Senior Honors Theses in Economics: Past and Future
Behavioral Economics
Seven of
the theses for which Ralph Byrns served as advisor in the past five 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), which some cognitive psychologists call expected utility
theory (EUT).
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 – see his Nudge,
with Cass Sunstein), Robert Shiller (Irrational
Exuberance and Subprime Solution), Steve Levitt
(Freakonomics), and Dan Ariely (Predictably
Irrational) also have a distinctly behavioral flavor. See also numerous
recent Freakonomics articles
written by Levitt and Dubner for the New York
Times Magazine.
Recent UNC Theses Focused on Behavioral Economics
|
Risk, Rationality, and
Familiarity with Basic Economic Theory Rebecca Martin, UNC BA with Highest Honors, 2008 |
Data |
This thesis
extended 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 examined how students performed 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. The
K-T questionnaire was administered to roughly 350 beginning students at the
start of Spring 2008, and at the end of the fall semester 2007. 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 have benefitted
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 was administered to Intro to Econ students at
the end of this semester [Fall 2007], and to new students at the first meeting
of Intro students at the beginning of Spring semester, 2008.
Race, Perceptions of Female
Attractiveness, and Potential Expected Income
(Sarah Simon, UNC BA with Honors, 2008.)
This
thesis extended 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 was to try to ascertain
whether differentials in perceived attractiveness are likely to be reflected in
future incomes. A final question addressed in this research is whether American
students differ from foreign students (in, e.g., Hong Kong) in their reactions
about how attractiveness might be expected to affect potential income. Outline
of the research: The database of Facebook
pictures of Caucasian women used by Fisk, randomly chosen, was replaced, with
half being pictures of non-Caucasian women randomly chosen, and half, pictures
of Caucasian women. Student judges from the University of North Carolina and
for Hong Kong were 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 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.
|
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.
|
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).
|
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
affects 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.
Possible Theses Topics for 2009
The Effects of Exposure to
Basic Economic Theory on Student Opinions about Political and Economic Issues
This
thesis would use surveys to ascertain the opinions of students on 37 questions originally
administered for the Washington Post to roughly 1200 PhD economists and a
similar number of members of the public at large. Some students will be asked these questions when
they initially enroll in introductory economics, in January 2009, but before
they have been exposed to economic theory. Other students will be asked the
same questions in December 2008, as they are completing their first economics
course. This research is intended to ascertain whether exposure to economics
courses has a discernible effect on student opinions. Do student perceptions about taxes, international
trade, immigration, the state of the economy, the importance of financial
markets, etc., change with exposure to economic theory?
Trends in College Enrollments
Analyzed by Gender: The Issue of Economic Efficiency
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 2007, 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.?
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 standard 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.” Some conventional
economists now recognize that calculation can absorb resources, much as the
acquisition of information does, and refer to the problem of computational
complexity.
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. Decision-makers 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.
