Market Design with Financially Constrained Agents
National Science Foundation, SES-0721053
Principal Investigator: Yeon-Koo Che
Many important resources, such as housing, health care, human organs, and education, are often allocated via methods that depart from "classic" markets. The current project provides an efficiency foundation for the use of such non-market methods and investigates ways to improve upon them, with a special attention given to the public school choice programs. The key element of the current research is the recognition that individuals are often financially constrained, some more so than others, due to their unequal access to capital markets. This inequality in financial constraints facing individuals causes standard markets to be prejudiced in allocation toward wealthy individuals, making the outcome inefficient, and not just inequitable, even compared with simple random assignment of resources, provided that the recipients of the resources are allowed to resell them. Allowing the recipients of benefits or entitlements to transfer them is crucial for non-market methods of resource allocation to be efficient, but this research also identifies the cost associated with unrestricted transferability in engendering speculation activities.
Having established a general efficiency foundation for non-market methods of resource allocation, the research then investigates ways to improve upon existing methods, particularly those adopted in public school choice programs. In particular, the project proposes a new procedure which would allow participants to signal their preferences, but in a way that limits their incentives to "game the system." The research demonstrates this new procedure of school choice to be more efficient than the existing procedures. The project next tests experimentally the efficiency benefit associated with the new procedure as well as the role of information available to the individuals participating in the program. Finally, the project investigates the roles housing markets play in the traditional residence-based regime as well as in the new school choice regime.
Broader Impact: Many issues under investigation have significant policy implications. The first part finds a tradeoff associated with allowing recipients of benefits or entitlements to transfer them. This finding has broad applications for many government programs, especially those which currently do not permit such transfers, including assignment of immigration visas and access to human organs. The second part on school choice design has more immediate and tangible applications. The proposed procedure has a potential for assigning students to schools more efficiently than the methods that have been used or proposed thus far and may influence the way school choice programs are designed in many parts of the country. Lastly, this project promises to yield a better understanding about the role of housing market in school selection in general, but particularly about the merit of the so-called "walk zone'' priority, which is often awarded to the students who reside in the immediate vicinity of a school to favor their assignment to that school.
Working Papers and Publications
Yeon-Koo Che & Fuhito Kojima, 2008. "Asymptotic Equivalence of Probabilistic Serial and Random Priority Mechanisms," Cowles Foundation Discussion Papers 1677, Cowles Foundation for Research in Economics, Yale University.
Full text: http://www.columbia.edu/~yc2271/files/papers/psrp28.pdf
Atila Abdulkadiroglu, Yeon-Koo Che and Yosuke Yasuda, 2009. “Resolving Conflicting Preferences in School Choice: The 'Boston' Mechanism Reconsidered”
Full text: http://www.columbia.edu/~yc2271/files/papers/Boston-WP2.pdf
Yeon-Koo Che and Ian Gale. 2007. "Market versus Non-Market Assignment of Ownership" Law and Economics Workshop.
Full text: http://www.columbia.edu/~yc2271/files/papers/nonmarket.pdf