Temporal stability of preferences and willingness to pay for natural areas in choice experiments: A test–retest

The main objective of this paper is to test the temporal stability of stated preferences and willingness to pay (WTP) values from a Choice Experiment (CE) in a test–retest. The same group of participants was asked the same choice tasks in an internet-based CE, conducted twice with a time interval of one year without interviewer interference. We examine choice consistency at individual choice task level and transferability of the underlying indirect utility function and associated WTP values. The results show that choices are consistent in 57 percent of the choice occasions. Comparison of the choice models over time shows that the estimated preference and scale parameters are significantly different, suggesting that choice behaviour changed between the two surveys. Differences between marginal WTP estimates for individual choice attributes are statistically significant only at the 10 percent level. However, we show that this can result in significantly different WTP values for policy scenarios. The instability of estimated mean WTP values for different policy scenarios asks for caution when including WTP values in cost-benefit analysis.     

Low-income fishermen's willingness-to-pay for fisheries and watershed management: An application of choice experiment to Lake Tana,Ethiopia

In this paper, choice experiment was applied for valuation of Lake Tana's fishery and watershed. Two attributes — fishing control and lake side plantation — were identified as relevant attributes for the choice experiment. A monetary attribute — payment for fishing permit was also included. Multinomial and random parameter logit models were used for estimation. All the attributes included were significant factors in affecting the probability of choosing an alternative scenario. The results suggest that fishermen were more concerned about fishing control as reflected by the higher value they give to fishing control than lake side plantation. Household income, years of education, and family size were found to be significant. The economic welfare measures we calculated for two scenarios show that fishermen were willing to pay 57 birr (US$5.3)) per month for a moderate improvement scenario and 93 birr (US$8.6) per month for an aggressive scenario for the two attributes.     

Does attribute order influence attribute-information processing in discrete choice experiments?

The existing empirical evidence shows that both contingent valuation and discrete choice experiment (DCE) methods are susceptible to various ordering effects. However, very few studies have analysed attribute-ordering effects in DCEs, and no study has investigated their potential influence on information-processing strategies, such as attribute non-attendance (ANA). This paper tests for attribute-ordering effects and examines whether the order of attributes describing the alternatives affects respondents’ propensity to attend to or ignore an attribute. A split-sample approach is used, where one sample received a DCE version in which the positions of the first and last non-monetary attributes are switched across the sequence of choice tasks compared with the other sample. The results show that attribute order does not affect welfare estimates in a significant way under the standard assumption of full attribute attendance, thus rejecting the notion of procedural bias. However, the welfare estimates for the attributes whose order was reversed and the share of respondents who ignored them differ significantly between the two attribute-ordering treatments once ANA behaviour is accounted for in the estimated choice models. These results highlight the important role of information-processing strategies in the design and evaluation of DCEs.     

The UK consumption function and structural instability: improving forecasting performance using a time-varying parameter approach

Previous studies indicate that the poor forecasting performance of constant parameter UK consumption expenditure models is caused by structural instability in the underlying data generating process. Typically, this instability is removed by reparameterization within the constant parameter framework. An alternative modelling strategy is to allow some, or all, of the parameters to vary over time. A UK non-durable consumption expenditure model with time-varying parameters is developed, based on the permanent income hypothesis of Friedman (1957). This model takes into account temporal changes in the average and marginal propensities to consume. The variation in the parameter estimates is given an economic interpretation in terms of the influence of omitted variables, namely UK financial liberalization and expectational changes. The forecasting performance of this model is superior to that of two widely used constant parameter models. Further tests show that, even if these constant parameter models are respecified as time varying parameter models, the authors' model still retains a superior forecasting performance.     

A comparison of preferences for two GLP-1 products – liraglutide and exenatide – for the treatment of type 2 diabetes

Abstract

Objective:

To use time trade-off (TTO) to compare patient preferences for profiles of two glucagon-like peptide (GLP-1) products for the treatment of type 2 diabetes (liraglutide and exenatide) that vary on four key attributes – efficacy (as measured by hemoglobin A_1C), incidence of nausea, incidence of hypoglycemia, and dosing frequency (QD vs. BID) – and measure the contribution of those attributes to preferences.

Methods:

A total of 382 people with T2DM were recruited to participate in an internet-based survey consisting of a series of health-related questions, a conjoint exercise and a set of time trade-off items. In the conjoint exercise, respondents were presented with eight pairs of hypothetical GLP-1 profiles, and completed a time-tradeoff exercise for each pair.

Results:

The product profile representing liraglutide was preferred by 96% of respondents and resulted in significantly higher health utilities (0.038) than the product profile representing exenatide (0.978 vs. 0.94, p?<?0.05). Estimated preference scores from the conjoint analysis revealed that efficacy measured by hemoglobin A_1C is the most important attribute, followed by nausea, hypoglycemia, and dosing schedule.

Limitations:

On-line participants may not represent ‘typical’ type 2 diabetes patients, and brief product profiles represented results from clinical trials, not clinical practice

Conclusion:

Based on the four attributes presented, patients prefer liraglutide over exenatide. Preference is based on superior efficacy and less nausea more than less hypoglycemia and once-daily dosing.     

A discrete choice experiment to assess patients’ preferences for HIV treatment in the rural population in Colombia

Abstract

Aim: To elicit patients’ preferences for HIV treatment of the rural population in Colombia.

Methods: A discrete choice experiment (DCE), conducted in a HIV clinic in Bogotá, was used to examine the trade-off between five HIV treatment attributes: effect on life expectancy, effect on physical activity, risk of moderate side-effects, accessibility to clinic, and economic costs to access controls. Attributes selection was based on literature review, expert consultation and a focus group with six patients. An efficient experimental design was used to define two versions of the questionnaire with each of 12 choice sets and a dominance task was added to check reliability. A mixed logit model was then used to analyse the data and sub-group analyses were conducted on the basis of age, gender, education, and sexual preference.

Results: A total of 129 HIV patients were included for analysis. For all treatment attributes, significant differences between at least two levels were observed, meaning that all attributes were significant predictors of choice. Patients valued the effect on physical activity (conditional relative importance of 27.5%) and the effect on life expectancy (26.0%) the most. Sub-group analyses regard age and education showed significant differences: younger patients and high educated patients valued the effect on physical activity the most important, whereas older patients mostly valued the effect on life expectancy and low educated patients mostly valued the accessibility to clinic.

Limitations: One potential limitation is selection bias, as only patients from one HIV clinic were reached. Additionally, questionnaires were partly administered in the waiting rooms, which potentially led to noise in the data.

Conclusions: This study suggests that all HIV treatment characteristics included in this DCE were important and that HIV patients from rural Colombia valued short-term efficacy (i.e. effect on physical activity) and long-term efficacy (i.e. effect on life expectancy) the most.     

The impact of offering two versus three alternatives in choice modelling experiments

Researchers designing choice modelling experiments have some latitude over the number of choice alternatives that can be offered in each choice set. There is some evidence that design dimensions, including the number of alternatives available in each choice set, can influence model outcomes. A key issue is whether referendum formats with binary options are preferable to choice sets with multiple alternatives.A choice modelling experiment was performed where questionnaires delivered to two split samples differed only according to whether two or three alternatives were offered to respondents in each choice set. The results indicate that more robust models could be constructed from the three-alternative split compared to the two-alternative split. One reason for the difference is that respondents tended to display serial non-participation in the two-alternative format, choosing an alternative consistently without regard for changes in the attributes. For practitioners of the CM technique, the results suggest that it may be preferable to offer more than two alternatives in a choice set that includes a status quo option.     

Test–Retest Reliability of Choice Experiments in Environmental Valuation

The paper presents the results of the first test–retest study on choice experiments in environmental valuation. In a survey concerning landscape externalities of onshore wind power in central Germany, respondents answered the same five choice sets at two different points in time. Each choice set comprised three alternatives described by five attributes, and the time interval between the test and the retest was eleven months. The analysis takes place at three different levels, investigating choice consistency at the choice task level and repeatability of the latent construct utility at the level of parametric models as well as at the level of willingness-to-pay estimates. At the choice task level we observed 59?% identical choices. The parametric analysis shows that the test and retest estimates are not equal, even when we control for scale, that is, differences in the error variance. However, comparing the marginal willingness-to-pay estimates among test and retest reveals only a statistically significant difference for one of the attributes. Overall, this indicates a moderate test–retest reliability taking into account that consistency at the choice task level overlooks the stochastic nature of the process underlying discrete choice experiments.     

Valuing improvements in urban water security: evidence of heterogeneity derived from a latent class model for eastern Australia

In many Australian cities the response to drought has included the imposition of mandatory constraints over how water is used by households, often termed ‘water restrictions’. A similar rationing approach has been witnessed in California’s recent drought. The aim of water restrictions is to slow the depletion of water storage but restrictions have also been criticized for the costs they impose on specific water users. In order to gain insight into the potential magnitude of the cost of water restrictions, this study uses a choice experiment to investigate the non-market values for specific attributes associated with the outcomes of drought restrictions. This information was sought to understand the community’s willingness to pay for attributes related to the extent, frequency and duration of water restrictions. The article reports a latent class choice model for a major city in eastern Australia and investigates heterogeneity in preferences towards increasing water availability during drought. This study departs from the existing literature by conducting the choice experiment in a context where water supply is relatively abundant. This unique framing of the choice experiment allows for a useful comparison with existing studies and also raises challenges about the interpretation of the data for planning purposes.     

Hurdle and Latent Class Approaches to Serial Non-Participation in Choice Models

In repeated choice modelling studies, it is often the case that individuals always select the status quo option. Although this pattern may reflect considered choices, they may also be the result of alternative decisions about whether to participate in the choice process at all. Alternative methods of dealing with this behaviour, each with associated implications for estimates of economic values, are presented. In particular we consider the alternative strategies of excluding such individuals from the data, using hurdle models to explicitly model this group, and propose the use of latent class models to endogenously allow for different preference structures. An advantage of the latent class approach is that the form of the non-participation need not be defined in advance. These approaches are considered using UK choice experiment data on food choices where the attributes include genetic modification of food. The latent class approach reveals the presence of two forms of non-participation in the data. 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.     

Non-attendance and dual processing of common-metric attributes in choice analysis: a latent class specification

There is a growing literature that promotes the presence of process heterogeneity in the way that individuals evaluate packages of attributes in real or hypothetical markets and make choices. A centerpiece of current research is the identification of rules that individuals invoke when processing information in stated choice (SC) experiments. These rules may be heuristics used in everyday choice making as well as manifestations of ways of coping with the amount of information shown in choice experiment scenarios. In this paper, using the latent class framework, we define classes based on rules that recognise the non-attendance of one or more attributes, as well as on the addition and the parameter transfer of common-metric attributes. These processing strategies are postulated to be used in real markets as a form of cognitive rationalization. We use a SC data set, where car driving individuals choose between tolled and non-tolled routes, to translate this new evidence into a willingness to pay (WTP) for travel time savings, and contrast it with the results from a model specification in which all attributes are assumed to be attended to and are not added up with parameter preservation. We find that the WTP is significantly higher, on average, than the estimate obtained from the commonly used full relevance and attribute preservation specification.     

Forecasting UK house prices: A time varying coefficient approach

Previous studies of UK house prices, developed from the demand and supply ofhousing or from the asset market approach have been poor in terms of robustness and ex-post forecasting ability. The UK housing market has suffered a number of structural changes, particularly since the early 1980s with substantial house price increases, financial market deregulation and the removal of mortgage market constraints through competition. Consequently, models which assume that the underlying data-generating process is stable and apply constant parameter techniques tend to suffer in terms of parameter instability. This article uses the Time Varying Coefficient (TVC) methodology where the underlying data-generating process in the UK housing market is treated as unstable. The estimation results of the TVC regression of UK house prices is compared with those obtained from three alternative constant parameter regressions. Comparisons of forecasting performance suggest the TVC regression out-performs forecasts from an Error Correction Mechanism (ECM), Vector Autoregressive (VAR) and an Autoregressive Time Series regression.     

Cost thresholds,cut-offs and sensitivities in stated choice analysis: Identification and implications

Within the discrete choice literature, there is growing recognition that some respondents do not process all attributes when evaluating their choice outcomes. Worryingly, the cost attribute is often among those attributes that are likely to be ignored by respondents. We use probabilistic decision process models (similar in form to latent class models, but where we define the classes to describe specific heuristics) to facilitate situations where respondents adopt cost thresholds and cut-offs. We further develop this model to address the potential confounding between preference heterogeneity and processing heterogeneity by simultaneously allowing for a segmentation of respondents based on their sensitivities to cost. Results, based on an empirical dataset on the existence value of rare fish species in Ireland, provide further confirmation that a share of respondents did not attended to cost. Importantly, however, when heterogeneity to cost levels is accounted for the inferred incidence of complete non-attendance is markedly lower, to the extent that when cost thresholds and cut-offs are also accommodated it almost disappears. This modelling approach leads to significant gains in model fit and has important implications for welfare analysis.     

USING ELICITED CHOICE PROBABILITIES TO ESTIMATE RANDOM UTILITY MODELS: PREFERENCES FOR ELECTRICITY RELIABILITY*

When choice data are not available, researchers studying preferences sometimes ask respondents to state the actions they would choose in choice scenarios. Data on stated choices are then used to estimate random utility models, as if they are data on actual choices. Stated and actual choices may differ because researchers typically provide respondents less information than they would have in actuality. Elicitation of choice probabilities overcomes this problem by permitting respondents to express uncertainty about behavior. This article shows how to use elicited choice probabilities to estimate random utility models and reports estimates of preferences for electricity reliability.     

Valuing changes in forest biodiversity

The paper attempts to improve current understanding of the economic value of biodiversity. Instead of the prevailing approach of using only one indicator of biodiversity (typically, species richness) we provide evidence that it is possible to value changes in a number of attributes which describe complex characteristics of biodiversity, based on ecological knowledge. The attributes used include structural, species and functional diversity. The empirical application is a choice experiment study conducted in the Białowieża Forest, Poland: our study is therefore also one of the first to cast light on the value of biodiversity protection in Eastern Europe. Interestingly, respondents valued passive protection regimes resulting in preservation of natural ecological processes. In addition, the respondents seemed to be concerned with the means, and not only the results of protection programmes.     

Bayesian Conjoint Choice Designs for Measuring Willingness to Pay

In this paper, we propose a new criterion for selecting efficient conjoint choice designs when the interest is in quantifying willingness to pay (WTP). The new criterion, which we call the WTP-optimality criterion, is based on the c-optimality criterion which is often used in the optimal experimental design literature. We use a simulation study to evaluate the designs generated using the WTP-optimality criterion and discuss the design of a real-life conjoint experiment from the literature. The results show that the new criterion leads to designs that yield more precise estimates of the WTP than Bayesian D-optimal conjoint choice designs, which are increasingly being seen as the state-of-the-art designs for conjoint choice studies, and to a substantial reduction in the occurrence of unrealistically high WTP estimates.     

Position-Dependent order effects on the prediction of consumer preferences in repeated choice experiments

Using stated choice data collected by experimental design with repeated choice tasks, this study developed an approach to quantify the position-dependent order effects on the prediction of preferences and marginal willingness to pay for product attributes. Results showed that repeated choice tasks allow learning to occur. Models with order effect adjustments showed significant improvements in goodness of fit. Attribute-specific polynomial trends showed the best fit among all models, which could possibly be explained by respondents’ familiarity and sensitivity to different product attributes. Repeated-choice experiments have a good potential to capture consumer preferences more accurately than the single-choice design. But order effects need to be taken into account for preferences and market prediction.     

Economywide impacts of climate change on agriculture in Sub-Saharan Africa

Two possible adaptation scenarios to climate change for Sub-Saharan Africa are analyzed under the SRES B2 scenario. The first scenario doubles the irrigated area in Sub-Saharan Africa by 2050, compared to the baseline, but keeps total crop area constant. The second scenario increases both rainfed and irrigated crop yields by 25% for all Sub-Saharan African countries. The two adaptation scenarios are analyzed with IMPACT, a partial equilibrium agricultural sector model combined with a water simulation module, and with GTAP-W, a general equilibrium model including water resources. The methodology combines the advantages of a partial equilibrium approach, which considers detailed water-agriculture linkages, with a general equilibrium approach, which takes into account linkages between agriculture and nonagricultural sectors and includes a full treatment of factor markets. The efficacy of the two scenarios as adaptation measures to cope with climate change is discussed. Due to the limited initial irrigated area in the region, an increase in agricultural productivity achieves better outcomes than an expansion of irrigated area. Even though Sub-Saharan Africa is not a key contributor to global food production (rainfed, irrigated or total), both scenarios help lower world food prices, stimulating national and international food markets.     

Bargaining with subjective mixtures

This paper reconsiders the Bargaining Problem of Nash (Econometrica 28:155–162, 1950). I develop a new approach, Conditional Bargaining Problems, as a framework for measuring cardinal utility. A Conditional Bargaining Problem is the conjoint extension of a Bargaining Problem, conditional on the fact that the individuals have agreed on a “measurement event”. Within this context, Subjective Mixture methods are especially powerful. These techniques are used to characterise versions of the Nash and the Kalai–Smorodinsky solutions. This approach identifies solutions based only on the individuals’ tastes for the outcomes. It is therefore possible to do Bargaining theory in almost complete generality. The results apply to Biseparable preferences, so are valid for almost all non-expected utility models currently used in economics.     

Sensitivity of Willingness to Pay Estimates to the Level of Attributes in Discrete Choice Experiments

This paper considers issues raised in the application of discrete choice experiments (DCEs) to estimating willingness to pay (WTP). The main issue addressed is the sensitivity of WTP estimates to the level of attributes. A DCE, concerned with preferences for alternative cervical screening programmes, was carried out with women in the Tayside area of Scotland. A split sample design was employed in which respondents were divided into two groups. Each group received a discrete choice questionnaire that varied with respect to the levels of three of the six attributes. The price attribute was one of the attributes that varied across questionnaires. Whilst estimated coefficients were not significantly different across five of the six attributes included in the experiment, mean WTP estimates were significantly different for four of the five welfare estimates. However, from a policy point of view, such a difference may not be important. Consideration is also given to other general methodological and policy issues that are raised when using DCEs to estimate WTP. The findings suggest the need for further research into the design and application of DCEs as a method for estimating WTP.