Model selection and misspecification in discrete choice welfare analysis

This study extends the work by Herriges and Kling (1997) to further evaluate the impact of discrete choice modelling techniques on welfare measures. Particularly, we evaluate the performance of the increasingly popular mixed logit model and the computational strategy for deriving discrete choice welfare measures. Our simulation results show that model misspecification can have profound effects on welfare measures. In general, the flexible mixed logit model performs relatively well in the presence of misspecification. However, when the nesting structure can be appropriately identified (via statistical tests and a priori knowledge/experience), the nested logit model provides more reliable welfare measures than the mixed logit model.     

Understanding Heterogeneous Preferences in Random Utility Models: A Latent Class Approach

A finite mixture approach toconditional logit models is developed in whichlatent classes are used to promoteunderstanding of systematic heterogeneity. The model is applied to wilderness recreationin which a branded choice experiment involvingchoice of one park from a demand system wasadministered to a sample of recreationists. The basis of membership in the classes orsegments in the sample involved attitudinalmeasures of motivations for taking a trip, aswell as their stated preferences overwilderness park attributes. The econometricanalysis suggested that four classes of peopleexist in the sample. Using the model toexamine welfare measures of some hypotheticalpolicy changes identified markedly differentwelfare effects than the standard singlesegment model, and provided insight into thedifferential impact of alternative policies.     

Quantifying landscape externalities of renewable energy development: Implications of attribute cut-offs in choice experiments

Renewable energy is worldwide seen as a key element necessary to address climate change. However, finding socially acceptable locations for renewable energy facilities and the accompanying infrastructure increasingly often faces fierce opposition. This paper quantifies the landscape externalities of renewable energies employing a choice experiment. In addition, it is investigated how accounting for non-compensatory choice behavior, i.e. attribute cut-offs, affects welfare measures and subsequently policy recommendations. The empirical application is Germany where we conducted a nationwide survey on the development of renewable energies. We first show that cut-off elicitation questions prior to the choice experiment at least partially influence preferences. We further find that most participants state cut-off levels for attributes. Many are, however, at the same time willing to violate the self-imposed thresholds when choosing among the alternatives. To account for this effect, stated cut-offs are incorporated into a mixed logit model following the soft cut-off approach. Model results indicate substantial taste heterogeneity in preferences and in the use of cutoffs. Also, welfare estimates are substantially affected. We conclude that welfare changes from renewable energy development could be strongly underestimated when cut-offs are ignored.     

Mixed Logit Model Performance and Distributional Assumptions: Preferences and GM foods

Using mixed logit models to analyse choice data is common but requires ex ante specification of the functional forms of preference distributions. We make the case for greater use of bounded functional forms and propose the use of the Marginal Likelihood, calculated using Bayesian techniques, as a single measure of model performance across non nested mixed logit specifications. Using this measure leads to very different rankings of model specifications compared to alternative rule of thumb measures. The approach is illustrated using data from a choice experiment regarding GM food types which provides insights regarding the recent WTO dispute between the EU and the US, Canada and Argentina and whether labelling and trade regimes should be based on the production process or product composition. This paper draws work commissioned by UK Department of Environment Food and Rural Affairs (DEFRA). The views presented in this paper are those of the authors alone and should not be regarded as those of DEFRA or of individuals within DEFRA.     

Alternative approaches to discrete working time choice in an AGE framework

We compare two options of integrating discrete working time choice of heterogenous households into a general equilibrium model. The first, known from the literature, produces household heterogeneity through a working time preference parameter. We contrast this with a model that directly incorporates a logit discrete-choice approach into an AGE framework. On the grounds of both calibration consistency and adequate accommodation of within-household interaction, we argue that the logit approach is preferable.     

Environmental diversity in recreational choice modelling

The accuracy of environmental valuation studies relies, to a great extent, on the suitability of the proxy measures used to capture individuals' preferences. While important advances have been achieved in the last years concerning the characterization of the physical background in which recreational choices are made, Travel Cost Method applications have failed to consider the heterogeneity of landscape and the spatial configuration of land use. This paper presents an empirical application to forest recreation in Mallorca (Spain), implementing a random parameter logit model to evaluate in terms of goodness-of-fit, model predictions and welfare measurements the effects of environmental diversity on the recreational site-choice process.     

On the Natural and Economic Difficulties to Fulfilling the Human Right to Water Within a Neoclassical Economics Framework

We present a neoclassical economic model of the human right to water using a nonrenewable resource model inclusive of a backstop technology. The right is interpreted as a minimum consumption requirement the government is obligated to fulfill in the event that any one household cannot do so independently. Differing by income levels, households maximize utility by purchasing a composite consumption good and water from two distinct, government-owned sources. Facing physical and financial constraints, the government uses fiscal policy to address potential human rights violations. Reducing the analysis to two periods, we develop a novel approach to compare total welfare levels from a joint human rights and neoclassical economics perspective. We define a human rights welfare standard and discuss cases in which traditional social welfare measures would exceed, violate, or meet this standard. We thus offer a unique way to merge economic analysis with human rights research.     

Recreation demand estimation and valuation in spatially connected systems

Recreation demand models frequently are used to explain outdoor recreation behavior and to estimate willingness to pay for changes in environmental quality at recreation sites. Among the most commonly used recreation demand models are site choice models based on the multinomial logit framework, which account for the spatial relationships between each recreator's home and multiple alternative destinations thereby capturing the substitution possibilities among recreation sites. However, standard applications of this framework typically do not account for the possibility of spatial connections among the sites via movements of the target species, such as fish in connected water bodies in recreational angling applications or terrestrial species in hunting or wildlife viewing applications. In this paper we examine aspects of environmental valuation and natural resource dynamics that generally are addressed separately. Specifically, we show that in such spatially connected systems, a “reduced form” application of the standard site choice modeling approach, using proxy measures of environmental quality rather than direct measures of species abundances, can produce biased estimates of willingness to pay for environmental improvements. Furthermore, we show that under some conditions poorly targeted environmental improvements in spatially connected systems can lead to welfare decreases. In such systems a structural model of recreator site choices and species sorting behavior and population dynamics may be required to fully account for the spatial linkages among sites and the feedback effects between recreators and the target species.     

GMM estimation of a structural demand model for yogurt and the effects of the introduction of new brands

The demand structure for yogurt is assumed to be properly described by a one level nested logit model that is applied to aggregate market data. Given the presence of endogenous regressors, suitably lagged endogenous variables (Arellano and Bover in J Econom 68:29–51, 1995; Blundell and Bond in J Econom 87:115–143, 1998) are proposed as instrumental variables. The validity of this set of instruments is discussed and price elasticities and marginal costs are recovered from the demand estimates. Total welfare gains associated to the introduction of two new brands by the same manufacturer are finally computed. Prices and profits decreased and total welfare increased.     

Combining Discrete and Continuous Representations of Preference Heterogeneity: A Latent Class Approach

Unobserved preference heterogeneity has been widely recognized as a critical issue not only for modelling choice behaviour, but also for policy analysis. This paper examines alternative approaches for incorporating heterogeneity in recreational demand. We apply a hybrid model combining discrete and continuous heterogeneity representations of tastes to capture the defining features of both the latent class and the random parameter logit specifications. This model allows for the joint estimation of discrete segments and within segment heterogeneity providing a richer interpretation of preference heterogeneity. A database of recreational trips to forest sites in Mallorca has been used to compare the empirical performance of this hybrid approach with common specifications such as the conditional logit, the random parameter logit, and the latent class model.     

Modelling the Effect of Chronic Wasting Disease on Recreational Hunting Site Choice Preferences and Choice Set Formation over Time

Chronic wasting disease (CWD) is a prion disease that affects deer, elk and other cervid wildlife species. Although there is no known link between the consumption of CWD affected meat and human health, hunters are advised to have animals from CWD affected areas tested and are advised against consuming meat from CWD infected animals (Government of Alberta 2010). We model hunter response to the knowledge that deer in a wildlife management unit have been found to have CWD in Alberta, Canada. We examine hunter site choice over two hunting seasons using revealed and stated preference data in models that incorporate preferences, choice set formation, and scale. We compare a fully endogenous choice set model using the independent availability logit model (Swait in Probabilistic choice set formation in transportation demand models. Dissertation, MIT, 1984) with the availability function approach (Cascetta and Papola in Transp Res C 9(4):249–263, 2001) that approximates choice set formation. We find that CWD incidence affects choice set formation and preferences and that ignoring choice set formation would result in biased estimates of impact and welfare measures. This study contributes to the broader recreation demand literature by incorporating choice set formation, scale and temporal impacts into a random utility model of recreation demand.     

Up the Proverbial Creek without a Paddle: Accounting for Variable Participant Skill Levels in Recreational Demand Modelling

We adapt the standard random utility model to take account of the heterogeneity of recreational preferences by using what we call a “skilled-based conditional logit framework”. By separating out our sample of whitewater kayakers into two exogenously identifiable groups (based on their skill level) and running separate conditional logits for each group we are able to take account of the fact that kayakers of different skill levels are looking for different characteristics from the whitewater site they choose to visit. We find that not taking into account the differences in the skill of the kayakers and the grade of the river will result in an overestimation of the welfare estimates associated with improvements to lower grade whitewater sites (which are frequented by basic/intermediated proficiency level kayakers) and underestimating welfare estimates associated with changes in the attributes of higher grade whitewater sites (which are frequented by advanced proficiency level kayakers).      

Do environmental benefits matter? Evidence from a choice experiment among house owners in Germany

Residential buildings strongly contribute to global CO₂ emissions due to the high energy demand for electricity and heating, particularly in industrialised countries. Within the EU, decentralised heat generation is of particular relevance for future climate policy, as its emissions are not covered by the EU ETS. We conducted a choice experiment concerning energy retrofits for existing houses in Germany. In the experiment, the approximately 400 sampled house owners could either choose a modern heating system or an improved thermal insulation for their home. We used standard and mixed logit specifications to analyse the choice data. We found environmental benefits to have a significant impact on choices of heating systems. However, they played no role in terms of insulation choices. Based on the estimated mixed logit model, we further obtained willingness-to-pay (WTP) measures for CO₂ savings.     

Analysis of School Enrollment in Ghana: A Sequential Approach

In this paper, we depart from the standard way of analyzing school enrollment by accounting explicitly for educational selectivity in order to examine the determinants of child school enrollment in Ghana. Using data from the Ghana Living Standard Survey round 6 (GLSS 6), we estimate a three‐step sequential logit model for the determinants of secondary school enrollment and its dependence on completing primary school. We find that family resources such as parental education, household income and the gender of the head of the household play a role in households' child schooling decisions. Educated parents are relatively more likely to enroll their children in primary school and keep them in school until they complete primary education. As well, we show that educated parents do not promote a gender‐biased investment in the schooling of children at the primary level. While household welfare does not influence children's entry into primary school importantly, their completion of primary school depends on household welfare. The study sheds more light on the pro‐male bias phenomenon regarding entry into primary school and primary school completion. Policies to promote the achievement of Sustainable Development Goal 4 in Ghana must be grade sensitive.     

A Nonparametric Revealed Preference Approach to Measuring the Value of Environmental Quality

We develop an approach to valuing non-market goods using nonparametric revealed preference analysis. We show how nonparametric methods can also be used to bound the welfare effects of changes in the provision of a non-market good. Our main context is one in which the non-market good affects the marginal utility of consuming a related market good. This can also be framed as a shift in the taste for, or quality of, the market good. A systematic approach for incorporating quality/taste variation into a revealed preference framework for heterogeneous consumers is developed. This enables the recovery of the minimal variation in quality required to rationalise observed choices of related market goods. The variation in quality appears as a adjustment to the price for related market goods which then allows a revealed preference approach to bounding compensation measures of welfare effects to be applied.     

Theory and Estimation of the Mortgage Payment Protection Insurance Decision

This paper examines the decision to take out mortgage payment protection insurance (MPPI) in the UK. The paper explains how MPPI has increased in importance over the last decade due to the government stating that Income Support for Mortgage Interest (ISMI) has crowded–out MPPI. A theoretical model of the mortgage protection insurance decision is developed which takes account of the welfare system. The model is estimated using logit analysis on 1995 Glasgow and Bristol data. Elasticities of the probability of take–up with respect to a variety of arguments are calculated, including the level of ISMI. The estimated elasticity with respect to ISMI is found to be very low, which suggests that the crowding–out motivation for the restructuring of Income Support for Mortgage Interest in October 1995 had little support in the data available at the time of the policy decision, and explains the continued low take–up rates since the 1995 restructuring.     

Measuring the Welfare Effects of Price Changes: A Convenient Parametric Approach

This paper shows how compensating and equivalent variations, and the equivalent income, resulting from a set of price changes can be calculated. A linear expenditure system is estimated for each of a range of total expenditure groups using cross-sectional budget data. The measures of welfare change can be used to determine the effect on the welfare of individuals in different income groups. Alternative social welfare functions can be used to evaluate the resulting distribution of equivalent income. The parametric approach is particularly useful where few data are available     

两种冲击视角下中国城乡经济波动福利成本研究*

经济波动福利成本的准确测算，一方面能够对经济波动根源的差异化理论做出评判，另一方面也直接关系到政府对于经济政策的选择。本文基于Lucas福利成本模型，采用BN分解公式对消费序列进行分解，估计了暂时性冲击和持久性冲击所造成的经济波动福利成本。本文研究证实，持久性冲击所造成的经济波动福利成本远大于暂时性冲击所造成的福利成本，避免不利供给因素所带来的持久性冲击，是增进城乡居民福祉的重要举措。这与当前所实施的“供给侧结构性改革”相契合。     

Disability in Italian households: income,poverty and labour market participation

This article studies the economic effects of disabled members on Italian households, with the aim of identifying a suitable target group for welfare policies. Survey of Households Income and Wealth data for the year 2000 is used. Preliminary results show significant differences in levels of income and poverty diffusion to the detriment of households with disabled members. We propose an exogenous explanation: the replacement ratio between disability benefits and expected labour income shows that disability benefits do not compensate the potential incomes of the disabled person and of the possible carer, except in households with severe socioeconomic disadvantages. We also propose an endogenous explanation: applying a logit model we show that the labour market participation of the possible carer is reduced in households with disabled persons. In order to increase the income of the households with disabled members, policy recommendations include the provision of care services and structural policies to improve employment, income and educational opportunities for households at greatest disadvantage.