Learning%20Dynamics%20for%20Mechanism%20Design

1
Learning Dynamics for Mechanism Design
An Experimental Comparison of
Public Goods Mechanisms
Paul J. Healy California Institute of Technology
2
Overview

• Institution (mechanism) design
• Public goods
• Experiments
• Equilibrium, rationality, convergence
• (How) Can experiments improve
• institution/mechanism design?

3
Plan of the Talk

• Introduction
• The framework
• Mechanism design, existing experiments
• New experiments
• Design, data, analysis
• A (better) model of behavior in mechanisms
• Comparing the model to the data

4
A Simple Example

• Environment
• Condo owners
• Preferences
• Income, existing park
• Outcomes
• Gardening budget / Quality of the park
• Mechanism
• Proposals, votes, majority rule
• Repeated Game, Incomplete Info

5
Mechanism Design

• Implementation g??(e)?F(e)

6
The Role of Experiments

• Field e unknown gt F(e) unknown
• Experiment everything fixed/induced except ?

7
The Public Goods Environment

• n agents
• 1 private good x, 1 public good y
• Endowed with private good only (gi)
• Preferences ui(xi,y)vi(y)xi
• Linear technology (?)
• Mechanisms

8
Five Mechanisms

• Efficient gt g??(e) ? PO(e)
• Inefficient Mechanisms
• Voluntary Contribution Mech. (VCM)
• Proportional Tax Mech.
• (Outcome-) Efficient Mechanisms
• Dominant Strategy Equilibrium
• Vickrey, Clarke, Groves (VCG) (1961, 71, 73)
• Nash Equilibrium
• Groves-Ledyard (1977)
• Walker (1981)

9
The Experimental Environment

• n 5
• Four sessions of each mech.
• 50 periods (repetitions)
• Quadratic, quasilinear utility
• Preferences are private info
• Payoff 25 for 1.5 hours

• Computerized, anonymous
• Caltech undergrads
• Inexperienced subjects
• History window
• What-If Scenario Analyzer

10
What-If Scenario Analyzer

• An interactive payoff table
• Subjects understand how strategies ? outcomes
• Used extensively by all subjects

11
Environment Parameters

• Loosely based on Chen Plott 96
• ? 100
• Pareto optimum yo (?bi - ?)/(?2ai)4.8095

ai bi ??i
Player 1 1 34 260
Player 2 8 116 140
Player 3 2 40 260
Player 4 6 68 250
Player 5 4 44 290
12
Voluntary Contribution Mechanism
Mi 0,6 y(m) ?imi
ti(m) ?mi

• Previous experiments
• All players have dominant strategy m 0
• Contributions decline in time
• Current experiment
• Players 1, 3, 4, 5 have dom. strat. m 0
• Player 2s best response m2 1 - ?i?2mi
• Nash equilibrium (0,1,0,0,0)

13
VCM Results
Nash Equilibrium (0,1,0,0,0)
Dominant Strategies
Player 2
14
Proportional Tax Mechanism
Mi 0,6 y(m) ?imi ti(m)(?/n)y(m)

• No previous experiments (?)
• Foundation of many efficient mechanisms
• Current experiment
• No dominant strategies
• Best response mi yi ? ?k?i mk
• (y1,,y5) (7, 6, 5, 4, 3)
• Nash equilibrium (6,0,0,0,0)

15
Prop. Tax Results
Player 1
Player 2
16
Groves-Ledyard Mechanism

• Theory
• Pareto optimal equilibrium, not Lindahl
• Supermodular if ?/n gt 2ai for every i
• Previous experiments
• Chen Plott 96 higher?? gt converges better
• Current experiment
• ? 100 gt Supermodular
• Nash equilibrium (1.00, 1.15, 0.97, 0.86, 0.82)

17
Groves-Ledyard Results
18
Walkers Mechanism

• Theory
• Implements Lindahl Allocations
• Individually rational (nice!)
• Previous experiments
• Chen Tang 98 unstable
• Current experiment
• Nash equilibrium (12.28, -1.44, -6.78, -2.2,
  2.94)

19
Walker Mechanism Results
NE (12.28, -1.44, -6.78, -2.2, 2.94)
20
VCG Mechanism Theory

• Truth-telling is a dominant strategy
• Pareto optimal public good level
• Not budget balanced
• Not always individually rational

21
VCG Mechanism Best Responses

• Truth-telling ( ) is a weak dominant
  strategy
• There is always a continuum of best responses

22
VCG Mechanism Previous Experiments

• Attiyeh, Franciosi Isaac 00
• Binary public good weak dominant strategy
• Value revelation around 15, no convergence
• Cason, Saijo, Sjostrom Yamato 03
• Binary public good
• 50 revelation
• Many play non-dominant Nash equilibria
• Continuous public good with single-peaked
  preferences
• 81 revelation
• Subjects play the unique equilibrium

23
VCG Experiment Results

• Demand revelation 50 60
• NEVER observe the dominant strategy equilibrium
• 10/20 subjects fully reveal in 9/10 final periods
• Fully reveal both parameters
• 6/20 subjects fully reveal lt 10 of time
• Outcomes very close to Pareto optimal
• Announcements may be near non-revealing best
  responses

24
Summary of Experimental Results

• VCM convergence to dominant strategies
• Prop Tax non-equil., but near best response
• Groves-Ledyard convergence to stable equil.
• Walker no convergence to unstable equilibrium
• VCG low revelation, but high efficiency
• Goal A simple model of behavior to
  explain/predict which mechanisms converge to
  equilibrium
• Observation Results are qualitatively similar to
  best response predictions

25
A Class of Best Response Models

• A general best response framework
• Predictions map histories into strategies
• Agents best respond to their predictions
• A k-period best response model
• Pure strategies only
• Convex strategy space
• Rational behavior, inconsistent predictions

26
Testable Predictions of the k-Period Model

• No strictly dominated strategies after period k
• Same strategy k1 times gt Nash equilibrium
• U.H.C. Convergence to m gt m is a N.E.
• 3.1. Asymptotically stable points are N.E.
• Not always stable
• 4.1. Global stability in supermodular games
• 4.2. Global stability in games with
  dominant diagonal
• Note Stability properties are not monotonic
  in k

27
Choosing the best k

• Which k minimizes??t mtobs ? mtpred ?
• k5 is the best fit

28
(No Transcript)
29
(No Transcript)
30
(No Transcript)
31
5-Period Best Response vs. Equilibrium Walker
32
5-Period Best Response vs. Equilibrium
Groves-Ledyard
33
5-Period Best Response vs. Equilibrium VCM
34
5-Period Best Response vs. Equilibrium PropTax
35
Statistical Tests 5-B.R. vs. Equilibrium

• Null Hypothesis
• Non-stationarity gt period-by-period tests
• Non-normality of errors gt non-parametric tests
• Permutation test with 2,000 sample permutations
• Problem If then the test
  has little power
• Solution
• Estimate test power as a function of
• Perform the test on the data only where power is
  sufficiently large.

36
Simulated Test Power
0.95
0.95
0.94
0.93
0.92
Frequency of Rejecting H0 (Power)
Prob. H0 False Given Reject H0
0.91
0.89
0.86
0.8
0.67
0
(W1-W2)/a
37
(No Transcript)
38
(No Transcript)
39
(No Transcript)
40
(No Transcript)
41
5-period B.R. vs. Nash Equilibrium

• Voluntary Contribution (strict dom. strats)
• Groves-Ledyard (stable Nash equil)
• Walker (unstable Nash equil) 73/81 tests reject
  H0
• No apparent pattern of results across time
• Proportional Tax 16/19 tests reject H0
• 5-period model beats any static prediction

42
Best Response in the VCG Mechanism

• Convert data to polar coordinates

43
Best Response in the cVCG Mechanism

• Origin Truth-telling dominant strategy
• 0-degree Line Best response to 5-period average

44
(No Transcript)
45
The Testable Predictions

• Weakly dominated e-Nash equilibria are observed
  (67)
• The dominant strategy equilibrium is not (0)
• Convergence to strict dominant strategies
• 2,3. 6 repetitions of a strategy implies
  e-equilibrium (75)
• Convergence with supermodularity dom. diagonal
  (G-L)

46
Conclusions

• Experiments reveal the importance of
• dynamics stability
• Dynamic models outperform static models
• New directions for theoretical work
• Applications for real world implementation
• Open questions
• Stable mechanisms implementing Lindahl
• Efficiency/equilibrium tension in VCG
• Effect of the What-If Scenario Analyzer
• Better learning models

47
An Almost-Trivial Game

• Cycling (including equilibrium!) for k3
• Global convergence for k1,2,4,5,

48
Efficiency
Efficiency Confidence Intervals - All 50 Periods
1
Efficiency
No Pub Good
0.5
Walker VC PT
GL VCG
Mechanism
49
(No Transcript)
50
(No Transcript)
51

• Voluntary Contribution Mechanism
• Results