“Demographics and the Political Sustainability of Pay-as-you-go Social Security”
(with Theodore C. Bergstrom), CESifo/Delta
Conference on Strategies for Reforming Pension Schemes, refereed conference
volume
Abstract: The net present value of costs and benefits from
pay-as-you-go social security is negative for young people and positive for the
elderly. If people vote their financial self-interest, there is a pivotal age
such that those who are younger favor smaller benefits and those who are older
favor larger benefits. We estimate the expected present value of benefits and
costs to US voters of each age and sex for a small permanent increase in social
security benefits. We find that if voters vote their self-interest, a
substantial majority will favor maintaining benefit levels at least at current
levels. Over the next four decades, as the population ages, maintaining social
security will become more expensive.
Consequently, the pivotal age at which voters begin to favor social
security will rise. At the same time, the median age of voters will also rise
and remain above the pivotal age.
Throughout the period, a majority of selfish voters would favor
maintaining current benefit levels. We also investigate support for changes in
the age at which benefits begin. Self-interest on this issue does not divide
simply between old and young. Young voters and some old voters will favor a
later benefits age, while middle-aged workers will oppose. Over the next 30
years, proposals for an immediate once-and-for-all increase in the benefits age
would be defeated in majority voting. Proposals for a deferred, gradual
increase in the benefits age might under some circumstances receive majority
support.
“A Route Choice Experiment with
an Efficient Toll,” forthcoming in Networks and Spatial Economics (Special
Issue on Transport Infrastructure)
Abstract: Traffic congestion is a substantial time cost
for many urban commuters. This paper
studies the response of subjects in an experimental setting in which subjects
choose between a short direct route that becomes increasingly congested as more
people travel on it and a more indirect route that does not become
congested. More specifically, I
investigate how subjects respond to the use of a toll that theory predicts will
minimize travel time costs. The
experimental results reported in this paper show that this toll comes very
close to achieving efficient use of the travel network.
“Exercising in
Herds: Treatment Size and Status
Specific Peer Effects in a Randomized Exercise Intervention” (with Philip
Babcock)
Abstract: In
a field experiment using university students, we find that subjects who have
been incentivized to exercise increase their
recreation center usage more if they have more friends who have been incentivized, and less if they have more friends in the
control group. Controls, however, are not influenced by their peers. Findings
highlight subtle effects of randomization, indicating that the fraction treated
in an experiment of this kind has a large influence on outcomes, and that spillovers may vary greatly by treatment status. The
methodology is applicable to other settings and
quantifies spillovers that previous approaches do not detect.
“A Comparison of Individual
and Group Behavior of a Route Choice Experiment, Examining Pure and Mixed
Strategies” (with Perry Shapiro)
Abstract: Economic modeling of group behavior often assumes
symmetric mixed-strategy behavior.
Although this seems reasonable, people are in fact heterogeneous and
thus could react differently to the same outcome. It turns out that various mixed-strategy
models examining a route choice experiment do poorly at predicting individual
behavior. However, since people react to
relatively good and bad outcomes differently, this helps to explain why some
mixed strategies do poorly at predicting individual behavior, but predict
aggregate behavior remarkably well.
Abstract: Heterogeneity is important in some settings. One such instance involves congested networks
with tolls, since people trade time for money at different rates. This paper reports results from two
experiments that examine these issues.
In both experiments, subjects choose between traveling on an indirect
route that does not congest and a direct toll route that congests as more
subjects travel on it. In the first
experiment, values of time are assigned to subjects. Subjects generally sort themselves with high
value-of-time subjects choosing the toll route.
I also find that as the cost of deviating from the equilibrium
prediction increases, subjects are more likely to make choices consistent with
equilibrium. In other words,
coordination problems diminish as value-of-time heterogeneity increases. The second experiment simulates a boring
commute in which subjects must wait after the experimental rounds are
finished. Subjects can give up money for
reduced waiting time in this experiment by traveling on the toll route. In this experiment, some subjects travel the
toll route frequently, giving up some of their payout in order to reduce their
waiting time after the experiment. These
choices are likely based on individuals’ values of time, since aggregate
behavior differs by session. There is
also evidence that subjects with time constraints travel the toll route more
often than other subjects.
Abstract: Consumers using networks with capacity constraints
can experience significantly negative consequences when quantity demanded
exceeds the capacity constraint. One
such example involves the use of electricity on days with the highest demand: When quantity demanded is ever so slightly
above production capacity at the time, electricity access must be cut off to
some people or the system will fail. This
paper experimentally tests the effectiveness of two methods to lower quantity
demanded to at or below capacity when demand is high. The first method, used in some sessions, involves
taxing high-use consumers and rebating this money to low-use consumers. The second method, used in the remaining
sessions, involves allowing a single communication period to coordinate efforts
to decrease the quantity demanded. During
the communication period, subjects make non-binding agreements to conserve
electricity. The first method works
better than the second, since subjects do not always adhere to their agreements. However, the agreements work surprisingly
well.