Working Paper: NBER ID: w30349
Authors: Marcel Fafchamps; Bereket Kebede
Abstract: We conduct a novel lab experiment in which pairs of subjects make separable decisions about allocative efficiency and equity in different agency structures. In terms of equity, subjects appropriate all surplus when they can, and share equally when they have to negotiate. They achieve high efficiency in general, albeit less so when the allocation of surplus is negotiated and negotiations fail. Allocative efficiency is reduced by input and output endowment effects, particularly in negotiated allocation games where subjects seek to create a sense of entitlement over joint surplus so as to achieve a more equitable income distribution. We find few differences across gender or gender pairings. Subjects are then given a choice between negotiating, paying to decide alone, or be paid to let their assigned partner decide. We find that demand for agency or delegation is sensitive to the price of agency, irrespective of gender. But female subjects are more likely to delegate to their partner if it is a male. We also find that a large fraction of both male and female subjects show a preference for negotiating that appears intrinsically motivated.
Keywords: No keywords provided
JEL Codes: D03; L23
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Agency structures (L85) | allocative efficiency (D61) |
| Agency structures (L85) | equity (D63) |
| Gender (J16) | preferences for negotiation (C78) |
| Gender (J16) | preferences for delegation (D72) |
| Negotiation (C78) | equitable distribution of income (D31) |
| Negotiation fails (C78) | efficiency losses (D61) |
| Input and output endowment effects (D11) | decision-making distortion (D91) |
| Female subjects (J16) | delegation to male partners (J12) |
| Gender differences in performance (D29) | null result (Y70) |