"How Good Are Privacy Guarantees? Platform Architecture and the Learning-Privacy Tradeoff" - Alireza Fallah      

Abstract: Many platforms deploy data collected from users for a multitude of purposes. While some are beneficial to users, others are costly to their privacy. The presence of these privacy costs means that platforms may need to provide guarantees about how and to what extent user data will be harvested for activities such as targeted ads, individualized pricing, and sales to third parties. In this work, we build a multi-stage model in which users decide whether to share their data based on privacy guarantees. We first introduce a novel mask-shuffle mechanism and prove it is Pareto optimal—meaning that it leaks the least about the users’ data for any given leakage about the underlying common parameter. We then show that under any mask-shuffle mechanism, there exists a unique equilibrium in which privacy guarantees balance privacy costs and utility gains from the pooling of user data for purposes such as assessment of health risks or product development. Paradoxically, we show that as users’ value of pooled data increases, the equilibrium of the game leads to lower user welfare. This is because platforms take advantage of this change to reduce privacy guarantees so much that user utility declines (whereas it would have increased with a given mechanism). Furthermore, we show that platforms have incentives to choose data architectures that systematically differ from those that are optimal from the user’s point of view. In particular, we identify a class of pivot mechanisms, linking individual privacy to choices by others, which platforms prefer to implement and which make users significantly worse off. 

Based on joint work with Daron Acemoglu, Ali Makhdoumi, Azarakhsh Malekian, and Asu Ozdaglar.

We will have some short presentations and a moderated discussion with panelists, who have experience in (re-)designing such applications, and then open discussion on possible research questions on various aspects, including policy design, optimisation, and implementation. We share knowledge on the specific features of applications, availability of real and generated datasets, and codes/software solutions. The aim is to better understand the current trends, new policy approaches, and practical challenges in implementation in the spirit of "Engineering Economics".

"School Choice with Information Inequality" - Negar Matoorian

Abstract:  “School choice policies are predicated on the assumption that parents have enough information to make an informed decision on where to send their children… districts don't always do a good job of disseminating information and explaining options to certain groups – low-income parents and those whose children are first-generation Americans, for example” (US News & World Report, Aug 2015). 

We study the consequence of unequal access to school value information in the popular mechanisms for allocation of seats in the public school system. We show that in the Deferred Acceptance mechanism students with inferior information suffer from curse of assignment and might try to mitigate this curse by ranking schools with higher capacity or schools they have priority at prior to schools that are known to have a higher educational quality. Our model lays the ground for studying the welfare consequences of policies such as information disclosure and priority design in school choice environments with asymmetric information.

"Deterministic Budget-Feasible Clock Auctions" - Dan Schoepflin

Abstract: We revisit the well-studied problem of budget-feasible procurement, where a buyer with a strict budget constraint seeks to acquire services from a group of strategic providers (the sellers). During the last decade, several strategyproof budget-feasible procurement auctions have been proposed, aiming to maximize the value of the buyer, while eliciting each seller's true cost for providing their service.  Our main result is a novel method for designing budget-feasible auctions, leading to solutions that outperform the previously proposed auctions in multiple ways.  First, our solutions take the form of descending clock auctions, rather than sealed-bid auctions, and thus satisfy a list of very appealing properties, such as obvious strategyproofness and unconditional winner privacy, making these auctions much more likely to be used in practice. Second, in contrast to previous results that heavily depend on randomization, our auctions are deterministic. In fact, our deterministic auctions achieve a constant-approximation for submodular  valuations, resolving a main open question in this literature.  Finally, we improve the best-known approximation factor for monotone submodular valuations, the focus of most of the prior work. 

In this informal presentation, I'll tell you about my personal experience with courts, commissions, nonprofits, and policymakers in the last three years on issues around voting and civil rights.  This includes expert work for redistricting litigation in seven states and one city, plus work designing the partisan gerrymandering test that made its way before Congress.  All welcome!  Lots of pictures!  Talk will be followed by a trip down the hill for happy hour.