Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Lies, Labels, and Mechanisms

Alex L. Brown, Ethan Park, Rodrigo A. Velez

Abstract

We test whether lying aversion can steer equilibrium selection in mechanism design. In a principal-worker environment, the direct mechanism admits two dominant-strategy equilibria: the designer's target and a worker-optimal outcome. We show this limitation persists for all robust mechanisms, then ask whether framing misreports as explicit lies helps. We develop a 2X2 experiment that varies direct vs. extended mechanisms with implicit vs. explicit messages. We find that framing misreporting of type as an explicit lie shifts play away from the worker-optimal outcome toward truthful reporting, raising designer payoffs with minimal efficiency loss. These findings indicate that lying aversion is an effective lever for aligning behavior with social objectives.

Lies, Labels, and Mechanisms

Abstract

We test whether lying aversion can steer equilibrium selection in mechanism design. In a principal-worker environment, the direct mechanism admits two dominant-strategy equilibria: the designer's target and a worker-optimal outcome. We show this limitation persists for all robust mechanisms, then ask whether framing misreports as explicit lies helps. We develop a 2X2 experiment that varies direct vs. extended mechanisms with implicit vs. explicit messages. We find that framing misreporting of type as an explicit lie shifts play away from the worker-optimal outcome toward truthful reporting, raising designer payoffs with minimal efficiency loss. These findings indicate that lying aversion is an effective lever for aligning behavior with social objectives.
Paper Structure (13 sections, 1 theorem, 6 equations, 3 figures, 4 tables)

This paper contains 13 sections, 1 theorem, 6 equations, 3 figures, 4 tables.

Table of Contents

  1. Introduction
  2. Lying Aversion as a Design Lever
  3. Hiring talent with adverse selection
  4. Standard robust implementation
  5. A limit of standard robust implementation
  6. Behavioral robust implementation
  7. The experiment
  8. Experimental design
  9. Experimental procedures
  10. Hypotheses
  11. Results
  12. Discussion and concluding remarks
  13. Additional Tables

Key Result

Proposition 1

Let $f:\Theta\rightarrow X$ be an scf; $\delta$ be an ex post equilibrium of the direct revelation mechanism of $f$, $(\Theta,f)$; and $g:=f\circ\delta$ the scf induced by this equilibrium. Let $(M,\varphi)$ be a mechanism for which there is an ex-post equilibrium $\sigma$ such that for each $\theta

Figures (3)

  • Figure 1: Screens for the $3\times3-$E (a, top) and $3\times3-$I (b, bottom) mechanisms. The direct mechanisms $2\times 2-$I and $2\times 2-$E used identical screens except there were no rows and columns in the table for the unanswered option, nor option for subjects to select.
  • Figure 2: (a, upper left) Rates of both workers reporting export by mechanism and player type-pair. (b, upper right) Rates of both players reporting type truthfully by mechanism and player type-pair. (c, bottom left) Histogram of the number of times (out of 10) players under the two direct mechanisms ($2\times 2-$E and $2\times 2-$I ) report being an expert type ($N=24$ and $30$, respectively). (d, bottom right) Histogram of the number of times (out of 10) players in the extended mechanisms ($3\times3-$E and $3\times3-$I ) report being an expert type ($N=24$ and $30$, respectively).
  • Figure 3: Each figure reports the estimated coefficients of two regression models. All are relative to the baseline $2\times 2-$I mechanism. The plots in black (right bracket intervals whose axis is labeled on the right) show positive effects; the plots in red (left intervals whose axis is labeled on the left in reversed values) show negative effects. (a, top) Effect of each mechanism on observance of truthful equilibrium (in black) and worker-optimal equilibrium (in red). (b, top) Effect of each mechanism on play of truthful action (in black) and deceptive action (in red) when subjects are beginner type. (c, bottom) Effect of each mechanism on mean worker combined (in red) and staffer (in black) payoffs per period. All regressions utilize cluster robust standard errors at the session level. The regressions in the top and bottom panels include dummy variables for worker-type pair (i.e., beginner-beginner, beginner-expert, expert-expert). Results in tabular form in Supplemental Appendix Table \ref{['tab:regs']}.

Theorems & Definitions (2)

  • Proposition 1: Repullo-1985-RES
  • proof : Proof of Proposition \ref{['Prop:Repullo']}