|
on Discrete Choice Models |
| By: | J.S. Cramer (University of Amsterdam) |
| Abstract: | In binary discrete regression models like logit or probit the omis- sion of a relevant regressor (even if it is orthogonal) depresses the re- maining <font face="Symbol">b</font> coefficients towards zero. For the probit model, Wooldridge (2002) has shown that this bias does not carry over to the effect of the regressor on the outcome. We find by simulations that this also holds for logit models, even when the omitted variable leads to severe misspecification of the disturbance. More simulations show that es- timates of these effects by logit analysis are also impervious to pure misspecification of the disturbance. |
| Keywords: | logit model; omitted variables; misspecification |
| JEL: | C25 |
| Date: | 2005–09–15 |
| URL: | http://d.repec.org/n?u=RePEc:dgr:uvatin:20050084&r=dcm |
| By: | Fainmesser, Itay; Fershtman, Chaim; Gandal, Neil |
| Abstract: | The NCAA college football ratings, in which the "so-called" national champion is determined, has been plagued by controversies the last few years. The difficulty arises because there is a need to make a complete ranking of teams even though each team has a different schedule of games with a different set of opponents. A similar problem arises whenever one wants to establish a ranking of patents or academic journals, etc. in which the raw data are (incomplete) bilateral citations or interactions among objects. This paper develops and estimates a simple consistent weighted ranking (CWR) scheme which, in the sports world, depends on four parameters (winning vs. losing and the relative importance of home vs. away games). In most ranking problems, there are not explicit criteria to evaluate the success of proposed rankings. NCAA college football has a special structure that enables the evaluation of each ranking scheme. Each season is essentially divided into two parts: the regular season and the post season bowl games. If a ranking scheme is accurate it should correctly predict a relatively large number of the bowl game outcomes. We use this structure to estimate the four parameters of our ranking function using "historical" data from the 1999-2003 seasons. |
| Keywords: | NCAA football; ranking; sport tournament |
| JEL: | C8 L83 |
| Date: | 2005–09 |
| URL: | http://d.repec.org/n?u=RePEc:cpr:ceprdp:5239&r=dcm |
| By: | Haluk Ergin (MIT); Tayfun Sönmez (Boston College) |
| Abstract: | Many school districts in the U.S. use a student assignment mechanism that we refer to as the Boston mechanism. Under this mechanism a student loses his priority at a school unless his parents rank it as their first choice. Therefore parents are given incentives to rank high on their list the schools where the student has a good chance of getting in. We characterize the Nash equilibria of the induced preference revelation game. An important policy implication of our result is that a transition from the Boston mechanism to the student-optimal stable mechanism would lead to unambiguous efficiency gains. |
| Keywords: | student assignment, Boston mechanism, induced preference revelation, mechanism design |
| JEL: | C78 D61 D78 I20 |
| Date: | 2005–09–12 |
| URL: | http://d.repec.org/n?u=RePEc:boc:bocoec:619&r=dcm |
| By: | Proost Stef (K.U.Leuven-Center for Economic Studies; UCL - CORE); Van der Loo Saskia (K.U.Leuven-Center for Economic Studies; UCL - CORE); Andre de Palma (Université de Cergy-Pontoise, ENPC and Member of Institut Universitaire de France, THEMA, 33,); Lindsey Robin (Department of Economics, University of Alberta) |
| Abstract: | A proposal has been made to build a new tunnel under the Scheldt river near the centre of Antwerp in order to relieve traffic congestion on the ring road and in an existing tunnel. The new tunnel is expected to cost more than €1 billion, and tolls have been suggested to help finance construction and to manage demand. This paper conducts a preliminary cost-benefit analysis of a new tunnel and three alternative tolling schemes, and compares them with a do-nothing scenario and an option to toll the existing tunnel without building a new one. The analysis is performed using a model that was recently developed as part of the European-Union funded REVENUE project. The two tunnels are treated as imperfect substitutes, and a multi-year accounting framework is adopted that accounts for emissions, accidents and noise externalities, road damage, revenues accruing to the national and regional governments from existing transport user charges, and the salvage value of the new tunnel. With the base-case parameter values it is found that building the tunnel is worthwhile with all three tolling regimes and yields a higher benefit than not building the tunnel and tolling the old one. Nevertheless, the net benefit from building the tunnel differs appreciably between tolling regimes, and it is sensitive to the value assumed for the marginal cost of public funds. |
| Keywords: | infrastructure investment, route choice, congestion, tolls |
| JEL: | D61 R41 R48 |
| Date: | 2005–09 |
| URL: | http://d.repec.org/n?u=RePEc:ete:etewps:ete0507&r=dcm |
| By: | G. LANINE; R. VANDER VENNET |
| Abstract: | The Russian banking sector experienced considerable turmoil in the late 1990s, especially around the Russian banking crisis in 1998. The question is what types of banks are vulnerable to shocks and whether or not bank-specific characteristics can be used to predict vulnerability to failures. In this study we employ a parametric logit model and a nonparametric trait recognition approach to predict failures among Russian commercial banks. We test the predictive power of both models based on their prediction accuracy using holdout samples. Both models performed better than the benchmark; the trait recognition approach outperformed logit in both the original and the holdout samples. As expected liquidity plays an important role in bank failure prediction, but also asset quality and capital adequacy turn out to be important determinants of failure. |
| Keywords: | Russian banks, bank failure prediction, logit model, trait recognition, forecasting accuracy |
| JEL: | C25 G21 G33 |
| Date: | 2005–08 |
| URL: | http://d.repec.org/n?u=RePEc:rug:rugwps:05/329&r=dcm |
| By: | Patrick Paul Walsh; Franco Mariuzzo; Ciara Whelan (Department of Economics, Trinity College) |
| Abstract: | Thresholds defined on the level and change in the HHI (Herfindahl- Hirschmann Index) applied to market shares seem to be the main instrument to select notified mergers for investigation in both the EU and US. We question the use of such a selection rule in di?erentiated products industries. We propose the use of a structural approach to apply HHI thresholds based on profit shares rather than market shares. We illustrate our point using product data for Retail Carbonated Soft Drinks (Price, Market Share and Characteristics). We estimate company (product) mark-ups consistent with a structural model of equilibrium, using demand primitives from a Nested Logit model and a Random Coe?cient model. We provide an example where the HHI thresholds based on profit shares identify potentially damaging mergers not captured by applying thresholds to output shares, or conversely, identify mergers of no concern that would be selected on the basis of output shares. |
| JEL: | K2 L11 L25 L40 L81 |
| Date: | 2005–08 |
| URL: | http://d.repec.org/n?u=RePEc:tcd:wpaper:tep10&r=dcm |
| By: | Robin Hogarth; Natalia Karelaia |
| Abstract: | An important problem faced by boundedly rational agents is to identify “regions of rationality,” i.e., the areas for which simple, boundedly rational models are and are not effective. To map the contours of such regions, we derive probabilities that models identify the best of m alternatives (m > 2) characterized by k attributes (k > 1). The models include a single variable (lexicographic), variations of elimination-by-aspects, equal weighting, hybrids of the preceding, and models exploiting dominance. We compare all with multiple regression. We illustrate the theory with twenty simulated and four empirical datasets. Fits between predictions and realizations are excellent. However, the terrain mapped by our work is complex and no single model is “best.” We further provide an overview by regressing the performance of the different models on factors characterizing environments. We conclude with suggestions for further research as well as implications for the concept of rationality in economics. |
| Keywords: | Decision making, Bounded rationality, Lexicographic rules, Choice theory |
| JEL: | D81 M10 |
| Date: | 2005–04 |
| URL: | http://d.repec.org/n?u=RePEc:upf:upfgen:828&r=dcm |
| By: | Libertad González Luna |
| Abstract: | This paper examines the effect of public assistance, labor market and marriage market conditions on the prevalence of single mother families across countries and over time. A multinomial logit derived from a random utility approach is estimated using individuallevel data for 14 countries. I find evidence that increases in the level of public support are significantly and positively associated with a higher incidence of both never married and divorced mothers. The results also suggest that single mothers are more prevalent when female wages are lower. Higher male earnings and employment opportunities in a woman’s marriage market appear to lead to fewer never married mothers, but more divorced mothers. Higher child support or alimony payments are associated with a higher prevalence of divorced mothers. |
| Keywords: | Single mothers, marriage, fertility, welfare benefits, marriage markets |
| JEL: | J12 J13 I38 |
| Date: | 2005–07 |
| URL: | http://d.repec.org/n?u=RePEc:upf:upfgen:876&r=dcm |