Setting Lower Bounds on Truthfulness

We present general techniques for proving inapproximability results for several paradigmatic truthful multidimensional mechanism design problems. In particular, we demonstrate the strength of our techniques by exhibiting a lower bound of 2 − 1 m for the scheduling problem with m unrelated machines (formulated as a mechanism design problem in the seminal paper of Nisan and Ronen on Algorithmic Mechanism Design). Our lower bound applies to truthful randomized mechanisms, regardless of any computational assumptions on the running time of these mechanisms. Moreover, it holds even for the wider class of truthfulness-in-expectation mechanisms. This lower bound nearly matches the known 1.58606 randomized truthful upper bound for the case of two machines (a non-truthful FPTAS exists). Recently, Daskalakis and Weinberg [17] show that there is a polynomial-time 2approximately optimal Bayesian mechanism for makespan minimization for unrelated machines. We complement this result by showing an appropriate lower bound of 1.25 for deterministic incentive compatible Bayesian mechanisms. We then show an application of our techniques to the workload-minimization problem in networks. We prove our lower bounds for this problem in the inter-domain routing setting presented by Feigenbaum, Papadimitriou, Sami, and Shenker. Finally, we discuss several notions of non-utilitarian fairness (Max-Min fairness, Min-Max fairness, and envy minimization) and show how our techniques can be used to prove lower bounds for these notions. No lower bounds for truthful mechanisms in multidimensional probabilistic settings were previously known.1 ∗Computer Science Department, Technion, Haifa, Israel. [email protected]. †School of Computer Science and Engineering, Hebrew University, Jerusalem, Israel. [email protected]. The current paper supersedes "Setting Lower Bounds on Truthfulness" that appeared as an extended abstract in the Proceedings of the 18th Annual ACM-SIAM Symposium on Discrete Algorithms (SODA07), pages 1143-1152, 2007. The current version includes a new lower bound result for Bayesian Incentive compatible Mechanisms.