Trade-off between binary and richer bidding schemes in karma mechanisms

Investigate the trade-off between the behavioral simplicity of a binary (or otherwise limited) bidding scheme and the theoretical efficiency benefits of richer (full-range) bidding schemes within the karma mechanism, specifically in settings where urgency has more than two levels. Characterize how simplicity impacts human adoption and performance relative to the efficiency loss predicted for binary bidding when urgency exceeds two levels, and quantify this trade-off empirically and theoretically.

Background

The paper experimentally compares a binary bidding scheme (bid 0 or half of current karma) to a full-range bidding scheme in a karma mechanism for repeated resource allocation. While the experiments (with two urgency levels) found no significant differences in realized efficiency gains between the binary and full-range schemes, theory predicts that binary bidding becomes less efficient when there are more than two urgency levels.

In discussing implementation considerations and broader applicability, the authors explicitly state that it remains an open question to test the trade-off between the behavioral simplicity of limited bidding actions and the theoretical efficiency advantages of richer bidding schemes. This highlights a need for future work to evaluate how simplicity aids adoption and decision quality versus the efficiency losses expected when urgency is more granular.

References

Theoretically, Nash equilibrium under binary bidding will lose in efficiency with more than two urgency levels, and it remains an open question to test what the trade-off is between behavioral simplicity of a binary (or otherwise limited) bidding scheme and the theoretical benefits that come with richer schemes.

Dynamic Resource Allocation with Karma: An Experimental Study (2404.02687 - Elokda et al., 3 Apr 2024) in Section 4 (Discussion)