|
on Discrete Choice Models |
| By: | Esmerelda A. Ramalho; Richard Smith (Institute for Fiscal Studies and University of Warwick) |
| Abstract: | Missing values are endemic in the data sets available to econometricians. This paper suggests a unified likelihood-based approach to deal with several nonignorable missing data problems for discrete choice models. Our concern is when either the dependent variable is unobserved or situations when both dependent variable and covariates are missing for some sampling units. These cases are also considered when a supplementary random sample of observations on all covariates is available. A unified treatment of these various sampling structures is presented using a formulation of the nonresponse problems as a modification of choice-based sampling. Extensions appropriate for nonresponse are detailed of Imbens’ (1992) effcient generalized method of moments (GMM) estimator for choice-based samples. Simulation evidence reveals very promising results for the various GMM estimators proposed in this paper. |
| JEL: | C25 C51 |
| Date: | 2003–07 |
| URL: | http://d.repec.org/n?u=RePEc:ifs:cemmap:wp07/03&r=dcm |
| By: | Francesco De Sinopoli; Giovanna Iannantuoni |
| Abstract: | Most of the voting models restrict themselves to the analysis of symmetric equilibria, i.e. equilibria in which ‘similar’ voters make ‘similar’ voting decisions. In this paper we investigate this assumption under costly plurality voting. In any pure strategy equilibrium, if two active voters have the same preference order over candidates, they do vote for the same candidate. However, as an example shows, this type of result cannot be hoped for mixed strategies equilibria. |
| Keywords: | Strategic Voting, Symmetric Equilibria |
| JEL: | C72 D72 |
| Date: | 2005–05 |
| URL: | http://d.repec.org/n?u=RePEc:cam:camdae:0521&r=dcm |
| By: | Renos Vakis (World Bank); Elisabeth Sadoulet (University of California, Berkeley); Alain de Janvry (University of California, Berkeley) |
| Abstract: | Farmers incur proportional and fixed transactions costs in selling their crops on markets. Using data for Peruvian potato farmers, we propose a method to measure these transactions costs. When opportunities exist to sell a crop on alternative markets, the observed choice of market can be used to infer a monetary measure of transactions costs in market participation. The market choice model is first estimated at the reduced form level with a conditional logit, as a function of variables that explain transactions costs. We then use these market choice equations to control for selection in predicting the idiosyncratic prices that would be received on all markets and the idiosyncratic proportional transactions costs that would be incurred to reach all markets. The net between the two gives us a measure of effective farm-level prices. This allows us to estimate a semi-structural conditional logit of the market choice model. In this model, the choice of market is a function of predicted effective farm-level prices, and of market information that accounts for fixed transactions costs. We can use the estimated coefficients to derive the price equivalence of the fixed cost due to information. We find that the information on market price that farmers receive from their neighbors reduces fixed transactions costs by the equivalent of doubling the price received, and is equal to four times the average transportation cost. |
| Keywords: | transactions costs, market choice, information, |
| Date: | 2003–10–01 |
| URL: | http://d.repec.org/n?u=RePEc:cdl:agrebk:962&r=dcm |
| By: | Gerlach, Stefan |
| Abstract: | This Paper discusses interest rate setting by the ECB between 1999 and 2004. I develop from the Monthly Bulletins quantitative indicators of the Governing Council’s assessment of inflation, economic activity, and M3 growth, and investigate their impact on its interest rate decisions. I also estimate reaction functions with ordered probit techniques, using the Monthly Bulletins to guide the choice of variables for the analysis. The results show that the ECB reacts strongly to economic sentiment indicators as measures of the state of the real economy. Furthermore, I find statistically significant reactions to inflation and M3 growth. |
| Keywords: | ECB; empirical reaction functions; ordered probit |
| JEL: | E43 E52 E58 |
| Date: | 2004–12 |
| URL: | http://d.repec.org/n?u=RePEc:cpr:ceprdp:4775&r=dcm |
| By: | Mosquera,M.A.; Borm,P.; Fiestras-Janeiro,M.G.; Garcia-Jurado,I.; Voorneveld,M. (Tilburg University, Center for Economic Research) |
| Abstract: | The class of maximin actions in general decision problems is characterized. |
| JEL: | C70 D81 |
| Date: | 2005 |
| URL: | http://d.repec.org/n?u=RePEc:dgr:kubcen:200554&r=dcm |
| By: | Bergh, Andreas (Department of Economics, Lund University); Engseld, Peter (Department of Economics, Lund University) |
| Abstract: | The standard method when analyzing the problem of cooperation using evolutionary game theory is to assume that people are randomly matched against each other in repeated games. In this paper we discuss the implications of allowing agents to have preferences over possible opponents. We model reputation as a noisy observation of actual propensity to cooperate and illustrate how reputation based choice of opponents can explain both the emergence and deterioration of cooperation. We show that empirical and experimental evidence of cooperation is consistent with our hypothesis that people behave so as to minimize the risk of damaging their reputation as nice, cooperative persons. |
| Keywords: | Cooperation; Prisoners Dilemma; Signaling; Reputation; Altruism; Institutions |
| JEL: | C70 C90 |
| Date: | 2005–04–28 |
| URL: | http://d.repec.org/n?u=RePEc:hhs:lunewp:2005_027&r=dcm |
| By: | François Gardes; Pierre Garrouste |
| Abstract: | Jean Ville gave, independently of Houthakker, and prior to him, a general one page proof of the integrability of demand functions in a revealed preference scheme. It happens that this essential contribution has been largely ignored in the literature. The comparison between Ville and Houthakker’s proofs makes room for discussing the assumptions necessary to encompass the discrete version of the acyclicity into a continuous version. |
| URL: | http://d.repec.org/n?u=RePEc:icr:wpicer:30-2004&r=dcm |
| By: | James Heckman (Institute for Fiscal Studies and University of Chicago); Rosa Matzkin; Lars Nesheim (Institute for Fiscal Studies) |
| Abstract: | We analyze equilibria in hedonic economies and study conditions that lead to identification of structural preference parameters in hedonic economies with both additive and nonadditive marginal utility and marginal product functions. The latter class is more general, allows for heterogeneity in the curvature of consumer utility, and can result in conditions that lead to bunching. Such bunching has been largely ignored in the previous literature. We then present methods to estimate marginal utility and marginal product functions that are nonadditive in the unobservable random terms, using observations from a single hedonic equilibrium market. These methods are important when statistical tests reject additive specifications or when prior information suggests that consumer or firm heterogeneity in the curvature of utility or production functions is likely to be significant. We provide conditions under which these types of utility and production functions are nonparametrically identified, and we propose nonparametric estimators for them. The estimators are shown to be consistent and asymptotically normal. When the assumptions required to use single market methods are unjustified, we show how multimarket data can be used to estimate the structural functions. |
| Date: | 2005–03 |
| URL: | http://d.repec.org/n?u=RePEc:ifs:cemmap:wp03/05&r=dcm |
| By: | Laura Blow (Institute for Fiscal Studies); Martin Browning (Institute for Fiscal Studies and University of Copenhagen); Ian Crawford (Institute for Fiscal Studies and University of Surrey) |
| Abstract: | Characteristics models have been found to be useful in many areas of economics. However, their empirical implementation tends to rely heavily on functional form assumptions. In this paper we develop a revealed preference-based nonparametric approach to characteristics models. We derive the minimal necessary and sufficient empirical conditions under which data on the market behaviour of individual, heterogeneous, pricetaking consumers are nonparametrically consistent with the consumer characteristics model. Where these conditions hold, we show how information may be recovered on individual consumer’s marginal valuations of product attributes. In some cases marginal valuations are point identi- fied and in other cases we can only recover bounds. Where the conditions fail we highlight the role which the introduction of unobserved product attributes can play in rationalising the data. We implement these ideas using consumer panel data on the Danish milk market. |
| Keywords: | Product characteristics, revealed preference |
| JEL: | C43 D11 |
| Date: | 2004–12 |
| URL: | http://d.repec.org/n?u=RePEc:ifs:cemmap:wp18/04&r=dcm |
| By: | Emawtee Bissoondoyal-Bheenick; Robert Brooks; Angela Y.N.Yip |
| Abstract: | The paper compares two alternative techniques for the modelling of the determinants of sovereign ratings, specifically, ordered probit and case-based reasoning. Despite the differences in approach the two alternative modelling approaches produce similar results in terms of which variables are significant and forecast accuracy. This suggests that either approach can be used, and that there is some robustness in the results. As regards significant variables, both models find that a proxy for technological development, specifically, mobile phone use, is the most important variable. Apart from the technology proxy, a range of conventional macroeconomic variables are found to be significant, in particular GDP and inflation. The models are then used to produce forecasts for 2002 and for a set of unrated countries. The forecast comparison indicates the critical role played by the technology proxy variable in the modelling. |
| Keywords: | Sovereign Ratings, Ordered Response Models, Case-Based Reasoning |
| JEL: | G15 |
| Date: | 2005–05 |
| URL: | http://d.repec.org/n?u=RePEc:msh:ebswps:2005-9&r=dcm |
| By: | José Apesteguía (Departamento de Economía-UPNA); Miguel A. Ballester (Departamento de Economía-UPNA) |
| Abstract: | Kalai, Rubinstein, and Spiegler (2002) propose the rationalization of choice functions that violate the “independence of irrelevant alternatives” axiom through a collection (book) of linear orders (rationales). In this paper we present an algorithm which, for any choice function, gives (i) the minimal number of rationales that rationalizes the choice function, (ii) the composition of such rationales, and (iii) information on how choice problems are related to rationales. As in the classical case, this renders the information given by a choice function completely equivalent to that given by a minimal book of rationales. We also study the structure of several choice procedures that are prominent in the literature. |
| Keywords: | Rationalization, Independence of irrelevant alternatives, Order partition, Computational effort. |
| URL: | http://d.repec.org/n?u=RePEc:nav:ecupna:'0501'&r=dcm |
| By: | Fan-chin Kung (Academia Sinica) |
| Abstract: | We study how individuals divide themselves into coalitions and choose a public alternative for each coalition. When preferences have consecutive support and coalition feasible sets are positively population- responsive, the proposed consecutive benevolence solution generates allocations belonging to the coalition structure core and that are also Tiebout equilibria. However, when each coalition follows a single-valued collective decision rule, the coalition structure core may be empty. Our results show that if individual preferences are, in a sense, similar and if members can be as well off when a coalition enlarges, then a stable formation of collective decision-making units can be guaranteed. A predetermined decision rule makes coalitions less stable. |
| JEL: | C62 C71 D71 |
| Date: | 2005–05–05 |
| URL: | http://d.repec.org/n?u=RePEc:wpa:wuwpga:0505002&r=dcm |
| By: | Greg Hannsgen (The Levy Economics Institute) |
| Abstract: | In dealing with the problematic relationship of morality to rational choice theory, neoclassical economists since Lionel Robbins have often argued that they can incorporate moral values into consumer theory by putting those values into the utility function. This paper tests the viability of such an approach in the context of international finance. The moral value at stake is autonomy, which may be lost when borrowers must submit to the edicts of international financial institutions. When such a value is inserted into the utility function of a small economy, the growth rate of consumption and the level of investment change. Furthermore, potential borrowers may lose their ability to credibly commit to paying back loans, resulting in a complete absence of borrowing where it might otherwise take place. The author argues that while this model illustrates the possibility of analyzing a noneconomic value (sovereignty) through rational choice theory, it also shows that standard methods of empirical inference, policy evaluation, and welfare analysis may fail in such a situation. To answer questions that mix morality and economics, economists must seek tools other than conventional rational choice theory. |
| Keywords: | : Values, Lionel Robbins, international debt, methodology |
| JEL: | B |
| Date: | 2005–05–02 |
| URL: | http://d.repec.org/n?u=RePEc:wpa:wuwpmh:0505001&r=dcm |
| By: | Riccardo SCARPA (Alma Mater Studiorum-Università di Bologna); Fiorenza SPALATRO (Alma Mater Studiorum-Università di Bologna); Maurizio CANAVARI (Alma Mater Studiorum-Università di Bologna) |
| Abstract: | This paper reports some preliminary results on a mixed logit random utility analysis of conjoint data from costumers' preferences over agricultural products. The data are collected via a telematic sample representative of Italian households. The survey instrument was implemented via a computer supported system. A multivariate normal full correlation structure is imposed in the mixed logit estimation and the implications of such a taste structure are examined. |
| JEL: | P Q Z |
| Date: | 2005–05–03 |
| URL: | http://d.repec.org/n?u=RePEc:wpa:wuwpot:0505003&r=dcm |