Matching and Market Design at INFORMS in San Francisco, Sunday November 9

There's a lot of market design at the INFORMS annual meeting, Nov 9-12.

On Sunday I'll start the day off with a talk from  10-10:50 called
"Market Design and the Economist as Engineer."

That will be followed by a cluster of talks organized by Itai Ashlagi called (embarrassingly)
Matching and Market Design (in honor of Al Roth), consisting of the following sessions

Cluster :	Matching and Market Design (in honor of Al Roth)
Session Information		: Sunday Nov 09, 11:00 - 12:30
Title:	Empirical Market Design
Chair:	Ramesh Johari,Stanford University,
Abstract Details
Title:	Quality Externalities and the Limits of Reputation in Two-Sided Markets
	Presenting Author:	Steve Tadelis,Professor, UC Berkeley, Haas School of Business, 2220 Piedmont Ave, Berkeley, United States of America, stadelis@haas.berkeley.edu
	Co-Author:	Chris Nosko,Booth School of Business, University Of Chicago, Chicago, United States of America, cnosko@chicagobooth.edu
Abstract:		Using data from eBay, we argue that platforms can mitigate externalities by actively screening sellers and promoting the prominence of better quality sellers. Exploiting the bias in feedback, we create a measure of seller quality and demonstrate the benefits of our approach through a controlled experiment that prioritizes better quality sellers to a random subset of buyers. .
Title:	On the Near Impossibility of Measuring the Returns to Advertising
	Presenting Author:	Randall Lewis,Economic Research Scientist, Google Inc., 1600 Amphitheatre Parkway, Mountain View Ca 94043, United States of America, randall@econinformatics.com
	Co-Author:	Justin Rao,Economic Research Scientist, Microsoft Research, New York City NY, United States of America, Justin.Rao@microsoft.com
Abstract:		Firms have a hard time measuring the causal impact of advertising expenditures on profit. In twenty-five online field experiments, individual-level sales are volatile relative to the per capita cost of a campaign--a small impact on a noisy dependent variable can generate positive returns. Experiments can need more than ten million person-weeks. Further, small selection biases can severely bias observational estimates. Weak informational feedback and technological advances shape ad marketplaces.
Title:	Corporate Prediction Markets: Evidence from Google, Ford, and Firm X
	Presenting Author:	Bo Cowgill,UC Berkeley, 1931 Diamond St Apt 3, SAN FRANCISCO Ca 94131, United States of America, bo.cowgill@gmail.com
	Co-Author:	Eric Zitezwitz,Dartmouth College, 6106 Rockefeller Hall, Hanover NH, United States of America, zitzewitz@dartmouth.edu
Abstract:		We examine data from prediction markets run by Google, Ford and Firm X (a large private materials company). Despite theoretically adverse conditions, we find these markets are relatively efficient, and improve upon the forecasts of experts at all three firms by as much as a 25% reduction in MSE. The most notable inefficiency is an optimism bias in the markets at Google and Ford. The inefficiencies that do exist become smaller over time for reasons we document.
Title:	At What Quality and What Price? Inducing Separating Equilibria as a Market Design Problem
	Presenting Author:	John Horton,Professor, NYU Stern School of Business, Kaufman Management Center, 44 West Fourth St, 8-81, New York NY 10012, United States of America, john.joseph.horton@gmail.com
	Co-Author:	Ramesh Johari,Stanford University, Huang 311, Stanford, United States of America, ramesh.johari@stanford.edu
Abstract:		A tool to promote revelation of buyers' price/quality preferences was experimentally introduced into an online labor market. In the treatment cells of the experiment, upon posting a job, buyers chose what price/quality level they were seeking from sellers. We find that buyers readily reveal their preferences and that this revelation---which itself was experimentally manipulated---strongly induced seller-side sorting.

Title:	******** Matching and Market Design
Chair:	Jacob Leshno,Columbia University,
Abstract Details
Title:	Matching in Networks
	Presenting Author:	Michael Ostrovsky,Associate Professor of Economics, Stanford Graduate School of Business, 655 Knight Way, Stanford CA 94305, United States of America, ostrovsky@stanford.edu
Abstract:		In this talk, I will present results on the existence and properties of stable outcomes in trading networks.
Title:	Matching with Peers in School Choice
	Presenting Author:	Atila Abdulkadiroglu,Professor, Duke University, 213 Social Sciences Building, Durham NC 27708, United States of America, atila.abdulkadiroglu@duke.edu
Abstract:		We develop a theory for matching of students to schools with peers and study various matching mechanisms with field data.
Title:	Endogenous preferences and the role of the mechanism in school choice
	Presenting Author:	Estelle Cantillon,Senior Research Fellow, Université Libre de Bruxelles (ECARES), 50, av FD Roosevelt, CP 114, Brussels 1050, Belgium, Estelle.Cantillon@ulb.ac.be
Abstract:		We consider a school choice model where preferences over schools are endogenous because students care about the quality of their peers. In such a setting, the mechanism affects the degree of preference polarization. We show how mechanisms can be designed to reduce polarization and improve the distribution of ranks of assigned schools in students’ preferences. A policy change in the city of Ghent (Belgium) provides a test for the predictions of the theory.
Title:	Evidence of Strategic Behavior in Hospital Claims Reporting
	Presenting Author:	Hamsa Bastani,Stanford University, Stanford, Stanford, United States of America, hsridhar@stanford.edu
	Co-Author:	Mohsen Bayati,Stanford Graduate School of Business, Stanford CA 94305, United States of America, bayati@gsb.stanford.edu
		Joel Goh,joelgoh@stanford.edu
		Stefanos Zenios,Charles A. Holloway Professor of Operations, Information, and Technology and Professor of Health Care Management, Stanford Graduate School of Business, 655 Knight Way, Stanford CA 94305, United States of America, stefzen@GSB.Stanford.Edu
Abstract:		We provide evidence from Medicare claims data that hospitals engage in upcoding behavior when reporting hospital-acquired infections that are no longer reimbursed by Medicare. In particular, we show that hospitals sometimes mark a hospital-acquired infection as present-on-admission, presumably in order to collect greater reimbursement.

*******

Title:	Matching Markets
Chair:	Yash Kanoria,Columbia Business School,
Abstract Details
Title:	Stable Matching in Large Economies
	Presenting Author:	Fuhito Kojima,Stanford University, 579 Serra Mall, Stanford CA 943055007, United States of America, fkojima@stanford.edu
Abstract:		Complementarities of preferences have been known to jeopardize stability of two-sided matching markets, yet they are a pervasive feature in many matching markets. In large markets, we demonstrate that if each firm's choice changes continuously as the set of available workers changes, then there exists a stable matching even if firm preferences exhibit complementarity. Building on this result, we show that there exists an approximately stable matching in any large finite economy.
Title:	The Prior-Independence Approach
	Presenting Author:	Inbal Talgam-Cohen,PhD Candidate, Stanford University, 86 Hulme Ct, Apt 108, Stanford CA 94305, United States of America, italgam@stanford.edu
	Co-Author:	Tim Roughgarden,Stanford, 353 Serra Street, Stanford, United States of America, tim@cs.stanford.edu
Abstract:		The matching literature has recently begun to consider priors on agents’ utilities. One of the barriers to adopting this potentially very fruitful approach is that priors add significant informational assumptions to the model. We survey a successful alternative approach from mechanism design called prior independence, which alleviates such assumptions while still reaping most benefits. We discuss both sampling-based methods and methods based on ensuring sufficient competition in the market.
Title:	The structure of the core in assignment markets
	Presenting Author:	Yash Kanoria,Columbia Business School, 404 Uris Hall, New York NY 10027, United States of America, ykanoria@columbia.edu
	Co-Author:	Daniela Saban,Columbia University, Uris Hall, 4I, New York, United States of America, dhs2131@columbia.edu
		Jay Sethuraman,Columbia University, IEOR Department, New York, United States of America, jay@ieor.columbia.edu
Abstract:		Assignment markets (Shapley & Shubik 1971) involve matching with transfers, as in labor markets and housing markets. We consider a two-sided assignment market with agent types and stochastic structure similar to models used in empirical studies. Each agent has a randomly drawn "productivity" associated with each type on the other side. We characterize how the structure of the core, i.e., the set of stable outcomes, is determined by market characteristics.