1 / 20

Outline

Outline. In-class experiment on IPV First-Price Auctions Data from Cox, Robertson, and Smith (1982) Glenn Harrison’s (1989) Critique Responses by Kagel and Roth (1992) and Merlo and Schotter (1992) Key Lessons. First-Price Auctions. N bidders, individual values are i.i.d. draws from

levana
Download Presentation

Outline

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Outline • In-class experiment on IPV First-Price Auctions • Data from Cox, Robertson, and Smith (1982) • Glenn Harrison’s (1989) Critique • Responses by Kagel and Roth (1992) and Merlo and Schotter (1992) • Key Lessons Experimental Economics

  2. First-Price Auctions • N bidders, individual values are i.i.d. draws from • Values are denoted by • Subject bids are • Subjects are risk-neutral Experimental Economics

  3. Game Theoretic Predictions • Risk-neutral Nash equilibrium (RNNE) • The winner is the person who has the highest xi (efficient allocation) • Mean sales price and variance are: Experimental Economics

  4. Example: (N=3, v=0.1, v = 4.90) • If 0.5, 2.3, 3.5 were drawn, then optimal bids would have been: • The winner is the person whose value is 3.5 • Mean sales price and variance are: Experimental Economics

  5. Cox, Robertson, and Smith (CSW): Theoretical and Observed Sales Prices Experimental Economics

  6. Empirical Regularities • Subject bids are consistently higher than the risk-neutral Nash equilibrium (RNNE) • The data are consistent with game theoretic predictions if subjects are risk-averse and each has a different constant relative risk aversion (CRRA) Experimental Economics

  7. Constant Relative Risk Aversion • Arrow-Pratt’s Relative Risk Aversion • Power Utility Function: U(y) = y r Experimental Economics

  8. Two Equations • Bid Function: • Power Utility Function: Experimental Economics

  9. Foregone Income • Foregone Income = Income from Predicted Bid – Income from Actual Bid • Metric 1: • Assume other bidders are risk neutral and use equation (1) (E(r) = 1.0) • Bidders are risk neutral and use utility function (2) to optimize (r=1.0) • Metric 2: • Assume other bidders are risk-averse and use equation (1) (E(r) = 0.7) • Bidders are risk neutral and use utility function (2) to optimize (r=1.0) Experimental Economics

  10. Foregone Income: An Example(Cox, Robertson and Smith’s Experiment) Experimental Economics

  11. Foregone Income: Metric 1Harrison’s Experiment Experimental Economics

  12. Foregone Income: Metric 2Harrison’s Experiment Experimental Economics

  13. Experimental Design • Three Treatment Variables • Experience (Played once before versus none) • RNNE robots versus human subjects • Points versus dollars • Dependent variables (Bid deviation and foregone expected income) • Missing cells Experimental Economics

  14. Issues of Debate • Dependent variable: “message” versus “payoff” ? • Is constant relative risk aversion (CRRA) the “right” theory for explaining over-bidding in independent first-price auctions? Experimental Economics

  15. Responses ? Experimental Economics

  16. Responses • Experimental tests on “low-cost deviation” conjecture • Responses are not random (over-bidding) • Raise the costs of deviation: Increase the conversion rate (CSW) (no effect when conversion rate is increased by a factor of 3) • Other predictions of “low-cost deviation” conjecture • Increase the range should reduce “over-bidding” (Table 2 from KR) • Merlo and Schotter: Shape of the payoff function cannot have any effect on subject behavior unless they are able to perceive it either deductively before experiment or learn during experiment • Theorists (deductively either rightly or wrongly) Choose what they predict is the optimal choice and persist in that choice never learn about the actual payoff function Harrison’s criticism would have no force • Experimentalists (learn) subjects won an average of 4.1 times and there are simply no enough data for them to detect the flatness of the payoff function Harrison’s criticism would hold little force (Table 1 in MS) Experimental Economics

  17. Responses • Experimental tests on “risk-aversion” theory • Can the same theory apply to other IPV auctions? • Second-price auctions: Dominant strategy to bid their value irrespective of risk attitudes (subjects consistently bid above their values by a small amount) • Multiple-unit discriminative auctions: Bids are significantly less than RNNE • What other predictions does risk-aversion make? • Profit earned as a % of predicted RNNE profit should decrease with increases in N (Table 4 from KR). Experimental Economics

  18. Camerer’s Review • In the kinds of tasks economists are most interested in, the overwhelming finding is that increased incentives do not change average behavior substantially (although the variance of responses often decrease) • There is no replicated study in which a theory of rational choice was rejected at low stakes in favor of a well-specified behavioral alternative, and accepted at high stakes. Experimental Economics

  19. Lessons • The power of replication (to verify a research finding) • Only robust research findings will survive • The power of control (i.e., super easy to test competing hypotheses) • Shift the focus of debate onto data • Knowledge accumulates based on experimental data not arm-chair theorizing • The boundaries of a theory (should a behavioral theory of first-price auctions generalize to second-price auctions?) Experimental Economics

  20. A Question Is Glenn Harrison’s article bad for experimental economics? Experimental Economics

More Related