2019 CMSS Research Seminars by Patrick Girard

Speaker: Patrick Girard (University of Auckland)

Title: “Mini-Series on Paradoxes”

Date, Time and Venue: Monday, [to start with] 11 and 25 March 2019, 14:00-15:00, 260-6115 [Business School Building, Level 6]

Abstract: Patrick will offer 2-3 seminars on paradoxes. He will start with historical/conceptual paradoxes (about God, being bald, and there being no change). He will then talk about modern logical paradoxes that involve truth, set membership and conditionals. He will end with paradoxes involving probability. The journey will take you from a conceptual/historical understanding of paradoxes in philosophy, go via problems in mathematics and logic, and end with more specific paradoxes behind the kind of mathematics that CMSS members are using in social choice theory and the like.

Everyone welcome!

2018 CMSS Research Seminar by Addison Pan

Speaker: Addison Pan (University of Auckland)

Title: “A Note on The Pivotality Condition”

Date, Time and Venue: Wednesday, 23 May 2018, 14:00-15:00, 260-319 [Business School Building, Level 3]

Abstract: This note provides simple derivations of the equilibrium conditions for different voting games with incomplete information. In the standard voting game à la Austen-Smith and Banks (1996), voters update their beliefs only based on the probability that they are pivotal. However, in voting games such as those in Ellis (2016) and Fabrizi and Pan (2017), given a closed and convex set of priors, the ambiguous averse voters select the prior from this set in a strategy-contingent manner. Therefore, it is shown in this paper that in ambiguous voting games the conditional probability of being pivotal alone is not sufficient to determine each voter’s best response.

Everyone welcome!

2018 CMSS Research Seminar by Arkadii Slinko

Speaker: Arkadii Slinko (University of Auckland)

Title: “What Do Multiwinner Voting Rules Do? Some Simulations Over the Two-Dimensional Euclidean Domain”, joint work with Edith Elkind, Piotr Faliszewski, Jean-Francois Laslier, Piotr Skowron and Nimrod Talmon

Date, Time and Venue: Wednesday, 9 May 2018, 14:00-15:00, 260-319 [Business School Building, Level 3]

Abstract: We visualize aggregate outputs of several multiwinner voting rules—SNTV, STV, Bloc, k-Borda, Monroe, Chamberlin–Courant, and PAV—for elections generated according to the two-dimensional Euclidean model. We consider three applications of multiwinner voting, namely, parliamentary elections, portfolio/movie selection, and shortlisting, and we use our results to understand which of these rules seem to be best suited for each application. In particular, we show that STV (one of the few nontrivial rules used in real high-stake elections) exhibits excellent performance, whereas the Bloc rule (also often used in practice) performs poorly. We also visualise three approximation algorithms for the computationally hard Chamberlin–Courant and Monroe rules. Our results show that the best approximation algorithms on offer (one of which is introduced in this paper) can be safely used instead of the original rules themselves.

Everyone welcome!

2018 CMSS Research Seminar by Piotr Faliszewski

Speaker: Piotr Faliszewski (Krakow)

Date, Time and Venue: Monday, 26 February 2018, 15:00-16:00, CaseRoom4/260-009 [Business School Building, Level 0]

Abstract: We start from examples of multiwinner voting rules, then dwell on a particular class of those rules, called committee scoring rules, which we consider in detail from axiomatic, algorithmic and experimental perspective. We promise beautiful pictures!

Everyone welcome!

2018 CMSS Research Seminar by Han Bleichrodt

Speaker: Han Bleichrodt (Erasmus University Rotterdam and ANU)

Date, Time and Venue: Friday, 23 February 2018, 12:00-13:00, room 260-6115 [Business School Building, Level 6]

Title: Testing Hurwicz Expected Utility

Abstract: Gul and Pesendorfer (2015) propose a new theory of ambiguity, they dub Hurwicz expected utility (HEU). HEU is the first axiomatic theory that is consistent with most of the available empirical evidence on decision under uncertainty. We show that HEU is also tractable and a particular subclass can readily be estimated and tested. We do this by requiring the probability weighting functions in the HEU representation to come from a two-parameter family. We investigate two predictions of HEU. The first prediction is that ambiguity aversion is constant across different sources of ambiguity. We investigate this utilizing the data of Abdellaoui et al. (2011). We observe support for it in their most extensive data set, but not in the other data set. The second prediction is that ambiguity aversion and first-order risk aversion (Segal and Spivak, 1990) are positively correlated. We perform an experiment to test this prediction. As the positive correlation revealed in the data is only slight to fair we conclude the evidence of a positive relation between ambiguity aversion and first order risk aversion is not conclusive.

Everyone welcome!