrive at college with a
well-developed sense of economic reasoning, or whether some aspects of economic
reasoning develop as students are initially exposed to
the material in an Introduction to Economics course. [Do entering UNC students
with little or no previous exposure to economics but with a predilection for
economic reasoning tend to major in economics moreso than similarly
inexperienced students without such a predilection?]
This thesis would entail examining how students perform on the Kahneman-Tversky questionnaire
when administered on the first day of an Introduction to Economics course
relative to their performance after taking this course. I administered the K-T
questionnaire to roughly 350 beginning students at the start of Fall 2005, and at the end of the semester. These data
include a lot of demographic information about the students. The resulting
database is not yet analyzed.
Alexandra
Samet analyzed the K-T questions for how alternative responses conform or fail
to conform to standard economic theory versus “prospect
theory,” the label Kahneman and Tversky applied to
choices that they perceived as inconsistent with SET. Samet identified several
questions that were misclassified by K-T as
inconsistent with SET. This thesis would benefit significantly by the
development of a better questionnaire containing newly formulated questions and
only the valid questions from the original K-T studies. This new questionnaire
should then be administered to Intro to Econ students at the end of this
semester [Fall 2006], and to new students at the first meeting of Intro
students at the beginning of Spring semester, 2007.
Race, Perceptions of Female Attractiveness,
and Potential Expected Income
(Matt Jarvis has expressed interest in
this thesis.)
This
thesis would extend the work of Fisk by addressing whether
the evaluations of attractiveness of females of different races or ethnicities are influenced by the race, ethnicity, gender, or other demographic
characteristics of “the judges.” Another aspect of the proposed research would
be to try to ascertain whether differentials in perceived attractiveness are
likely to be reflected in future incomes. Outline of the research: Half
of the database of Facebook pictures
of Caucasian women used by Fisk, randomly chosen, will be
replaced by pictures of non-Caucasian women randomly chosen from
students at the colleges Susan Fisk used to build her database. Student judges will also be asked to estimate how much they would offer to
pay these women if the judges were employers potentially hiring these women as
new college graduates.
The idea
that underpins parts of this proposed research derives from recently published
research results in which individual student volunteers who were paid were,
first, asked to estimate how many puzzles they could complete in a given time
period, and then these students were tested to see how many puzzles they in
fact could solve per period. A panel of paid student “judges” was asked to assign attractiveness scores to the students
who had been tested for proficiency as puzzle solvers. The results of these
surveys were that, on average, students with higher attractiveness scores
believed they could solve more puzzles than students with lower attractiveness scores, and the panel of judges supported the view that more
attractive students were more likely to solve more puzzles. However, the “real world”
test of relative proficiency as puzzle solvers revealed zero correlation
between attractiveness and the average number of puzzles solved. These results
suggest that more attractive people are more confident
that they have significant aptitude for tasks than less attractive students,
and by inference, that employers may systematically offer higher pay to
potential employees perceived as more attractive.
Trends in College Enrollments Analyzed by
Gender: The Issue of Economic Efficiency
(Anna Jones has expressed interest in
this thesis.)
This
thesis would address possible reasons for the reversal of college enrollments
by gender that has occurred during the past half century or so, and whether the
current structure of K-12 education and the criteria for the awarding of
scholarships and other incentives to acquire higher levels of education may
systematically yield economically inefficient results.
Background:
In 1960, roughly 65 percent of all college students were male. By 2005, roughly
60 percent of all college students are female. Moreover, scholarships seem increasingly
to be awarded to beginning female students relative to
the incentives offered to beginning male college students. The normative
criterion of economic efficiency suggests that the expected societal gains from
education, including expected gains to the educated individuals, should, at the margin, be equal per dollar spent for all
identifiable demographic groups. Recent
research suggests that females students tend to learn more if taught by female
teachers, and that male students learn more if taught by male students.
Moreover, there is some evidence that females mature earlier than males in
numerous dimensions, including intellectual and social aspects of
personality. In a nutshell, are male
students inefficiently disadvantaged relative to female students, if the
awarding of, e.g., Robertson or Morehead scholarships is based on criteria that
have biases in their predictive values for gains from education, lifetime
productivity, etc.?
Optimal Price Discrimination and Software
Piracy: Efficiency Issues
(Zach Clayton has expressed interest in
this thesis.)
This
thesis would address the circumstances under which software piracy results in
inefficiently low incentives for software development, the possibility that
there is an optimal level of piracy that benefits software developers and the society at large, and how effectively price structures for
software use price discrimination to maximize the long run profitability of
software developers.
Standard Economic Theory: Some Assumptions about Behavior
1.
Human beings are
self-interested, and they seek pleasure and try to avoid pain.
2.
Human beings are
rational and forward-looking.
3.
Human beings are
time consistent –choices at any moment are assumed consistent with choices
people expect to make at future points in time.
4.
Human beings tend
to be somewhat risk averse.
5.
Human beings must
optimize because choices are bounded by limited resources.
Self Interest and Rational Optimization?
Self Interest:
Naïve Egoism: U = U(x1,
x2, …, xn) … [xi = some good]
Present Aim Rationality: U = U(X, Y) [X, Y = vectors]
Production and Resources:
The creation of value (utility?)
Labor / Land / Capital / Entrepreneurship
1. Choices
are rational if we expect the
consequences to be consistent with our goals. Goals that are inconsistent with
each other may also be a symptom of irrationality.
2. Choices
are irrational if we know
them to be inconsistent with achieving our goals.
3. A
behavior is arational if, for example, it entails our tastes –
preferences between chocolate and licorice, for example, or our aesthetic
sensibilities.
The conventional economists’ view of
behavior: humans are rational and do
the best they can with what they have. Mistakes are consequences of imperfect
information and imperfect foresight.
Behavioral Economics: Standard
Economic Assumptions are Unrealistic.
We are not always good at optimizing. If people were as rational and
forward-looking as economists assume, every chess match would yield the same
result, which might be a stalemate, or perhaps white
would win, or black might win. None of us is smart enough to know with
certainty what that ending would be. Numerous paths might lead to the
monotonously predictable ending of each game, but every counter to every
possible move by an opponent would generate a countermove yielding the same
result. Chess would be no more intellectually challenging than tic-tac-toe.
How “self-interested” are
people?
Robert Frank (Cornell University – Behavioral
Economist) confronted three groups of volunteers with a prisoners’
dilemma problem.
1.
Grad students studying
economics.
2.
Undergrads
studying economics.
3.
Students
with no background in economics courses.
Results? Members of group 3 trusted each
other and cooperated most. Group 2 was more trusting and cooperative than Group
1. This raises the issue of whether people who choose to major in economics are
inherently less trusting, or does increased exposure to “the economic way of thinking”
program students to cynically view others as narrowly
self-interested?
How “rational”
are people?
Cognitive
psychologists Daniel Kahneman [Nobel Prize in Economics, 2002] and Amos Tversky
[henceforth, K-H] built a 21 item questionnaire to try to ascertain how rational
people are, and how consistent they are in dealing with risk.
K-T Results?
The results of the K-T surveys of college students at
Stanford. Princeton, Toronto, and
Tel Aviv suggested strongly that people are far less rationally consistent and
far less systematic in their treatment of risk than economists suppose.
Kahneman’s
“Peak-End Rule”
1. First derivatives of income/consumption matter more
than levels. It’s not where you are that determines
happiness, but the path that you followed to get there. People on an upward
trajectory (ex: recent promotions or increases in wealth) tend to be much
happier than people at the same level who have recently experienced some type
of reversal.
2.
People are often
poor forecasters of what will make them happy, but they do tend to know when
they are happy.
Some Definitions:
Anomaly: Anomalies are exceptions to standard expectations
about events or behavior. In economics, the term anomaly increasingly refers to
behavior not in accord with conventional economic theory. For example, people
who habitually set their clocks or watches ahead a few minutes are exhibiting
anomalous behavior, because economic analysis assumes that rational people
never intentionally try to fool themselves. Open the prospect theory link for more discussion.
Prospect theorists categorize various anomalies as problems of:
(a) framing –
people sometimes make different choices when the same problem is presented in
different ways. Example: If the grocery store cashier rings up “on sale”
microwave dinners at $2 each regardless of whether the price is stated as “$2
each” or “2/$4” and if people buy more microwave dinners in even numbered lots
(2, 4, 6, 8, …) when the price is stated as “2/$4,” then framing affects
consumer decisions.
(b) nonlinear preferences – people may make choices that seem
inconsistent with assumptions about preference functions. Consider
transitivity. If A is preferred to B and B is preferred to C, then when people
choose C over A, they are not behaving in accord with economic rationality.
(c) risk aversion and risk seeking – some
individuals will simultaneously and knowingly take unfair bets to avoid risk
(e.g., by buying life insurance) and unfair bets that increase risk
(e.g., playing slot machines at casinos).
(d) source – the
mechanism may matter even if the probable outcomes of activities are identical.
People may pay more for a good because of the way it is
packaged than they will an item that they know to be identical but
packaged differently, even if they intend to immediately discard the packaging.
(e) loss
aversion – potential losses loom greater than relatively equal
potential gains. The observed asymmetry
in these differences is far too large to be explained
solely by income effects.
Bounded rationality: Nobel-prize winner Herbert
Simon coined the term “bounded rationality” to describe limits to people’s ability to rationally solve
complex problems. Conventional economic theory assumes that every agent
possesses infinite amounts of infinitely rapid brainpower to
optimally solve every problem. Simon observed that many people fail to
understand probabilities, and that people universally lack the ability to
process information with consistency, precision, and accuracy. Consequently, Simon asserted that consumers often fail to maximize
their utility, and that business decisionmakers often fail to maximize profits,
so that behavior should be assumed “satisficing” instead of maximizing per se. Simon pointed out that instead
of the precise maximization assumed by economists, most people rely heavily on
cognitive shortcuts, also known as heuristics, e.g.,
they make lists, or develop “rules of thumb.”
Bounded self-interest: Behavioral theorists use the
phrase bounded self interest to describe the
phenomenon of altruism. Research indicates
that individuals who believe themselves treated fairly in the marketplace tend
to engage in charitable activities to benefit other people, contrary to the
view that self interest as a motive is narrowly
egoistic. (These findings hint that charitable giving is income elastic.)
Bounded willpower: Behavioral economists have
observed that individuals frequently exhibit bounded will power, in that people
sometimes knowingly act in ways in the short run that they recognize as
inconsistent with their long run self interests. For example, a person who
smokes may be very aware of the long run consequences of smoking, but
procrastinate on quitting despite knowledge of the likely long term
consequences. Procrastination and addiction (to, e.g., gambling, drugs, or
alcohol.) are among several categories of behavior that reflect problems
associated with bounded will power.
Heuristics: Heuristics are the mental shortcuts people use when evaluating decisions because they cannot perform all the mental gymnastics necessary to perfectly process information so that their decisions are mathematically optimal, given all the information that is available and known. An example of the use of a heuristic would be a shopper who “eyeballs” the contents of a grocery cart to estimate whether or not the shopper has enough cash on hand to pay for the groceries. The shopper could, instead, know the answer with certainty by summing the prices of the items as each was placed in the cart, and continuously adding the appropriate sales tax. The “eyeball” approach is a convenient heuristic.
Risk aversion: Risk aversion refers to the relative willingness of
people to choose alternatives with lower expected (probabilistic) values if
their choices reduce uncertainty. Decisionmakers who exhibit this
characteristic are termed risk averse. The extent of risk aversion is roughly measured by the expected value an economic agent
is willing to forego to gain greater certainty. Risk-averse individuals are
willing to make only bets that hold the prospect of loss (as opposed to merely
holding onto their money) only if the bets are unfairly
weighted in their favor. The assumption that people are consistently
risk aversion is a workhorse of neoclassical utility theory, but
research in the area of prospect theory increasingly calls this
assumption into question.
Risk lover: Individuals are described as
risk seeking if they are willing to choose alternatives with lower expected
(probabilistic) values in order to increase the possibilities of gain.
Risk-seeking individuals (also known as risk lovers) are willing to make a bet
that is unfairly weighted against them. A risk seeker will accept an expected
payoff of less than X in preference to the certainty of X. Suppose, for
example, that you have a dollar that is yours to keep if you don’t make a bet. If you were a risk seeker
you might buy a $1 lottery ticket on which the odds of winning a million
dollars were only 1 in 5 million.
