Abstract: The authors present the distribution of the maximum random utility in the GEV model and that of the maximum random utilities within the separable subsets in the GEV model with an additively separable generator function. They then formulate the inclusive value of the entire choice set in the GEV model and that of a separable subset of alternatives in the GEV model with an additively separable generator function and relate them to the consumer's surplus.

Abstract: This study discusses, from the Bayesian standpoint, whether the comparison or choice between two competing models depends on the characteristics of a third model lurking in the background. In the qualitative choice literature, an obvious analog is the independence of irrelevant alternatives property. Usually, Bayesian posterior odds analysis satisfies a similar property but standard Bayesian model selection does not. The analysis also considers a random utility model in which the utility function of the reader is only partially known to the statistician. The implications of this framework for estimation and prediction are explored.

Abstract: Price discrimination in monopolistically competitive markets affects firms' joint profits through several pecuniary and nonpecuniary externalities. Discrimination is a public good if the net effect is positive. Using a random utility shopping destination choice model we investigate the effect of a downtown parking coupon program that discriminates in favor of suburban consumers and against consumers based downtown. The program appears profitable for downtown stores collectively, but in the noncooperative Nash equilibrium stores do not participate. Participation is thus subject to free-riding. As the subsidy rate required to induce participation rises, profits fall. Whatever the subsidy rate, social surplus declines.

Abstract: Random utility theory offers a powerful modeling tool for group decision, voting and persuasion. I combine and extend three recent developments: (1) probabilistic choice models of approval voting, (2) general random utility representations for a variety of order relations, and (3) stochastic theories of persuasion. The present topset voting model merges (1) and (2) by generalizing probabilistic choice models for approval voting to arbitrary binary preference relations. Whereas the econometric literature emphasizes parametric random utilities (e.g. generalized logit and probit), I investigate parametrizations of the probabilistic relations. A new model for paneled subset choice combines (1) and (3).

Abstract: A stated preference approach is used to predict likely responses to market reform in Zimbabwe. A random utility model is estimated using data on consumers' stated preferences for alternative types of maize meal. The model is used to predict demand patterns resulting from market reforms. Predicted demand is then compared with actual demand following market reforms. An analysis of the welfare effects of alternative policy scenarios shows that coupling the removal of consumer maize subsidies with improved access to a broader range of maize meal products ameliorated many of the adverse effects of subsidy removal, especially for lower-income groups.

Abstract: A new estimator links migration data to a random utility 'voting with your feet' model to compare the relative living standards of pairs of regions. It is argued that this estimator has a firmer theoretical basis and uses migration information more efficiently than previous methods. An algorithm converts pairwise comparisons into rankings of the U.S. states for 1970, 1980, and 1990. The rankings indicate living standards were highest in the Northwest (1970, 1980) and the south Atlantic coast (1990). A nonparametric test suggests that the system was in disequilibrium in 1980 (probably due to energy price shocks), but near equilibrium in 1970 and 1990.

Abstract: The authors adapt the theoretical state-preference model to value nonmarket public goods under individual uncertainty about use, illustrating with an assessment of willingness-to-pay to prevent acid rain lake damage in the northeast United States. Individual usage uncertainty is modeled via probabilities of participation in trout fishing. Changes in environmental quality are valued using a random utility model to explain yes/no responses to a contingent valuation question. The authors produce quantitative welfare measures: individual fitted and simulated passive- and active-use values, individual expected consumer surplus, option price, option value, and complete individual willingness-to-pay loci.

Abstract: A new choice forecasting model is developed here to answer a practical question: how to forecast housing demand at a disaggregate level. Being different from the previous housing choice models, this model is derived from housing sub-market structure based on a random utility approach. Housing choice is thought to be a joint choice of all the components associated with a dwelling. These components create a huge bundle of dwelling alternatives resulting in an empirical calculation problem. To avoid this problem, the model developed in this paper separates the joint choice behaviour into two stages: the choices of the key dwelling components which construct the housing sub-markets and the choices of the non-key dwelling components which distinguish individual dwellings in each housing sub-market. Some policy simulations are presented based on this model.

Abstract: A new approach is proposed for calculating the expected market share. It is assumed that consumers patronize a facility according to a utility function, selecting the facility with the highest utility value. However, consumers' ratings of the utility components are stochastic by some random distribution. Therefore, the buying power of customers located at the same point is divided among several facilities. A probability that a consumer patronizes a certain facility can be calculated. Consequently, the expected market share by competing facilities can be estimated. This calculation is more than 1,000 times faster than repeating a simulation enough times to achieve a reasonable accuracy. The distance decay calculated using the new approach is approximately exponential. A procedure for finding the optimal location anywhere in the plane for a new facility that maximizes the market share is also introduced.

Abstract: This paper examines faculty time allocation decisions that are fundamental to the functioning of a university. A random-utility approach yields a grouped-data, multinominal logit model and predicts that time allocation decisions depend systematically on both personal and institutional attributes. The empirical results for a random sample of U.S. arts and sciences faculty indicate that structural differences between universities with different research orientations account for most of the significant differences in faculty time allocations. Faculty characteristics reinforce institutional missions, however, and thus condition university policies for change- for example, attempts to mandate greater time to teaching in research universities.

Abstract: In this paper we develop a general class of dynamic brand choice models, called LIGHTNING BOLT (LB) models, which are consistent with the theory of random utility maximization of consumer choice behavior. The underlying random utility process is Markov, and the inter-temporal evolution of the (utility maximizing) brand choice process is also Markov. The model permits parsimonious parameterizations of the random utility process in brand choices with the resulting switching probabilities being functionally related to explanatory variables. The model allows for structural state dependence (feedback), habit persistence (inertia), and unobserved heterogeneity. The theoretical development shows that several well known stochastic brand choice models can be deducted from random utility maximization theory. From a managerial perspective, the usefulness of the proposed model stems from its ability to separate out the effects of habits, state dependence and heterogeneity. Strong state dependence effects imply incentives for inducing trial of a brand (e.g., product sampling).

Abstract: The value of choice-freedom is a measurement of the variety of choice of an individual and is defined as the maximum amount a consumer would be willing to pay for the option which is infrequently or not at all used by the consumer. This paper shows that the value of choice-freedom can be derived from random utility theory and measured by an entropy defined by choice probabilities. In addition, the interrelationships among choice-freedom, average utility, and indirect utility is unified into only one equation, called the choice equation, and its properties are examined.

Abstract: Severe declines in Pacific Northwest salmon stocks, coupled with increasing recreational demands, and judicial decisions supporting Native American fishing rights create challenges for fishery agencies. This article explores the welfare effects on recreational anglers of alternative salmon allocation policies to meet Native American treaty rights. A discrete choice random utility model, coupled with a poisson trip frequency model, is used to analyze these welfare effects. The model is fit to survey data from the Willamette River spring chinook fishery, an important recreational fishery in Oregon. Management options have dramatically different welfare effects.

Abstract: Random Utility Models are commonly used the recreation demand literature to model the choice among a set of qualitatively different recreation sites. Often, due to computational or data constraints, analysts do not include all possible alternatives in the choice set. This research is concerned with choice set boundaries. In particular, we ask, "how far away from a person's home must a site be before it is no longer considered part of the choice set in estimation?" We examine this question by estimating a single model with varying choice set boundaries using a data set of recreational fishing activity in Maine. In this analysis, we show that there exists some threshold distance beyond which adding additional sites to the choice set has negligible effects on the estimation results.

Abstract: This essay presents a measure-theoretic version of the random utility model with no substantive restrictions upon the choice space. The analysis is based upon DeFinetti's Coherency Axiom, which characterizes a set function as a finitely additive probability measure. The central result is the equivalence of the random utility maximization hypothesis and the coherency of the choice probabilities over all allowable constraint sets.

Abstract: This paper presents a theory describing the emergence and the evolution of preference relations over (real) time. The subject or consumer is regarded as immersed in a random environment of the Poisson type. Events arise at random times, which consist of tokens of information about the alternatives. Some of these tokens have the potential to create or modify some edges of the graph of the preference relation. General axioms are given that lead to a stochastic process of Markovian character. In a special case of these axioms, a random utility model is obtained as an asymptotic result. Specifically, it is shown that the Markov process converges to a random walk on a unique ergodic set composed of all the rankings of the set of objects. Precise asymptotic results are derived and the explicit expression for the probability distribution on the set of rankings is obtained. In another special case, the asymptotic random walk is on the set of all weak orders. It is also shown how the theory can account for the resemblance between the preference relations obtained, from the same subject, at time "t" and "t" + sigma.

Abstract: In this paper, the authors describe a model using a household production framework to link measures of nonpoint source pollution to fishing quality and a random utility model to describe how that quality influences sport fishing parties' decisions in North Carolina. The results provide clear support for using a model that evaluates the effects of pollution on the activities and decisions associated with the fishing activity once a trip is taken. Site selection decisions are then conditioned on the anticipated quality of fishing sites. The framework also has the advantage of linking the spatial, technical, and economic information required to evaluate the management plans required for estuaries under the National Estuarine Program.

Abstract: Fourteen papers written in honor of Amartya Sen, Lamont University Professor and Professor of Economics and of Philosophy at Harvard University, and his contributions to individual and social choice theory and development. Papers focus on freedom and flexibility, capabilities, exclusion, and the supply of goods; existence, efficiency, and remedial policy; rationalizing choice probabilities using the random utility model; private information and the shape of the redistribution frontier; majority rule, social welfare functions, and game forms; symbolic utility; valued opinions or opinionated values; rational fools and cooperation in a poor hydraulic economy; primary education and economic development in China and India; children and intrahousehold inequality; the entitlement approach to famine; differing fortunes of the poor in Bangladesh and Indonesia; and mass unemployment as a social problem. Includes biographical and bibliographical information on Amartya Sen. Index.

Abstract: The budget-constrained random utility model (RUM) gives utility-consistent measures of welfare, but requires the length of the planning period be specified. An alternative is to treat the RUM probabilities as behavioral and calculate consumer surplus. This paper shows that such calculations lead to the same welfare measures as RUM calculations. The paper provides support and an alternative justification for the standard welfare measured.

Abstract: We present a random utility model of recreation site choice that incorporates an aggregate demand function for number of trips during a season. We derive the trip demand function using conventional demand theory and use it to calculate seasonal welfare changes due to improvements in site characteristics or addition of new sites. The model is based on Bockstael, et al.'s participation function.

Abstract: This paper discusses and evaluates three issues to which the analyst has not paid much attention when describing recreational site choices and implementing the nested random utility model (RUM) for recreation benefit measurement: (1) the effects of party composition on recreational decisions; (2) the implications of estimating a nested RUM on the functional structure of the underlying indirect utility function; and (3) the variation in benefit estimates across quality improvement scenarios. A three-level nested RUM is estimated and the benefits of water quality improvements are measured for marine recreational fishing in the Albemarle-Pamlico Estuary of North Carolina.

Abstract: This paper presents a Poisson model of expected angler catch during a sportfishing trip and employs the expected catch in a random utility model of site choice. The approach permits greater heterogeneity in expected catch and in individual welfare estimates from policies such as creel limits.

Abstract: Measurement of nonmarket values often involves subjective judgments. Since these judgments may influence results, they should be carefully considered. The author focuses on an aspect of subjective choice relating to the estimation of random utility models. Such models require specification of each recreationalist's choice set. Whether an individual perceives his choice set as composed of all possible alternatives, a few popular alternatives, or collections of spatially aggregated alternatives is an important judgment affecting the conclusions.

Abstract: A new approach is developed for integrating consumers' risk perceptions with stated purchase behavior when consumption decisions must be made with incomplete information. The application involves health risks from exposure to pesticide residues on fresh produce. Unlike traditional food demand analysis, the present approach treats produce choices as discrete outcomes, resulting in a random utility model. Empirical results from a pilot survey suggest a clear linkage between perceptions and behavior in response to new risk information. Consumers' stated preferences for safer produce were primarily influenced by price differences and perceived risks, not by the technical risk information provided alone. However, the linkage between behavior and valuation was less clear cut. The risk/price tradeoffs entailed by contingent discrete choices indicate high price premia for small risk reductions and little variation in price premium across alternative risk reductions.

Abstract: This study determines the likely effect of cost-share incentives on participation in the Tennessee Forest Stewardship Program and identifies other factors that may contribute to participation. A random utility model is used to determine the probability that a landowner will choose to participate in the program. A binary choice model is specified to represent the dichotomous decision and a logit procedure is used to fit the model. Data are obtained from mail surveys of 4,000 randomly selected landowners. Results indicate that attitudes and knowledge of forestry programs may be more influential in a landowner's decision to participate than monetary incentives.

Abstract: The generalized extreme value model was developed by D. McFadden for the case with discrete choice sets. The present paper extends this model to cases with both discrete and continuous choice sets and choice sets that are unobservable by the analyst. The author also proposes behavioral assumptions that justify random utility functions (processes) that have a max-stable structure, i.e., utility processes where the finite dimensional distributions are of the multivariate extreme value type. Finally, he derives nonparametrically testable implications for the choice probabilities in the continuous case.

Abstract: A stated preference model and a revealed preference model for recreational site choice are examined and compared. Both models are based on random utility theory and the data are obtained from the same individuals. The stated preference model is based on the respondent's choice from hypothetical choice sets. Attributes in the stated preference model are based on the ranges of the actual levels of attributes in the revealed preference choice set and are presented to respondents using a fractional factorial statistical design. The results show that while independently estimated models appear to reflect different underlying preferences, joint estimation of the model parameters, including estimation of the relative scale parameter, provides evidence that the underlying preferences are in fact similar. Furthermore, combining the revealed and stated preference information yields other benefits in estimation.

Abstract: This paper develops an individual level choice model which embeds a rational choice set generation process. The choice set generation process is motivated by a goal derived categorization paradigm where consumers make uncertain judgments about the ability of brands to meet the consumption or usage goal motivating the choice decision. Recognizing this uncertainty, brand selection is modeled as a decision under uncertainty within a random utility framework. The modeling of the process builds on previously suggested formalizations based on search theory and decision costs. Derivations are provided for a flexible and parsimonious representation of the choice probabilities which are characterized neither by IIA nor the regularity property. Some initial empirical results using scanner data are reported.

Abstract: Random utility models are widely applied in studies of recreation demand. The model is particularly useful when the number of recreation sites from which individuals may choose is large. Yet, when the number gets too large, say in the hundreds, estimation becomes burdensome. The authors present an analysis suggested by Daniel McFadden (1978) for dealing with large numbers of sites. They estimate a model using randomly drawn opportunity sets. The authors use each person's chosen site plus a random draw of as few as eleven other sites (when hundreds are available) to estimate a plausible behavioral model.

Abstract: Random utility models are commonly used to model the choice among a set of alternatives. Often, due to data or computational constraints, the analyst must use aggregated alternatives to estimate the model. These aggregates are defined by averaging characteristics of alternatives over prespecified groups. In this analysis, the authors demonstrate that unless some very restrictive conditions hold, the use of aggregated alternatives will lead to biased results. Then, using a data set of recreational fishing in Wisconsin, they show the effects that this bias can have on the results estimated from the model.

Abstract: W. Alonso's (1964) bid-rent theory and the discrete-choice random-utility theory appear in the literature as well-established alternative frameworks to model urban land use. As both approaches share the support of microeconomic theory, the main issue addressed in this paper is the theoretical comparison of the two approaches. It is demonstrated that in perfectly competitive land markets these approaches are equivalent, therefore they should be understood as complementary rather than alternative. The case of markets subject to speculative land prices is then explored for the cases of speculative supply and/or speculative demand, with the discovery that both approaches are theoretically equivalent in every case studied, thus extending the previous conclusion for the general case. These conclusions provide the base for an integrated and more comprehensive urban economic theory and for the bid-choice land-use model.

Abstract: The paper introduces the notion of a probabilistic mixture of a family of relational structures. This concept is of interest both from a measurement-theoretic and from a choice-theoretic point of view, since it offers a probabilistic generalization of classical deterministic measurement concepts (such as additive conjoint measurement) and, at the same time, a natural generalization for binary choice systems induced by rankings. Probabilistic mixtures with finite domains are shown to be characterizable in terms of finite systems of linear equations and inequalities and to be closely related to random-utility-like representation concepts.

Abstract: In this paper, an individual level choice model is developed for the utility maximizing consumer which explicitly recognizes the use of heuristics to confine choice to a subset of available alternatives. A rational choice set formation process is integrated probabilistically within a random utility framework. Conceptually, the model integrates previously suggested formalizations based on search theory and decision costs. Operationally, the model relies on the random utility assumptions of the multinominal logit model.

Abstract: In conventional random-utility models, such as the multinomial logit model, it is assumed that a decisionmaker's choice set is independent of his or her preferences conditional on the explanatory variables of the models. However, there are many situations in which the decisionmaker chooses the choice set, thereby making the independence assumption implausible. For example, if information about alternatives is costly, an individual may choose to learn in detail about only a small group of them. This paper is concerned with the modeling of discrete choice with endogenous choice sets. It is assumed that costly information forces the decisionmaker to acquire detailed knowledge of only a small group of alternatives that she or he selects. Models are developed for choice-set generation in this context and for choice conditional on endogenous choice sets.

Abstract: Probabilistic choice systems in the generalized extreme value (GEV) family embody two restrictions not shared by the covariance probit model. First, the unobserved components of random utility are homoskedastic across individuals and alternatives. Second, the degree of similarity among alternatives is also assumed to be constant across individuals. This paper considers extensions to models in the GEV class which relax these two restrictions. An empirical application concerning the demand for cameras is developed to demonstrate the potential significance of the heterogeneous logit model.

Abstract: In the literature of random utility maximization models, the preference for variety has been handled mainly by assumption of a stochastic choice parameter. The present model shows that a stochastic choice parameter can be determined in a model in which an agent is shown to have a deterministic preference for variety, as well as stable preferences for physical attributes of consumption choices.

Abstract: Rotondo has developed a constant utility generalization of Luce's choice model that can handle various examples that contradict the choice model. This paper axiomatizes a generalization of Rotondo's model and discusses unidimensional and multidimensional versions of that generalization; several additional incomplete characterizations of the model are also presented. The class of random advantage models, which generalizes the classical random utility family, is presented and related to the generalized Rotondo model.

Abstract: There has been substantial increase in the use of random utility models (RUM) to estimate the values people would place on reducing marine pollution. This paper identifies and evaluates the importance of four methodological issues that arise in using the RUM framework for estimating the effects of estuarine quality on the choice of recreation sites and the benefits from reducing pollution. The issues include: (a) specifying the the set of alternatives to be indicated in the choice set, (b) evaluating how the definition of what constitutes an elemental alternative affects the plausibility of the independence of irrelevant alternatives; (c) examining the effects of the definition of alternatives for benefit measures; and (d) evaluating the implications of multiple pollutants for describing the factors influencing people's recreation site choices.

Abstract: In 1979, Bishop and Heberlein introduced an appealing variant of the contingent valuation method which required only yes or no responses (1979). Hanemann developed a utility-theoretic interpretation of the yes/no responses which has helped popularize the approach (1984). Recently Cameron has offered another interpretation, also utility-theoretic (1987), which she argues is more general than the Hanemann random utility model. This paper compares the deterministic models suggested by Hamemann and Cameron, showing them to be dual to each other. It shows that correct specification of either function must exclude endogenous variables such as quantity demanded. It also demonstrates that using utility-theoretic models allows one to compare travel cost models and contingent valuation models on the basis of their implied behavior, not on the unobservable surplus or variation measure.

Abstract: Consumer surplus estimates are random variables. While they are generally recognized as stochastic, little attention was given to their properties prior to Bockstael and Strand's (1987) evaluation of conventional practices for using recreation demand models in benefit measurement. Their paper, as well as all the research it stimulated, adopted a similar strategy, namely to judge the methods for estimating demand or random utility models based on the properties of their respective consumer surplus estimates. This paper proposes a different strategy, to define estimators based on the properties of their implied consumer surplus estimates.

Abstract: The gray literature in the field of nonmarket benefit measurement has made extensive use of the random utility (or discrete choice) model in recent years, but few applications appear in the literature. This article provides such an application, illustrating the technique with preliminary results from a regional study modeling east cost sportfishing behavior. The article discusses some of the strengths and weaknesses of the random utility model. It also illustrates how data regularly collected by the National Marine Fisheries Service can be supplemented with economic survey data to estimate these discrete choice behavioral models.

Abstract: The question of model identification is analyzed for the semiparametric random utility model of discrete choice. Attention is focused on settings where agents face a common choice between a set of J+1 alternatives, but where actual choices are only partially observed. Necessary conditions are derived for the setting where the only data on actual choices consists of a binary indicator for one of the alternatives. Sufficient conditions are developed in this setting for a linear in parameters specification of indirect utility.

Abstract: In this paper, the authors examine the consequences of adopting the random utility hypothesis as an approach for randomizing a system of demand equations. Random utility models are appealing since they allow the usual assumption of deterministic utility-maximizing behavior by each consumer to coexist with the apparent randomness across individuals that is exhibited by data. Their results show that the use of random utility models implies that the disturbances of the demand equations may not be homoskedastic, but must be functions of prices and/or income.

Abstract: It is argued that the statistical basis of the analysis of stated-preference data is rather underresearched. When stated-preference data is used to pr ovide relative evaluations of attributes, the implications are probab ly not serious; however, the forecasting issues are more important. T hese issues, relating to model specification and different sources of error, are made explicit in relation to the random-utility theory co mmonly used with revealed-preference models. It is concluded that the use of more complex estimation methods should be encouraged, and tha t there remains a need for "external" data on real choices if credi ble forecasts are to be produced.

Abstract: A simultaneous discrete and continuous random utility model for housing demand is presented. Household utility depends on three characteristics of hou sing: a discrete number of bedrooms and continuous measures of hous-i ng quality and distance to the central business district. Empirical r esults for a sample of renter households from San Francisco are compa red for a univariate structure in which each choice variable has its own utility or disutility function, a simultaneous structure assuming a discrete-continuous analog of the independence of irrelevant alter natives, and one for which this assumption is relaxed. The results in dicate that the third structure is most appropriate.

Abstract: Choice probabilities generated by D. McFadden's (1973) random utility model are both multinomial logit and have the Independence of Irrelevant Alternatives (IIA) property. Various authors have suggested specification tests of this property. This paper identifies the alternative against which these tests are constructed and obtains classical specification tests. It is shown that both the null hypothesis and local asymptotic power of the classical tests is the same as for a proposed Hausman test. Finally the discussion throws some light on the nature of the IIA property and on the performance of the logit model in applications where one would expect the IIA property to be violated but where the alternative set is fixed.

Abstract: This paper develops a semiparametric method for estimating the nonrandom part V(.) of a random utility function U(v, omega) = V(v) + e(omega) from data on discrete choice behavior. Here v and omega are, respectively, vectors of observable and unobservable attributes of an alternative, and e(omega) is the random part of the utility for that alternative. The method is semiparametric because it assumes that the distribution of the random parts is known up to a finite dimensional parameter theta, while not requiring specification of a parametric form for V(.). The nonstochastic part V(.) of the utility function U(.) is assumed to be Lipschitzian and to possess a set of properties, typically assumed for utility functions. The estimator of the pair (V, theta) is shown to be strongly consistent.

Abstract: A generalization of the multinomial logit (MNL) model is developed for cases where discrete alternatives are ordered, by allowing stochastic correlation among alternatives in close proximity. The model belongs to the Generalized Extreme Value class and is therefore consistent with random utility maximization. An extension can handle cases where observations have been selected from a truncated choice set. A two-stage procedure using MNL computer software provides a specification test for MNL against the proposed model. Two empirical applications are briefly described.