Abstract
We consider two natural notions of strategyproofness in random object-assignment mechanisms based on ordinal preferences. The two notions are stronger than weak strategyproofness but weaker than strategyproofness. We demonstrate that the two notions are equivalent, provide a geometric characterization of the new intermediate property which we call convex strategyproofness, and then show that the (generalized) probabilistic serial mechanism is, in fact, convexly strategyproof. We finish by showing that the property of weak envy-freeness of the random serial dictatorship can be strengthened in an analogous manner.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
First-order stochastic dominance here is defined with respect to the agent’s true preferences.
See also Lubin and Parkes (2012) for a survey of other ways of relaxing strategyproofness.
We restrict our attention to strict preference orders. The main reason for that is that the version of the (generalized) PS mechanism which allows for indifferences (Katta and Sethuraman 2006; Budish et al. 2013) fails to even be weakly strategyproof. However, the strictness assumption is not crucial. A result essentially identical to our Proposition 1 can be derived for non-strict preference orders, as well.
We allow there to be more than one copy of each object.
That is to say, whenever \(x\succsim y\) for some \(x,y\in \mathbb {R}^n\) then \(\alpha x+z\succsim \alpha y+z\) for all \(z\in \mathbb {R}^n\) and \(\alpha \ge 0\).
The reason we use the Polyhedral Separating Hyperplane Theorem, as opposed to a weaker separating result, is that it gives us the strict inequality in this sentence. In the next paragraph, it becomes apparent that the strict inequality is crucial in showing that \(\overline{u}\) is compatible with the preference order.
Budish et al. (2013) generalize the mechanism by adding group-specific quotas and show that it remains weakly strategyproof. The results in this paper hold for their Generalized Probabilistic Serial mechanism as well.
Budish et al. (2013) use a similar arguments so the following Lemma would hold in their setting as well.
Note that Example 1 can be used to show that convex envy-freeness is strictly stronger than weak envy-freeness. Namely, let \(g(a),g(b^1),\) and \(g(b^2)\) instead refer to the probability-share distributions of three agents dividing three objects among themselves (note that the probability shares for each object sum up to 1), and let their preferences be identical: they all prefer objects with smaller indices over objects with larger indices. Then that allocation would satisfy weak envy-freeness. However, it would not satisfy convex envy-freeness since the agent corresponding to g(a) would envy a convex combination of the other two agents’ allocations as shown in the example.
References
Abdulkadiroğlu A, Sönmez T (1998) Random serial dictatorship and the core from random endowments in house allocation problems. Econometrica 66(3):689–701
Aliprantis C, Border K (2006) Infinite dimensional analysis: a Hitchhiker’s guide. Springer, Berlin Heidelberg
Bogomolnaia A, Moulin H (2001) A new solution to the random assignment problem. J Econ Theory 100(2):295–328
Budish E, Che YK, Kojima F, Milgrom P (2013) Designing random allocation mechanisms: theory and application. Am Econ Rev 103(2):585–623
Hadar J, Russell W (1969) Rules for ordering uncertain prospects. Am Econ Rev 59(1):25–34
Katta AK, Sethuraman J (2006) A solution to the random assignment problem on the full preference domain. J Econ Theory 131(1):231–250
Kojima F, Manea M (2010) Incentives in the probabilistic serial mechanism. J Econ Theory 145(1):106–123
Lubin B, Parkes D (2012) Approximate strategyproofness. Curr Sci India 103(9):1021–1032
McLennan A (2002) Ordinal efficiency and the polyhedral separating hyperplane theorem. J Econ Theory 105(2):435–449
Mennle T, Seuken S (2013) Partially strategyproof mechanisms for the assignment problem Mimeo. University of Zurich, Zurich
Acknowledgments
I am grateful to Timo Mennle, Hervé Moulin, the anonymous referees of this paper, and, in particular, to Haluk Ergin for helpful comments and suggestions.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Balbuzanov, I. Convex strategyproofness with an application to the probabilistic serial mechanism. Soc Choice Welf 46, 511–520 (2016). https://doi.org/10.1007/s00355-015-0926-z
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00355-015-0926-z