Mixture models of choice under risk

This paper is concerned with estimating preference functionals for choice under risk from the choice behaviour of individuals. We note that there is heterogeneity in behaviour between individuals and within individuals. By ‘heterogeneity between individuals’ we mean that people are different, in terms of both their preference functionals and their parameters for these functionals. By ‘heterogeneity within individuals’ we mean that the behaviour may be different even by the same individual for the same choice problem. We propose methods of taking into account all forms of heterogeneity, concentrating particularly on using a Mixture Model to capture the heterogeneity of preference functionals.     

Nonstationary discrete choice: A corrigendum and addendum

We correct the limit theory presented in an earlier paper by Hu and Phillips [2004a. Nonstationary discrete choice. Journal of Econometrics 120, 103–138] for nonstationary time series discrete choice models with multiple choices and thresholds. The new limit theory shows that, in contrast to the binary choice model with nonstationary regressors and a zero threshold where there are dual rates of convergence (n^1/4$n^{1/4}$ and n^3/4$n^{3/4}$), all parameters including the thresholds converge at the rate n^3/4$n^{3/4}$. The presence of nonzero thresholds therefore materially affects rates of convergence. Dual rates of convergence reappear when stationary variables are present in the system. Some simulation evidence is provided, showing how the magnitude of the thresholds affects finite sample performance. A new finding is that predicted probabilities and marginal effect estimates have finite sample distributions that manifest a pile-up, or increasing density, towards the limits of the domain of definition.     

Masking identification of discrete choice models under simulation methods

We present examples based on actual and synthetic datasets to illustrate how simulation methods can mask identification problems in the estimation of discrete choice models such as mixed logit. Simulation methods approximate an integral (without a closed form) by taking draws from the underlying distribution of the random variable of integration. Our examples reveal how a low number of draws can generate estimates that appear identified, but in fact, are either not theoretically identified by the model or not empirically identified by the data. For the particular case of maximum simulated likelihood estimation, we investigate the underlying source of the problem by focusing on the shape of the simulated log-likelihood function under different conditions.     

A Poisson mixture model of discrete choice

In this paper, we introduce a new Poisson mixture model for count panel data where the underlying Poisson process intensity is determined endogenously by consumer latent utility maximization over a set of choice alternatives. This formulation accommodates the choice and count in a single random utility framework with desirable theoretical properties. Individual heterogeneity is introduced through a random coefficient scheme with a flexible semiparametric distribution. We deal with the analytical intractability of the resulting mixture by recasting the model as an embedding of infinite sequences of scaled moments of the mixing distribution, and newly derive their cumulant representations along with bounds on their rate of numerical convergence. We further develop an efficient recursive algorithm for fast evaluation of the model likelihood within a Bayesian Gibbs sampling scheme. We apply our model to a recent household panel of supermarket visit counts. We estimate the nonparametric density of three key variables of interest-price, driving distance, and their interaction-while controlling for a range of consumer demographic characteristics. We use this econometric framework to assess the opportunity cost of time and analyze the interaction between store choice, trip frequency, search intensity, and household and store characteristics. We also conduct a counterfactual welfare experiment and compute the compensating variation for a 10%-30% increase in Walmart prices.     

On specification testing of ordered discrete choice models

We discuss how to test the specification of an ordered discrete choice model against a general alternative. Two main approaches can be followed: tests based on moment conditions and tests based on comparisons between parametric and nonparametric estimations. Following these approaches, various statistics are proposed and their asymptotic properties are discussed. The performance of the statistics is compared by means of simulations. An easy-to-compute variant of the standard moment-based statistic yields the best results in models with a single explanatory variable. In models with various explanatory variables the results are less conclusive, since the relative performance of the statistics depends on both the fit of the model and the type of misspecification that is considered.     

A theory of stochastic choice under uncertainty

In this paper we propose a characterization of stochastic choice under risk and under uncertainty. We presume that decision makers’ actual choices are governed by randomly selected states of mind, and study the representation of decision makers’ perceptions of the stochastic process underlying the selection of their state of mind. The connections of this work to the literatures on random choice, choice behavior when preference are incomplete; choice of menus; and grades of indecisiveness are also discussed.     

Binary choice models with discrete regressors: Identification and misspecification

This paper explores the inferential question in semiparametric binary response models when the continuous support condition is not satisfied and all regressors have discrete support. I focus mainly on the models under the conditional median restriction, as in Manski (1985). I find sharp bounds on the components of the parameter of interest and outline several applications. The formulas for bounds obtained using a recursive procedure help analyze cases where one regressor’s support becomes increasingly dense. Furthermore, I investigate asymptotic properties of estimators of the identification set. I describe a relation between the maximum score estimation and support vector machines and propose several approaches to address the problem of empty identification sets when the model is misspecified.     

Separability of stochastic production decisions from producer risk preferences in the presence of financial markets

Separation results, as they are usually understood, refer to conditions under which a firm’s production decisions are independent of its risk attitudes. Well-understood situations where separation occurs typically include those where technically feasible production opportunities are replicable in financial markets. This paper gives necessary and sufficient conditions for separation that go beyond these well-understood spanning conditions. To do so, we present a unified treatment of the production and financial decisions available to a firm facing frictionless financial markets and a stochastic production technology under minimal assumptions about the firm’s technology and objective function.     

Revealed preference theory on the choice of lotteries

The choice behavior of a decision-maker is said to be consistent with expected utility maximization if there exists a utility function defined on the set of prizes such that the decision-maker chooses lotteries with the highest expected utility. We present a revealed preference characterization of choice behavior that is consistent with expected utility maximization. A necessary and sufficient condition for expected utility maximization is that there does not exist a way to compound lotteries such that the probability distribution over the final prizes generated by the chosen lotteries of each observation is equal to that generated by the rejected lotteries of each observation. Our result is quite general and can be applied to any compact set of prizes and any choice correspondence.     

A Cautionary Note on Estimating the Standard Error of the Gini Index of Inequality: Comment

We observe that care must be taken when using a regression‐based approach to construct a standard error for the Gini coefficient.     

The Hazard Rate of Foreign Direct Investment: A Structural Estimation of a Real‐option Model*

The hazard rate of investment is derived within a real‐option model, and its properties are analysed so as to directly study the relation between uncertainty and investment. Maximum likelihood estimates of the hazard are calculated using a sample of multinational enterprises (MNEs) that invested in Central and Eastern Europe over the period 1990–98. Employing a standard, non‐parametric specification of the hazard, our measure of uncertainty has a negative effect on investment, but the reduced‐form model is unable to control for nonlinearities in the relationship. The structural estimation of the option‐based hazard is instead able to account for the nonlinearities and exhibits a significant value of waiting, although the latter is independent of our measure of uncertainty. This finding supports the existence of alternative channels through which uncertainty can affect investment.     

Nonparametric Bayesian modelling of monotone preferences for discrete choice experiments

Discrete choice experiments are widely used to learn about the distribution of individual preferences for product attributes. Such experiments are often designed and conducted deliberately for the purpose of designing new products. There is a long-standing literature on nonparametric and Bayesian modelling of preferences for the study of consumer choice when there is a market for each product, but this work does not apply when such markets fail to exist as is the case with most product attributes. This paper takes up the common case in which attributes can be quantified and preferences over these attributes are monotone. It shows that monotonicity is the only shape constraint appropriate for a utility function in these circumstances. The paper models components of utility using a Dirichlet prior distribution and demonstrates that all monotone nondecreasing utility functions are supported by the prior. It develops a Markov chain Monte Carlo algorithm for posterior simulation that is reliable and practical given the number of attributes, choices and sample sizes characteristic of discrete choice experiments. The paper uses the algorithm to demonstrate the flexibility of the model in capturing heterogeneous preferences and applies it to a discrete choice experiment that elicits preferences for different auto insurance policies.     

Global identification of risk preferences with revealed preference data

The concept of parameter identification (for a given specification) is differentiated from global identification (which specification is right). First-order conditions for production under risk are shown to admit many alternative specification pairs representing risk preferences and either perceived price risk, production risk, or the deterministic production structure. Imposing an arbitrary specification on any of the latter three determines which risk preference specification fits a given dataset, undermining global identification even when parameter identification is suggested by typical statistics. This lack of identification is not relaxed by increasing the number of observations. Critical implications for estimation of mean-variance specifications are derived.     

Habits,Complementarities and Heterogeneity in Alcohol and Tobacco Demand: A Multivariate Dynamic Model*

In this paper we test the existence of forward‐looking behaviour in a multivariate model for alcohol and tobacco consumption. The theoretical framework, based on a dynamic adjustment cost model with forward‐looking behaviour, is enhanced to include the intertemporal interactions between the two goods. The analysis of the within‐period preferences completes the intertemporal model, allowing to evaluate the static substitutability/complementarity relationships. The empirical strategy consists in a two‐step estimation procedure. In a first stage, we obtain the parameters of the demand system, while in a second stage Euler equations are estimated. Results, based on a cohort data set constructed from a series of cross‐sections of the Italian Household Budget Survey, reveal a significant complementarity relationship between alcohol and tobacco. Estimation of the Euler equations does not lead to rejection of the hypothesis of intertemporal dependence, providing evidence for a forward‐looking behaviour in alcohol and tobacco consumption. Moreover, we find significant intertemporal interactions that support the adjustment cost setting in a multivariate model with rational expectations.     

Complete monotonicity of the representative consumer’s discount factor

A univariate real-valued function is said to be completely monotone if it takes positive values and alternate the signs of its higher order derivatives, starting from everywhere negative first derivatives. We prove that the representative consumer’s discount factor of a continuous-time economy under uncertainty is a power function of some completely monotone function of time satisfying certain boundary conditions if and only if it may be derived from a group of consumers having constant and equal relative risk aversion, and constant and yet possibly unequal discount rates.     

The effects of information about health hazards in food on consumers’ choice process

This study examines the effects of context (health hazard), direction (positive versus negative) and intensity of information about health hazards on consumers’ choice processes. We propose that choice of frequently purchased food commodities, ceteris paribus, is based on a single dimension—taste. We develop a set of hypotheses regarding the type of choice process to be employed in various information types and empirically test them in a field experiment design. Our results indicate that a single-dimension choice process is employed under a nonsevere message and a multidimensional process under high-intensity negative information.     

Robustness of Logit Analysis: Unobserved Heterogeneity and Mis‐specified Disturbances*

In probit and logit models, the β coefficients vary inversely with the variance of the disturbances. The omission of a relevant orthogonal regressor leads to increased unobserved heterogeneity, and this depresses the β coefficients of the remaining regressors towards zero. For the probit model, Wooldridge (Econometric Analysis of Cross Section and Panel Data, MIT Press, Cambridge, MA, 2002) has shown that this bias does not carry over to the effect of these regressors on the outcome. We find by simulations that this also holds for the logit model, even when omitting a variable leads to severe mis‐specification of the disturbance. More simulations show that logit analysis is quite insensitive to pure mis‐specification of the disturbance as such.     

On the equivalence of location choice models: Conditional logit, nested logit and Poisson

It is well understood that the two most popular empirical models of location choice - conditional logit and Poisson - return identical coefficient estimates when the regressors are not individual specific. We show that these two models differ starkly in terms of their implied predictions. The conditional logit model represents a zero-sum world, in which one region’s gain is the other regions’ loss. In contrast, the Poisson model implies a positive-sum economy, in which one region’s gain is no other region’s loss. We also show that all intermediate cases can be represented as a nested logit model with a single outside option. The nested logit turns out to be a linear combination of the conditional logit and Poisson models. Conditional logit and Poisson elasticities mark the polar cases and can therefore serve as boundary values in applied research.     

A Note on Additive Separability and Latent Index Models of Binary Choice: Representation Results*

The standard binary choice model in econometrics has the choice determined by a latent index crossing a threshold. The latent index is almost always assumed to be additively separable in observable and unobservable regressors, and most commonly linear in all regressors. This note provides a class of non‐separable latent index functions which will have equivalent representations as additively separable or linear index functions. These results demonstrate that assuming a linear or additively separable latent index function is less restrictive than previously recognized.