Working Paper: CEPR ID: DP12479
Authors: Charlie Bull; Pascal Courty; Maurice Doyon; Daniel Rondeau
Abstract: Substantial efforts have been devoted to understanding deviations from optimal behavior in games. Cason and Plott (2014, hereafter CP) propose that sub-optimal behavior may be explained by game form misconception (GFM), a failure of game form recognition, rather thanby non-standard preferences or framing effects. Following CP's application of the GFM theory to the Becker-DeGroot-Marschak mechanism (Becker et al., 1964, hereafter BDM), this paper explores whether GFM can robustly explain bidding mistakes by inexperienced subjects. We derive two new tests of the GFM hypothesis based on comparing subject behavior in the misconceived task (BDM) and on the task it is misconceived for (a first price auction). While we do replicate Cason and Plot's original results, our additional tests are inconsistent with a first price misconception explaining observed deviations from optimal bidding in the BDM. At a minimum, additional forms of misconception are necessary to explain observed bidding behavior.
Keywords: game form recognition; game form misconception; mistake; Becker-DeGroot-Marschak; preference elicitation
JEL Codes: C8; C9
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| inexperienced subjects (C90) | systematic deviations from optimal bidding behavior (D44) |
| misconceptions about the game form (C73) | systematic deviations from optimal bidding behavior (D44) |
| BDM mechanism (C69) | systematic deviations from optimal bidding behavior (D44) |
| confusion with first-price auction (D44) | systematic deviations from optimal bidding behavior (D44) |
| inability to distinguish between BDM and first-price auction (D44) | no significant difference in bidding strategies (D44) |
| increase in competition (two random posted prices) (D41) | no increase in aggressive bidding behavior (D44) |