November 11: Daniel Rehsmann (QSMS Seminar)

Daniel Rehsmann (University of Vienna) will present “Choose Your Auction: Mechanism Design for a Bidder” (with Dmitriy Knyazev)” on November 11th, 2024, at 10:30 AM, in room QA406.

Abstract:   

This study discusses the maximization of a bidder’s utility in auctions by leveraging the information about a bidder’s value to formulate the auction’s rules. To make the analysis interesting, the research focuses on auction formats perceived as fair and unbiased, in line with common EU or WTO procurement regulations. In our main setup, we do not allow the auction to pay the bidders and characterize a preferable auction format as a second-price auction with pooling regions. The analysis then extends to include transfers towards bidders and demonstrates how a substantial interim utility can be guaranteed to a bidder without running a deficit in equilibrium. The theory is applied to a model of favoritism, discussing whether forms of preferential treatment in auctions are preventable or detectable.

October 7: Márton Benedek (QSMS Seminar)

Márton Benedek (KRTK, Corvinus University) will present “Computing the nucleolus of cooperative games” on October 7th, 2024, at 10:30 AM, in room QA406.

Abstract:   

The nucleolus offers a desirable payoff-sharing solution in cooperative games, thanks to its attractive properties—it always exists and lies in the core (if the core is nonempty), and it is unique. The nucleolus is considered as the most ‘stable’ solution in the sense that it lexicographically minimizes the dissatisfactions among all coalitions. In this talk we are focusing on the different computation methods of the nucleolus: the traditionally most widely employed scheme using a sequence of linear programs (LPs), as well as less traditional methods, such as single-LP and nonlinear approaches, and the state-of-the-art lexicographical descent method. We will also focus on the role of characterization sets, and briefly discuss some of the recent applications.

October 4: Toygar T. Kerman  (QSMS Seminar)

Toygar Kerman (Corvinus University) will present “Nested Self-Selectivity” (with Semih Koray, Bilkent University) on October 4th, 2024, at 10:30 AM, in room QA406.

Abstract:   

A society that will make a choice from a set of alternatives might also need to choose the choice rule that will be employed to make the choice. In such a situation, the notion of self-selectivity requires a social choice function (SCF) to choose itself among a set of rival SCFs. However, verifying self-selectivity might be difficult as the set of rival SCFs may be very large. We show that self- selectivity of an SCF (that satisfies independence of irrelevant alternatives) relative to some very large sets of SCFs can be verified by checking self-selectivity relative to a much smaller set of SCFs that is nested in the large set. Moreover, we show that if an SCF is self-selective relative to two different sets of rival functions, then it is self-selective relative to the intersection and union of these sets.

June 12: Alexander Usvitskiy (QSMS Seminar)

Alexander Usvitskiy (HSE University) will present the paper “Support Networks in Contests ” (with Anastasia Antsygina, Mariya Teteryatnikova, James Tremewan) on June 12th, 2024, at 10:30 AM, in room QA406.

Abstract:   

Many real-life competitive environments allow for a third party to be indirectly involved in the competition through supporting one or both conflicting parties. Such support can come from trade partners, colleagues, or allies, who can in turn benefit from the supported party’s success. In this paper we model such environments by introducing a two-stage game, in which two competing players have an opportunity to form pairwise support links with the third player before proceeding to the contest stage. We analyze under what conditions agents have incentives to form support links in view of the future conflict, what networks (if any) can be pairwise stable, and how the structure of support network affects players’ individual and total efforts in the contest. Our main result is that stable network connectivity is nonmonotone in the supporter’s benefit from the supported party’s success. Specifically, we find that as the benefit increases, network connectivity rises as linking becomes more attractive to the supporter, but then it falls as linking becomes less attractive to the competing players – the comparative statics which we test with a laboratory experiment.

June 10: Boris Knapp (QSMS Seminar)

Boris Knapp (Corvinus University of Budapest) will present the paper “Price Discrimination via Waiting Lists” on June 10th, 2024, at 10:30 AM, in room QA406.

Abstract:   

Several markets, in particular some for luxury products, exhibit long periods of excess demand. In the short run, prices below the market clearing price can be explained by demand uncertainty or sticky prices, for example. In the long run, however, they are much more challenging to rationalize. This paper shows that a monopoly can use a covert tying practice that takes the form of a priority list to increase its profits beyond the standard monopoly profit. The market outcome is characterized by a (relatively) low price of the luxury good and a priority list that grants only some consumers the right to purchase it.

June 5: Leonardo Madio (QSMS Seminar)

Leonardo Madio (University of Padova) will present the paper “Design and Governance of Quality on a Digital Platform” on June 5th, 2024, at 3:00 PM, in room QA406.

Abstract:   

This paper studies the incentives of a platform to curate quality of third- party products it hosts. There are two types of sellers. “Normal sellers” have heterogeneous intrinsic quality and can enhance their quality through effort. Bad sellers have zero quality and never exert effort. Quality is unobserved to consumers who rely on a rating system to form beliefs on product quality. We show that the platform may intentionally host bad sellers to induce more effort by normal sellers, even if it can identify and block bad sellers at zero cost. We also show that a social planner might also want to have some bad sellers on board. Moreover, tolerating bad sellers can also be a more effective tool to induce more effort than making the rating system more stringent.

May 22: Kevin Techer (QSMS Seminar)

Kevin Techer (University Jean Monnet Saint-Etienne) will present the paper “Hazardous waste transportation: a cost allocation analysis” on May 22nd, 2024, at 10:30 AM, in room QA406.

Abstract:   

This paper studies hazardous waste transportation problems. Due to their dangerous nature, the transportation of such waste implies a risk of incident having irreversible consequences on the environment. This problem has lead to a body of legal statutes that monitor the generation, storage and transportation of hazardous waste. Assuming that the transport of hazardous waste is done in a cooperative manner on a transport network, this paper investigates how to share the cost of maintaining such network among the involved agents. We analyze the hazardous transportation problem from the viewpoint of axiomatic analysis. We consider several axioms that are interpreted through different environmental law principles and provide a characterization of a new allocation rule: the responsibility rule. Then we show that the responsibility rule coincides with the multi-choice Shapley value of an appropriate multi-choice game.

May 8: Artem Razumovskii (QSMS Seminar)

Artem Razumovskii (CERGE-EI) will present the paper “Interim Deadline for Procrastinators” on May 8th, 2024, at 10:30 AM, in room QA406.

Abstract:   

People are partially time inconsistent and many have difficulties committing to a detailed schedule for a project. I study optimal interim deadlines and how they affect the behavior and resulting welfare of the present-biased agent. I consider a model in which there are three types of agent in terms of how the agent understands her present bias: naive, sophisticated, and partially-sophisticated. For each type, there is a unique design for an exogenous interim deadline that maximizes the agent’s welfare. However, only the sophisticated agent would self-impose an optimal interim deadline, while the naïve agent would not apply a self-imposed deadline at all. The partially-sophisticated agent sets a nonoptimal self-imposed deadline and can even decrease her own welfare by imposing it. The main result is that the partially-sophisticated agent who is relatively less present-biased would decrease her own welfare by using a self-imposed deadline, and the partially-sophisticated agent who is relatively more present-biased would increase her welfare given the same degree of sophistication.

May 13: Jinglei Huang (QSMS Seminar)

Jinglei Huang (Tsinghua University) will present the paper “The Shapley Value and the Nucleolus of a Two-Sided Platform Game ” on May 13th at 10:30 AM online.

MS Teams (click here to join) (Meeting ID: 389 673 827 294 Passcode: pezarH)

Abstract:   

The paper introduces a new coalitional game with transferable utility — called a two-sided platform game. The participation of the platform user on one side benefits the other side, and the platform can be established if and only if there is more than one entrepreneur. Well-known point solutions and set solutions are investigated. It turns out that the kernel and the nucleolus coincide, and both the Shapley value and the nucleolus have simple expressions. The paper sheds light on platforms and antitrust issues. When there is more than one platform entrepreneur, the utility share of each entrepreneur is relatively low in both point solutions.

March 25: Roland Molontay (QSMS Seminar)

Roland Molontay (BME, department of Stochastics) will introduce the HSDSlab: Network science and machine learning in empirical social sciences on March 25th at 16:00 PM, room QA323.

Abstract:   

In this talk, I will review some recent works of the Human and Social Data Science Lab (HSDSLab). HSDSLab is a young research group based in the Institute of Mathematics at the Budapest University of Technology and Economics. HSDSLab conducts both methodology-oriented basic research in data and network science and applied research with a human-centered and societal focus. The talk will revolve around three main topics: (1) Educational data science. (2) Social media analysis. (3) Science of science. I will sketch some of our data-driven research projects from the educational domain, including identifying students at risk of dropping out using explainable artificial intelligence, assessing the predictive validity of the admission system, and quantifying the relationship between student evaluation of teaching and grade inflation. Next, I will present our findings about the virality of social media posts, especially regarding the local and global context of image-with-text memes. Moreover, I will also briefly mention our recent project on a novel measure to evaluate the impact of scientific journals.