Past Talks s1: Spring, 2008
'Science of Liberty', Spring 2008
Jens Grosser
of Florida State University
Department of Political Science
Friday, April 25th, from 4:00 to 5:30
:
"Candidates, voters, and endogenous group formation: An experimental study"
Jens Großer
Thorsten Giertz
ABSTRACT
We experimentally study two-candidate elections with simple majority voting. Distinct groups of voters are endogenously formed by candidates' policy offers, i.e. by their distribution of a fixed budget across the voters. We allow candidates to favor some voters at the expense of others. Elections with compulsory and voluntary costly voting are distinguished. Our experimental results show that policy offers include more voters when voting is compulsory. Moreover, we observe increasing voluntary participation in the voters' benefit-differentials between both policy offers. Finally, we present evidence for the development of political bonds between voters and long-lived parties, which have the opportunity to coordinate their policy offers across legislative periods.
----------------------------------------------------------------------------------
Self-interest through agency:
An alternative rationale for the principal-agent relationship
and
The Delgation of Social Behavior
with
Roberto Weber
of Carnegie Mellon University
This Friday, April 4th, from 4:30to 6:00
Abstracts:
Self-interest through agency: An alternative rationale for the principal-agent relationship
Economic actors are usually assumed to enter into principal-agent relationships due to efficiency gains from comparative advantage. This research explores whether such relationships may also arise because they allow principals to obtain a selfish, inequitable outcome without behaving themselves in an explicitly selfish manner. An experiment is reported in which principals either decide repeatedly how much money to share with recipients or select an agent to make the decision on their behalf. As hypothesized, principals keep more and recipients receive less when the allocation decision is made through agents. Attachments1
The Delegation of Social Behavior
Economic actors frequently delegate their decision making to agents. The usual interpretation for such agency relationships is that they are motivated by gains from exchange - for instance, due to comparative advantage. In this talk I discuss two experimental papers demonstrating alternative rationales for why people might want to delegate decision making to others. In one case, we demonstrate that having others make decisions on our behalf allows us to circumvent moral constraints on our behavior, and thus behave more self-interestedly than we would ourselves. We use the well-known "dictator game" to show that acting through agents produces outcomes that enhance self-interest at the expense of fairness. People reward agents for acting self-interestedly on their behalf, even though people rarely act as self-interestedly when making decisions directly. In a second (in progress) paper, we explore the extent to which delegation of decision making allows groups to solve the free-rider problem in public goods games. We observe that groups electing agents to make decisions on the groups' behalf reach cooperation levels close to the full-contribution maximum. However, we also observe cases in which the elected agent appeals to and rewards only a minimum-winning coalition, and delegation thereby serves as a mechanism through which a slight majority can coordinate on exploiting a minority.
----------------------------------------------------------------------------------
Agent Computational Economic Modeling
an ICES Double Lecture
with
Blake LeBaron
of Brandeis University,
and the National Bureau of Economic Research
and
Robert Axtell
of GMU, the Santa Fe Institute,
and the Brookings Institution
This Friday, March 28th, from 3:00 to 6:00
with a reception to follow
We at the Interdisciplinary Center for Economic Science are proud to present two top scholars in the field of Simulated Agent Computational Economic Modeling (ACE), Blake LeBaron and GMU's own Robert Axtell. ACE is a powerful tool of increasing value in the study of economies and markets. At 3:00, Blake LeBaron will present his recent work on agent based financial markets, and how sensitive the end wealth distribution can be to the functional form of agents' utility. Then at 4:45, Robert Axtell will present his efforts towards constructing an artificial economy of one hundred million heterogeneous-agent, and explain how it may help us understand relationships between aggregate variables. To better showcase this rising methodology, the format for our regular Friday seminar has been adjusted as follows:
- 3:00 - 4:00 Le Baron will present Wealth Evolution and Distorted Financial Forecasts
- 4:00 - 4:15 Open discussion, question and answer
- 4:15 - 4:45 Coffee break (refreshments provided)
- 4:45 - 5:45 Axtell will present Macroeconomics from the Bottom Up with Very Large Scale Agent Models.
- 5:45 - 6:00 Open discussion, question and answer
- 6:00 - 7:00 Open Reception (beer, wine, all are invited!)
Abstracts:
Rob Axtell, Macroeconomics from the Bottom Up with Very Large Scale Agent Models
Macroeconomics is the study of economic phenomena in the aggregate. For example, macroeconomists typically posit specific relationships between total savings (by individuals) and aggregate investment (by businesses), or try to infer empirically the relationship between the economy-wide rate of unemployment and and the overall rate of inflation. The existence of such aggregates is not controversial, but the functional relations between aggregates is not well understood, and this has led to the search for "the microfoundations of macroeconomics." Because analytical solution of models involving large numbers of heterogeneous agents is very difficult and possibly intractable, macroeconomic theorizing as practiced today utilizes 'representative agents'--often a single consumer and firm. In this talk I will take a different approach, using a large number of heterogeneous, boundedly rational agents (consumers and firms) who interact directly with one another, potentially out of equilibrium, and whose cumulative actions give rise to an 'emergent macroeconomy'. In such models aggregates exist (
e.g., total savings, overall unemployment rate), but their evolution is determined by economic decisions and actions at the individual level, not on other aggregates. I shall describe barriers and bottlenecks to our goal of realizing a 100 million agent artificial economy 'in silico' using software agents, and speculate on the extent to which 'more is different' in economics.
Blake Lebaron, Wealth Evolution and Distorted Financial Forecasts
Evolutionary metaphors have been prominent in both economics and finance. They are often used as basic foundations for rational behavior and efficient markets. Theoretically, a mechanism which selects for rational investors actually requires many caveats, and is far from generic. This paper tests wealth based evolution in a simple, stylized agent-based financial market. The setup borrows extensively from current research in finance that considers optimal behavior with some amount of return predictability. The results confirm that with a homogeneous world of log utility investors wealth will converge onto optimal adaptive forecasting parameters. However, in the case of utility functions which differ from log, wealth selection alone converges to parameters which are economically far from the optimal forecast parameters. This serves as a strong reminder that wealth selection and utility maximization are not the same thing. Therefore, suboptimal financial forecasting strategies may be difficult to drive out of a market, and may even do quite well for some time.
----------------------------------------------------------------------------------
Do Strategic Substitutes Make Better Markets? A Comparison of Bertrand and Cournot Markets as Trading Institutions
Speaker: Professor Douglas Davis, Virginia Commonwealth University
When: Friday, March 21, 4:00-5:30pm
Where: Room 335, Truland Building, Arlington Campus
----------------------------------------------------------------------------------
Do You Know That I Am Biased? An Experiment
Speaker: Professor Sandra Ludwig, University of Munich
When: Friday, March 7, 4:00-5:30pm
Where: Room 335, Truland Building, Arlington Campus
Abstract
This experiment explores whether individuals know that other people are biased. We confirm that overestimation of abilities is a pervasive problem, but observe that most people are not aware of it, i.e. they think others are unbiased. We investigate several explanations for this result. As a first one, we discuss a possible unfamiliarity with the task and the subjects�� inability to distinguish between random mistakes and a real bias. Second, we show how the relation between a subject��s belief about others and his belief about himself might be driven by a false consensus effect or self-correction mechanism. Third, we identify a self-serving bias when comparing how a subject evaluates his own and other people��s biases.
Attachments
----------------------------------------------------------------------------------
Preemption Games: Theory and Experiment
Speaker: Professor Ryan Opera, University of California, Santa Cruz
When: Friday, Feburary 29, 4:00-5:30pm
Where: Room 335, Truland Building, Arlington Campus
Abstract
Several investors face an irreversible investment opportunity whose value V is governed by Brownian motion with upward drift and random expiration. The first investor i to seize the opportunity before expiration receives the current V less a privately known cost Ci ; the other investors receive nothing. We characterize Bayesian Nash Equilibrium (BNE) for this game, extending previously known results.
We also report a laboratory experiment with 72 subjects randomly matched into 600 triopolies. As predicted in BNE, subjects in triopolies invested at lower values than in monopolies, changes in Brownian parameters significantly altered investment values in monopoly but not in triopoly; and the lowest cost investor in a triopoly usually preempted the others. Evidence was mixed on other BNE predictions, e.g., whether higher cost brings smaller markups. Overall, subjects' earnings came rather close to the BNE prediction.
----------------------------------------------------------------------------------
Price Dynamics in an Exchange Economy
Speaker: Professor Steven Gjerstad, Purdue University
When: Friday, Feburary 15, 4:00-5:30pm
Where: Room 555, Truland Building, Arlington Campus
Abstract
The pure exchange model is the foundation of the neoclassical theory of value, yet equilibrium predictions and models of price adjustment for this model remained untested prior to the experiment reported in this paper. With the exchange economy replicated several times, prices and allocations converge sharply to the competitive equilibrium in continuous double auction (CDA) trading. Convergence is evaluated by comparing the extent of price adjustment within each market replication (or trading period) to the extent of adjustment across trading periods: most observed price adjustment occurs within trading periods, so price adjustment data are evaluated with the Hahn process model (Hahn and Negishi [1962]), which is a disequilibrium model of within-period trades. Estimation demonstrates that the model is consistent with observed price paths within each period of the exchange economy. The model is augmented with an additional assumption �C based on observations from this experiment �C that the initial trade price in period t+1 is randomly drawn from the interval between the minimum and maximum trade prices in period t. The estimated within-period adjustment rule, combined with this across-period adjustment rule, generates price paths similar to data from an experiment session.
----------------------------------------------------------------------------------
Applying Psychology to Economic Incentive Design: Using Incentive Preserving Rebates to Increase Acceptance of Critical Peak Pricing
Speaker: Dr. Robert Letzler
When: Friday, Feburary 1, 4:00-5:30pm
Where: Room 555, Truland Building, Arlington Campus
Abstract
This project extends the idea that we should address policy problems by improving incentives to add that aligning the incentives' presentation with the way people make economic decisions can help people make better choices and help achieve policy goals. It applies ideas from behavioral economics to design practical electricity pricing policies. The cost of generating power fluctuates enormously from hour to hour but most customers pay time invariant prices. The mismatch between the fluctuating generation cost and the fixed retail price creates billions of dollars in deadweight losses. Customers who participate in Critical Peak Pricing (CPP) programs use less power during high-priced periods than do customers on traditional, time invariant rates. CPP customers report high satisfaction levels and often save 10% or more. Yet, roughly 99% of customers reject opportunities to switch to critical peak pricing. The psychology literature documents heuristics that people use to decide under risk. The conventional CPP presentation leads several of these heuristics astray. For example, customers using these heuristics would put too much weight on the risks and losses involved with paying more to get less during the high priced ``events'' relative to the weight they put on their steady stream of savings. This paper departs from the hypothesis that one or more of these heuristics underlies customer resistance. Hence, I suggest Incentive Preserving Rebates that change the presentation of CPP to address these heuristics. Incentive Preserving Rebates reframe events as opportunities to get rebates rather than as periods of extremely high prices. Incentive Preserving Rebates change the presentation, but change neither marginal incentives nor each customer's total annual payments. I then explore the implications of incentive preserving rebates for customers who participated in a California pilot program.
----------------------------------------------------------------------------------
When Bioterrorism was No Big Deal
Speaker: Professor Werner Troesken
When: Friday, January 25, 4:00-5:30pm
Where: Room 555, Truland Building, Arlington Campus
Abstract
It To better understand the potential economic repercussions of a bioterrorist attack, this paper explores the effects of several catastrophic epidemics that struck American cities between 1690 and 1880. The epidemics considered here killed between 10 and 25 percent of the urban population studied. A particular emphasis is placed on smallpox and yellow fever, both of which have been identified as potential bioterrorist agents. The central findings of the paper are threefold. First, severe localized epidemics did not disrupt, in any permanent way, the population level or long-term growth trajectory of those cities. Non-localized epidemics (i.e., those that struck more than one major city) do appear to have had some negative effect on population levels and long-term growth. There is also modest evidence that ill-advised responses to epidemics on the part of government officials might have had lasting and negative effects in a few cities. Second, severe localized epidemics did not disrupt trade flows; non-localized epidemics had adverse, though fleeting, effects on trade. Third, while severe epidemics probably imposed some modest costs on local and regional economies, these costs were very small relative to the national economy.
 |
|
