ALEXEY KUSHNIR

Research statement     Google Scholar page

Several labor markets, including the job market for new Ph.D. economists, have recently developed formal signaling mechanisms. We show that such mechanisms are harmful for some environments. While signals transmit previously unavailable information, they also facilitate information asymmetry that leads to coordination failures. In particular, we consider a two-sided matching game of incomplete information between firms and workers. Each worker has either the same "typical" known preferences with probability close to one or "atypical" idiosyncratic preferences with the complementary probability close to zero. Firms have known preferences over workers. We show that under some technical condition if at least three firms are responsive to some worker's signal, the introduction of signaling strictly decreases the expected number of matches.

Many labor markets share three stylized facts: employers cannot give full attention to all candidates, candidates are ready to provide information about their preferences for particular employers, and employers value and are prepared to act on this information. In this paper we study how a signaling mechanism, where each worker can send a signal of interest to one employer, facilitates matches in such markets. We find that introducing a signaling mechanism increases the welfare of workers and the number of matches, while the change in firm welfare is ambiguous. A signaling mechanism adds the most value for balanced markets.

We consider a standard social choice environment with linear utilities and independent, one-dimensional, private types. We prove that for any Bayesian incentive compatible mechanism there exists an equivalent dominant strategy incentive compatible mechanism that delivers the same interim expected utilities for all agents and the same ex ante expected social surplus. The short proof is based on an extension of an elegant result due to Gutmann et al. (Annals of Probability, 1991). We also show that the equivalence between Bayesian and dominant strategy implementation generally breaks down when the main assumptions underlying the social choice model are relaxed, or when the equivalence concept is strengthened to apply to interim expected allocations.

We consider social choice environments with private values and correlated types when each agent's matrix of conditional probabilities satisfies the full rank condition. We show that for any mechanism there exists a dominant strategy incentive compatible mechanism that delivers the same interim expected utilities to all agents and generates at least the same social surplus. In addition, if there is a social alternative that is inferior to the other alternatives for all agents the dominant strategy incentive compatible mechanism matches exactly the social surplus. These results extend to environments with interdependent values satisfying the single crossing condition.

We provide a unified and simple treatment of reduced-form implementation for general social choice problems and extend it to environments with value interdependencies. We employ the geometric approach developed by Goeree and Kushnir (2015) to characterize the set of feasible interim agent values (agent utilities excluding transfers) that we identify by deriving the analytical expression of its support function. As an application, we use the reduced-form implementation to analyze second-best mechanisms in environments with value interdependencies.

We extend the equivalence between Bayesian and dominant strategy implementation established by Gershkov et al. (Econometrica, 2013) to environments with non-linear utilities satisfying the property of increasing differences over distributions and the convex-valued assumption. The new equivalence result produces novel implications to the literature on the principal-agent problem with allocative externalities, environmental mechanism design, and public good provision.

For any linear transformation and two convex closed sets, we provide necessary and sufficient conditions for the transformation of the intersection of the sets to coincide with the intersection of their images. We also identify conditions for non-convex closed sets, continuous transformations, and multiple sets. We demonstrate the usefulness of our results via an application to the economics literature of mechanism design.

We provide sufficient conditions for a monotone function with a finite set of outcomes to be cyclically monotone. Using these conditions, we show that any monotone function defined on the domain of gross substitutes is cyclically monotone. The result also extends to the domain of generalized gross substitutes and complements.

We consider a model of monopolistic competition with several heterogeneous sectors and endogenous labor supply. For low (high) values of the labor supply elasticity, we show that there is always a unique equilibrium. For medium values of the labor supply elasticity, there are either zero or two equilibria.

We apply basic techniques from convex analysis to develop a simple and unified treatment of optimal mechanisms. Our approach clarifies the literature on reduced form implementation and generalizes it to general social choice settings with possibly multidimensional and correlated types. For linear one-dimensional settings, we incorporate incentive compatibility using well-known results from majorization theory and prove the equivalence of Bayesian and dominant strategy implementation. We then derive the optimal mechanism for any linear objective of agent values using standard micro-economic tools such as Hotelling's lemma. Finally, we extend our results to concave objectives of agent values and payments by providing a fixed-point equation characterizing the optimal mechanism.

We prove that supply correspondences are characterized by two properties: the law of supply and being homogeneous of degree zero.

The unequal distribution of dividends implies the unequal distribution of the profit share of workers’ product of labor. In a Mirrleesian framework when dividends cannot be expropriated, we show that a progressive distribution of dividends creates a positive dividend effect on labor income taxes. Our numerical simulations show the dividend effect to be approximately four percentage points. We analyze the dividend effect under different market structures and its interplay with other forms of taxation.

Nguyen et al. (2016) introduce the Bundled Probabilistic Serial and Bundled Probabilistic Serial under Limited Complementarities mechanisms for allocating bundles of indivisible objects under ordinal preferences over bundles. They present efficiency, fairness, and incentive properties of the two mechanisms, showing both to be ordinally efficient, weakly envy-free, weakly strategy-proof, and asymptotically strategy-proof. We provide counterexamples demonstrating that the Bundled Probabilistic Serial mechanism is not weakly strategy-proof and the Bundled Probabilistic Serial under Limited Complementarities mechanism is not weakly envy-free, weakly strategy-proof, or asymptotically strategy-proof. We also revise proofs for the remaining properties.

We consider the problem of allocating courses to students in post-secondary institutions. We propose a mechanism that assigns course seats based on student preferences and respects course priorities. This mechanism uses fake money and competitive equilibrium to allocate courses without transfers and has desirable theoretical properties in terms of stability, efficiency, fairness, and strategy-proofness. In simulations drawing from real-world university data, we demonstrate that its outcomes improve student mean utility and allocation fairness over the outcomes of several celebrated mechanisms currently used in practice.

In this paper, we revisit the multiple-good monopolist problem wherein a seller is designing the optimal mechanism for the sale of multiple goods facing a buyer with multiple dimensions of private information. Using a first-order approach to simplify the problem, we use network flow methods to identify sufficient conditions for optimality of candidate allocations. Our main result is that within a large class of mechanisms, optimality under the first-order approach is equivalent to a first-order stochastic dominance condition in the direction of trade. We apply this result to several settings and identify conditions on the distribution of types under which deterministic and random bundles are optimal, and derive the optimal mechanism for several families of distributions with correlated types.

We study the problem of minimizing a functional over the class of convex curves. Such problems often arise in economics, especially in mechanism design, where incentive-compatibility constraints reduce the designer's optimization problem to a search over convex utility functions. We characterize solutions under general assumptions, distinguishing between linear and non-linear objectives. In the nonlinear case, our key methodological contribution is to use Minty's parametrization to transform the original problem into an equivalent, more tractable one. As applications, we obtain new results on optimal mechanisms for selling multi-attribute goods under single-unit demand and for regulating a monopoly with unknown fixed and marginal costs. We also show how several other economic problems can be reformulated within our framework.

This paper analyzes the effects of grade-based priority on incentives, welfare, and equity in course allocation using theory and empirics based on the actual course allocation data from Universidad Carlos III de Madrid.