In mechanism design, the optimal mechanism is usually not linear in the agent's type. However, linear mechanisms are easy for implementation. Are there any studies on the performance comparison between optimal and linear mechanisms. I guess a theoretical comparison is probably hard, as the payoff in the linear mechanism does not have closed-form, but a numerical analysis is not very difficult. Thanks.
2
-
$\begingroup$ This doesn't answer your question, but you might be interested in this literature on robustness in linear contracts, in particular, Calibrated incentive contracts by Sylvain Chassang. $\endgroup$– NickJMar 18, 2015 at 3:56
-
$\begingroup$ Thank you for the reference you suggested. This is helpful. $\endgroup$– JustinMar 19, 2015 at 15:48
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
Robustness and linear contracts by Carroll says that linear is optimal (in his class of environments). Tim Roughgarden from Stanford also studies `the price of anarchy' related to using simple but suboptimal mechanisms.
