|
on Discrete Choice Models |
| By: | Fuhito Kojima (Stanford University); M. Utku Ünver (Boston College) |
| Abstract: | The Boston mechanism is a popular student placement mechanism in school choice programs around the world. We provide two characterizations of the Boston mechanisms. We introduce two new axioms, respect of preference rankings and rank-respecting Maskin monotonicity. A mechanism is the Boston mechanism for some priority if and only if it respects preference rankings and satisfies consistency, resource monotonicity, and rank-respecting Maskin monotonicity. In environments where each type of object has exactly one unit, as in house allocation, a characterization is given by respect of preference rankings, individual rationality, population monotonicity, and rank-respecting Maskin monotonicity. |
| Keywords: | Mechanism design, matching, school choice, market design, Boston mechanism |
| JEL: | C78 D78 |
| Date: | 2010–02–04 |
| URL: | http://d.repec.org/n?u=RePEc:boc:bocoec:729&r=dcm |
| By: | Koji Miyawaki (National Institute for Environmental Studies); Yasuhiro Omori (Faculty of Economics, University of Tokyo); Akira Hibiki (National Institute for Environmental Studies) |
| Abstract: | Block rate pricing is often applied to income taxation, telecommunication services, and brand marketing in addition to its best-known application in public utility services. Under block rate pricing, consumers face piecewise-linear budget constraints. A discrete/ continuous choice approach is usually used to account for piecewise-linear budget constraints for demand and price endogeneity. A recent study proposed a methodology to incorporate a separability condition that previous studies ignore, by implementing a Markov chain Monte Carlo simulation based on a hierarchical Bayesian approach. To extend this approach to panel data, our study proposes a Bayesian hierarchical model incorporating the individual effect. The random coefficients model result shows that the price and income elasticities are estimated to be negative and positive, respectively, and the coefficients of the number of members and the number of rooms per household are estimated to be positive. Furthermore, the AR(1) error component model suggests that the Japanese residential water demand does not have serial correlation. |
| Date: | 2010–02 |
| URL: | http://d.repec.org/n?u=RePEc:tky:fseres:2010cf717&r=dcm |