Stated and Inferred Attribute Attendance Models: A Comparison with Environmental Choice Experiments

There is increasing evidence that respondents to choice experiment surveys do not consider all attributes presented in the choice sets. Not accounting for this ‘attribute non‐attendance’ leads to biased parameter estimates, and hence biased estimates of willingness to pay. Various methods exist to account for non‐attendance in the analysis of choice data, with limited agreement as to which method is ‘best’. This paper compares modelling approaches that can account for non‐attendance, based on stated and inferred attribute non‐attendance. Respondents' stated non‐attendance is incorporated in the specification of multinomial and mixed logit models. Inference of non‐attendance is based on equality constrained latent class models. Results show that model fit is significantly improved when attribute non‐attendance is taken into account, and that welfare estimates are lower when incorporating non‐attendance. The inference based on equality constrained latent class models provides the best model fit. There is little concordance between stated and inferred non‐attendance, suggesting that respondents may not answer attendance statements truthfully.     

Estimating values for recreational fishing at freshwater dams in Queensland*

In this paper, estimates of value for recreational fishing are reported for three major freshwater impoundments in Queensland, Australia, using both travel cost and contingent valuation methods. Policy analysts often require estimates of value when analysing the importance of recreation against other uses of impoundments, or when considering the potential for further investments, such as with fish stocking programs. Different forms of the travel cost method are used to estimate separate consumer surpluses associated with two key subgroups of recreational anglers: frequent and occasional anglers. A contingent valuation study is used to estimate the marginal values associated with a potential improvement in fishing experience. The results of the travel cost analysis provide strong evidence that recreational values vary between different groups of anglers and across sites, while the contingent valuation estimates provide values for additional marginal benefits of recreational angling.     

Assessing Consumer and Producer Preferences for Animal Welfare Using a Common Elicitation Format

We assess pig farmers’ willingness‐to‐accept (WTA) higher farm animal welfare (FAW) standards and consumers’ willingness‐to‐pay (WTP) for thus enhanced standards. The analysis is based on discrete choice experiments with nearly identical choice sets for both farmers (N=140) and consumers (N=554). Based on preference estimates from a random parameter logit (RPL) model, supply and demand curves for high‐welfare pork in Germany are estimated and market equilibria are derived for alternative levels of FAW. We find that estimates of consumer WTP are significantly positive for all FAW attributes: consumers value more surface space per pig, more bedding and manipulable material, less surgical interventions and shorter transportation times. In contrast, our model revealed significant producer WTA estimates only for surface area per pig and the amount of bedding material on offer, but not for the other FAW attributes. Farmers who expect to continue farming and engage in direct marketing are more likely to adopt higher FAW standards. Male consumers and those who find price more important than brand, origin or taste are less likely to buy high‐welfare pork, as are consumers who never purchase organic meat products. Market simulations for high‐welfare pork indicate increasing divergence between demand and supply with rising FAW standards. We estimate a market share of 44.6% for pork produced in compliance with an entry‐level FAW programme with standards only slightly above the legal minimum. Programmes with more demanding standards are estimated to gain much smaller market shares.     

On the use of flexible mixing distributions in WTP space: an induced value choice experiment

In this study, we use data from an induced value choice experiment to compare estimates from mixed logit models in willingness to pay (WTP) space using different parameter distributional assumptions. Specifically, we test differences in WTP estimates when using flexible parameter mixing distributions (i.e. Legendre polynomials, step functions and splines) and conventional parameter distributions (normal and lognormal). Similar WTP estimates are obtained. However, we observe that WTP estimates are statistically different from the induced value when conventional distributions are assumed, but they are not when more flexible distributions are assumed. This suggests that flexible distributions can provide more reliable WTP estimates.     

Models of Site‐choice for Walks in Rural Ireland: Exploring Cost Heterogeneity

Farmland can confer significant public good benefits to society aside from its role in agricultural production. In this article, we investigate preferences of rural residents for the use of farmland as a recreational resource. In particular, we use a choice experiment to determine preferences for the development of farmland walking trails. Our modelling approach uses a series of mixed logit models to assess the impact of alternative distributional assumptions for the cost coefficient on the welfare estimates associated with the provision of the trails. Our results reveal that using a mixture of discrete and continuous distributions to represent cost heterogeneity leads to a better model fit and lower welfare estimates. Our results further reveal that Irish rural residents show positive preferences for the development of farmland walking trails in the Irish countryside.     

An equilibrium displacement approach to oligopoly market analysis: an application to trade in the Korean infant formula market

This article uses an equilibrium displacement model (EDM) to assess the impact of a change in trade policy in the Korean infant formula market that is assumed to be oligopolistic. Domestic oligopolists compete in a market for a homogenous good with importing firms who enjoy the same level of market power as the domestic producers. Within an EDM framework, we investigate how the market adjusts to a new equilibrium in response to an external policy shock that improves access to foreign goods. Our empirical work consists of two steps. We first econometrically estimate the market demand elasticity and market power parameter. Then, using these estimates, we simulate our EDM model to assess the distributional effects of tariff cuts and other changes in marginal cost. Simulation results show that, with a lower market price and increased consumption, the change in consumer surplus resulting from tariff reduction is unambiguously positive, but that the change in social welfare can be in either direction, depending on who receives market rents associated with imports. When importing firms are domestic, the domestic welfare change is definitely positive. However, in the case of foreign importing firms, the direction of the welfare change crucially depends on the demand elasticity.     

What Do Farmers Want From Agri‐Environmental Scheme Design? A Choice Experiment Approach

Agri‐environmental schemes (AES) have had a limited effect on European agriculture due to farmers’ reluctance to participate. Information on how farmers react when AES characteristics are modified can be an important input to the design of such policies. This article investigates farmers’ preferences for different design options in a specific AES aimed at encouraging nitrogen fixing crops in marginal dry‐land areas in Spain. We use a choice experiment survey conducted in two regions (Aragón and Andalusia). The analysis employs an error component random parameter logit model allowing for preference heterogeneity and correlation amongst the non‐status quo alternatives. Farmers show a strong preference for maintaining their current management strategies; however, significant savings in cost or increased participation can be obtained by modifying some AES attributes.     

Organic‐ and Animal Welfare‐labelled Eggs: Competing for the Same Consumers?

To understand the market for sustainable foods, it is important to allow for heterogeneous preferences. However, most studies of consumer preferences for sustainable foods only investigate average consumer preferences. They do not take into account that some consumer segments attempt to purchase as much sustainable food as possible, others are almost indifferent to the notion of sustainable food, and still others consider sustainable food a complete hoax. The aim of this study is to explore the preferences for various types of premium eggs across three consumer segments. We conduct a choice experiment including 900 Norwegian consumers and perform a behavioural segmentation based on the frequency of organic food purchase. We find that the segment purchasing the most organic food is, as expected, willing to pay a significant premium for organic eggs over eggs displaying only enhanced animal welfare. However, most consumers, who only occasionally purchase organic products, are unwilling to pay more for organic eggs than for enhanced animal welfare eggs, suggesting diminishing marginal utility for additional attributes. We find that a third consumer segment attempts to avoid organic eggs, even when they cost the same as other eggs. Our findings suggest that organic products will be unsuccessful in acquiring a larger share of the market as long as most consumers are unwilling to pay a premium for organic products with all their cost increasing sustainable attributes over products that have only a single sustainable attribute, in our case enhanced animal welfare.     

Using choice experiments to value catchment and estuary health in Tasmania with individual preference heterogeneity*

Choice experiments (CE) have become widespread as an approach to environmental valuation in both Australia and overseas. However, there are few valuation studies that have addressed natural resource management (NRM) changes in Tasmania. Furthermore, few studies have focussed on the estimation of estuary values. The CE study described in this paper aims to analyse community preferences for NRM options in the George catchment, Tasmania. Catchment health attributes were: the length of native riverside vegetation; the number of rare native animal and plant species in the George catchment; and area of healthy seagrass beds in the Georges Bay, which was used as a measure of estuary condition. Mixed logit models with interactions between socio‐economic variables and the choice attributes were estimated to account for systematic and random taste heterogeneity across respondents. Results reveal considerable variation in preferences towards the attributes and show that value estimates are significantly impacted by the way in which we account for preference heterogeneity. Preference heterogeneity thus needs to be considered when estimating community willingness‐to‐pay for environmental changes. This study further shows little responsiveness to the presented changes in estuary seagrass area.     

Effect of Experimental Design on Choice-Based Conjoint Valuation Estimates

In this article, we investigate the effect of several commonly used experimental designs on willingness-to-pay in a Monte Carlo environment where true utility parameters are known. All experimental designs considered in this study generated unbiased valuation estimates. However, random designs or designs that explicitly incorporated attribute interactions generated more precise valuation estimates than main effects only designs. A key result of our analysis is that a large sample size can substitute for a poor experimental design. Overall, our results indicate that certain steps can be taken to achieve a manageably sized experimental design without sacrificing the credibility of welfare estimates.     

Examining demand for urban river rehabilitation in Indonesia: Insights from a spatially explicit discrete choice experiment

The demand for urban river rehabilitation can be measured through stated preference surveys such as choice experiments, providing information on the welfare estimates of a particular approach. We deploy such a technique in the context of plans to rehabilitate a major river in Jakarta, Indonesia. The current plan focuses on widening and canalizing the downstream segment of the river within Jakarta’s administrative boundary. We hypothesize that residents would demand (and thus be willing to pay for) additional components of an ecological rehabilitation program in the form of riverside park space and upstream forest conservation outside of Jakarta’s jurisdiction. We develop a spatially-explicit discrete choice experiment in which households register their preferences for channel widening, park space, forest conservation, and a monthly fee to fund the rehabilitation. Using mixed logit models we find significant and substantial demand for both park space and forest conservation, with a lower bound on the total willingness to pay (WTP) of greater than US $4 million per year for park space and nearly US $6 million per year to support reforestation in the upper catchment. These estimates are based on households within the catchment, but we find that demand did not seem to decay with distance so the upper bound on total WTP could be substantially higher. We also find that household income level has a strong effect on marginal WTP for forest conservation, minimal effect on marginal WTP for park space, and that location along the river influenced WTP for park space and channel widening. This provides further evidence that there is substantial demand for river rehabilitation in developing world cities, and that choice experiments can provide information relevant to land use planning.     

The Optimal Length of an Agricultural Carbon Contract

In this paper, we present the economic determinants of the optimal length of a carbon offset contract. We find that because of a declining capacity of the soil to sequester carbon, the optimal length of the carbon contract is finite (the marginal benefit of remaining in the contract is declining over time, whereas marginal opportunity cost is rising). We also explore the effect of varying key parameter values on the optimal length in the contract. If the contract requires the farmer to sequester at a higher rate, the farmer chooses the contract for a shorter length of time, and this may decrease rather than increase social welfare. If society places a higher value on carbon accumulation, the contract is chosen for a longer length of time. Finally, if both the farmer and society have a higher discount rate, the model provides a somewhat surprising result. The overall time in the contract, and benefits from carbon accumulation are higher when the common discount rate is higher.     

Go climb a mountain: an application of recreation demand modelling to rock climbing in Scotland

In this paper, we apply random utility modelling techniques to rock‐climbing in Scotland. Attributes relevant to choices over rock‐climbing sites were identified from focus groups with climbers, along with a categorisation of principal climbing areas. A survey of climbers yielded 267 responses, which were then used as the basis for modelling. We compare a standard multi‐nominal logit model with a random parameters approach, and look at seasonal differences in behaviour, and at the implications of different treatments of travel time. The random utility models showed that most of the attributes selected were significant determinants of choice. Welfare estimates of changes in site attributes are presented, which are relevant to policy choices currently facing land managers.     

Unraveling determinants of inferred and stated attribute nonattendance: Effects on farmers’ willingness to accept to join agri-environmental schemes

Attribute nonattendance (ANA) has received very little attention in the context of willingness to accept (WTA), although an increasing number of studies analyze the preferences of ecosystem service providers toward incentive-based schemes. We add to the understanding of ANA behavior by analyzing stated and inferred ANA in a choice experiment investigating farmers’ WTA for participating in agri-environmental schemes (AES) in southern Spain. We use mixed logit models, following Hess and Hensher for the inferred ANA approach. Evidence is found of ANA behavior for both stated and inferred approaches, with models accounting for ANA clearly outperforming those that do not account for it; however, we produce no conclusive results as to which ANA approach is best. WTA estimates are only moderately affected, which to some extent is consistent with the low level of non-attendance found for the monetary attribute. Stated and inferred approaches show very similar WTA estimates. Additionally, we investigate sources of observed heterogeneity related to ANA behavior by using a sequence of bivariate probit models for each attribute. Overall, our results hint at a positive relationship between ease of scheme adoption and nonattendance to attributes. However, further research is still needed in this field.     

Consumer willingness to pay for chicken welfare attributes in Kenya

Abstract

This study provides new insights on consumer preferences for chicken welfare attributes in a developing country context. Further, the inclusion of chicken transportation methods as an attribute is a useful contribution to empirical literature. Choice experiment survey data from 200 chicken consumers in Kenya were analyzed using random parameter logit model. Consumers had a positive preference for use of certified transportation, humanely slaughtered chicken and welfare labeling. However, there was a negative preference for using antibiotics in chicken production. Relative to the current price of chicken meat, consumers were willing to pay a premium of 30% for use of certified transport, 72% for animal welfare labeling, 135% for humane slaughter, 236% for nonuse of growth hormones and 40% less for chicken reared in confined systems.     

Convergent validity between revealed and stated recreation demand data: Some empirical evidence from the Basque Country,Spain

In a travel cost exercise, reported past visits to mount Jaizkibel, a natural area located in the Basque Country (Spain), are compared for convergent validity to stated intended future trips under the assumption that the natural resource's conditions will remain the same. In line with the results obtained by other studies, the empirical evidence of this application suggests that revealed preferences (RP) and stated preferences (SP) do not produce consistent data, i.e. do not achieve convergent validity. The paper deals with the convergent validity literature in continuous-choice studies by using two-staged count data models for recreation demand. Differences in preference structures and welfare estimates are tested assuming both common and different data generating processes for the RP and SP data.     

A Comparison of Conjoint Analysis Response Formats

A split-sample design is used to evaluate the convergent validity of three response formats used in conjoint analysis experiments. We investigate whether recoding rating data to rankings and choose-one formats, and recoding ranking data to choose one, result in structural models and welfare estimates that are statistically indistinguishable from estimates based on ranking or choose-one questions. Our results indicate that convergent validity of ratings, ranks, and choose one is not established. In addition, we find that people frequently use "ties" in responses to rating questions, and that the option not to choose any of the alternatives ("opt-out") affects some preference estimates.     

Consumers' willingness to pay for food quality - The case of eggs

The aim of this paper is to introduce an approach to estimating the marginal willingness to pay (WTP) for food quality and safety using data on actual consumer purchases. Marginal WTP for specific food product varieties can be derived from estimated demand functions by inverting the relations, thereby expressing the prices consumers are willing to pay as functions of quantity demanded. The task addressed in the paper is isolation of the WTP for quality attributes embedded in the varieties. Two models based on the AIDS-specification are proposed and discussed. To illustrate the applicability of the models, the case of five varieties of eggs is presented. It is found that Danish consumers are willing to pay a relatively high premium for improved animal welfare and organic production methods and somewhat less for food safety.     

Valuing Australian botanic collections: a combined travel‐cost and contingent valuation study*

Economic values of biological collections in three Australian botanic gardens in Canberra, Melbourne and Sydney were estimated using the travel‐cost method (TCM) and the contingent valuation method (CVM). The TCM component of the study produced average per‐trip consumer surplus (CS) values of $39 and $18 for single‐ and multiple‐site visitors, respectively, for each botanic garden, resulting in an estimate of approximately $194 million for the total social welfare generated by trips to the three sites. Marginal willingness to pay (WTP) for access to botanic gardens was also investigated through payment vehicles of entry fees or higher parking charges using the CVM component. The analysis revealed a positive mean WTP of between $3 and $5 per trip per person. The difference between the CVM and TCM results reflect the different underlying concepts of value under investigation: average CS per visit for the TCM and the utility arising from a marginal visit for the CVM. Marginal changes in CS from the TCM were derived. The confidence intervals from the TCM marginal values overlap the WTP estimates from the CVM. These findings will be useful for resource management decisions in the botanic gardens collection in Australia.