The choice of the mixing density in random parameter logit models remains a complex issue. Freque... more The choice of the mixing density in random parameter logit models remains a complex issue. Frequently, trade-offs need to be made between ease of estimation, shape flexibility and behavioural relevance. We contrast a set of alternative distributions, including the asymmetric triangular, in a choice experiment on flood risk exposure and estimate our models in a Bayesian WTP-space framework. Results show that the triangular distribution can be applied, but is outperformed by the lognormal distribution in terms of model fit. The Gibbs Sampler reveals that simple distributional forms are not likely to be consistent with the underlying distribution of preferences.
ABSTRACT In this paper we extend the behavioural scope of discrete choice models for leisure acti... more ABSTRACT In this paper we extend the behavioural scope of discrete choice models for leisure activity-travel choices. More specifically, we investigate to what extent choices for leisure activities and related travels are driven by the satisfaction of needs. In addition to conventional attributes (such as activity costs), our regret based discrete choice model incorporates latent variables representing the anticipated level of individual needs-satisfaction by a particular leisure activity. The latent variables are calibrated with the help of subjective indicators of needs-satisfaction associated with the leisure activities. Results show that needs-satisfaction allows us to decompose a substantial share of the unobserved heterogeneity in leisure activity-travel decisions across respondents. Identifying the structural drivers of anticipated needs-satisfaction also enables a better prediction of leisure activity choice.
Stated choice (SC) surveys are a key tool for studying travel behaviour and are used to inform po... more Stated choice (SC) surveys are a key tool for studying travel behaviour and are used to inform policy decisions in many countries. Recently, the best-worst (BW) variant of SC has rapidly increased in popularity in fields as diverse as transport, marketing and health research. Key arguments for their use have been that they provide more information per choice task, and are superior to standard ranking due to a perceived propensity by respondents to identify and respond more consistently to extreme options. Despite numerous applications of different variants of BW, the stability of preferences across stages has not been tested in any rigorous manner to date, with analysts relying on an assumption that only differences in scale exist between stages. This paper fills the gap by investigating the stability of response patterns across stages for sequential and simultaneous variants of BW SC. Using datasets from three different studies, we show that, regardless of the dataset used, the obtained estimates are not stable across stages, going beyond the differences in scale commonly taken into account. In all our datasets, we found significant and substantial differences between marginal utility estimates and implied monetary valuations across stages. These results thus question the simple use of BW data as a data augmentation approach, especially in applied work which is aimed primarily at forecasting choices rather than preference orderings. Our findings also support recent discussions about framing effects in BW surveys. The unique presence of corresponding data from ranking and BW surveys allows us to give added weight to these conclusions, while they also show that choice task composition alone can play a key role in leading to differences in sensitivities.
Australian Journal of Agricultural and Resource Economics, 2015
ABSTRACT This study tests the transferability of the nonmarket values of water conservation for d... more ABSTRACT This study tests the transferability of the nonmarket values of water conservation for domestic and environmental purposes across three south European countries and Australia applying a common choice experiment design. Different approaches are followed to test the transferability of the estimated values, aiming to minimise transfer errors for use in policy analysis, comparing both single- and multicountry transfers, with and without socio-economic adjustments. Within Europe, significant differences are found between implicit prices for environmental water use, but not for domestic water use. In the Australian case study, alleviating restrictions on domestic water use has no significant value. Pooling the three European samples improves the transferability of the environmental flow values between Europe and Australia. Results show that a reduction in transfer error is achieved when controlling for unobserved and observed preference heterogeneity in the single- and multicountry transfers, providing additional support for the superiority of socio-economic adjustment procedures in value transfer.
Journal of Environmental Economics and Management, 2012
This paper studies patenting decisions by firms in relation to the negotiation and signing of the... more This paper studies patenting decisions by firms in relation to the negotiation and signing of the Helsinki and Oslo protocol as part of the Convention on Long-Range Transboundary Air Pollution. We use a uniquely constructed patent data set on SO2 abatement technologies filed in 15 signatory and non-signatory countries in the period 1970–1997. The data distinguish between so-called ‘mother’ patents,
ABSTRACT This paper introduces to the field of marketing a regret-based discrete choice model for... more ABSTRACT This paper introduces to the field of marketing a regret-based discrete choice model for the analysis of multi-attribute consumer choices from multinomial choice sets. This random regret minimization (RRM) model, which has recently been introduced in the field of transport, forms a regret-based counterpart of the canonical random utility maximization (RUM) paradigm. This paper assesses empirical results based on 43 comparisons reported in peer-reviewed journal articles and book chapters, with the aim of finding out to what extent, when, and how RRM can form a viable addition to the consumer choice modeler's toolkit. The paper shows that RRM and hybrid RRM–RUM models outperform RUM counterparts in a majority of cases, in terms of model fit and predictive ability. Although differences in performance are quite small, the two paradigms often result in markedly different managerial implications due to considerable differences in, for example, market share forecasts.
Climate change may cause most harm to countries that have historically contributed the least to g... more Climate change may cause most harm to countries that have historically contributed the least to greenhouse gas emissions and land-use change. This paper identifies consequentialist and nonconsequentialist ethical principles to guide a fair international burden-sharing scheme of climate change adaptation costs. We use these ethical principles to derive political principles -historical responsibility and capacity to pay -that can be applied in assigning a share of the financial burden to individual countries. We then propose a hybrid 'common but differentiated responsibilities and respective capabilities' approach as a promising starting point for international negotiations on the design of burden-sharing schemes. A numerical assessment of seven scenarios shows that the countries of Annex I of the United Nations Framework Convention on Climate Change would bear the bulk of the costs of adaptation, but contributions differ substantially subject to the choice of a capacity to pay indicator. The contributions are less sensitive to choices related to responsibility calculations, apart from those associated with land-use-related emissions. Assuming costs of climate adaptation of USD 100 billion per year, the total financial contribution by the Annex I countries would be in the range of USD 65-70 billion per year. Expressed as a per capita basis, this gives a range of USD 43-82 per capita per year. ß
The main objective of this study is to examine how repeated choice affects preference learning in... more The main objective of this study is to examine how repeated choice affects preference learning in stated preference experiments. We test different hypotheses related to preference learning by analyzing response patterns and asking respondents in a choice experiment to report their experienced certainty when going through the choice tasks. In a split-sample test, we show that follow-up choice certainty questions are procedural invariant. The self-reported certainty results indicate that learning occurs, but econometric testing procedures do not identify any significant impact of learning effects on parameter estimates or variance across choice tasks. Additional tests of choice consistency suggest that preferences in the choice experiment are stable and coherent.
This paper presents the first value of statistical life (VSL) meta-analysis that empirically esti... more This paper presents the first value of statistical life (VSL) meta-analysis that empirically estimates correction factors for 'out of context' benefits transfer (BT) purposes. In the field of mortality risk reductions elicited willingness to pay values in one risk context, say road safety, are frequently applied in other risk contexts like air pollution. However, differences in risk perception and the population at risk across contexts are likely to result in diverging VSL estimates. In a meta-analysis of 26 international stated preference studies, a Bayesian model is estimated regressing contingent values for mortality risk reductions, originating from three different risk contexts, on the characteristics of the risk reduction itself and additional variables characterizing the underlying studies. A willingness to pay (WTP) premium for mortality risk reductions in the air pollution and general mortality risk context relative to improving road safety is observed. Evaluated at the mean, road safety VSL estimates should be multiplied by a factor 1.8 before being applicable in the air pollution context. Moreover, in an illustrative BT exercise we find limited overlap in the set of context specific predictive VSL distributions. Consequently, 'out of context' BT results in a substantial overor underestimation of the VSL.
Climate change may cause most harm to countries that contribute least to greenhouse gas emissions... more Climate change may cause most harm to countries that contribute least to greenhouse gas emissions. This paper identifies deontology, solidarity and consequentialism as the principles that can serve as a basis for a fair international burden sharing scheme of adaptation costs. We translate these principles into criteria that can be applied in assigning contributions of individual countries, namely historical responsibility,
The choice of the mixing density in random parameter logit models remains a complex issue. Freque... more The choice of the mixing density in random parameter logit models remains a complex issue. Frequently, trade-offs need to be made between ease of estimation, shape flexibility and behavioural relevance. We contrast a set of alternative distributions, including the asymmetric triangular, in a choice experiment on flood risk exposure and estimate our models in a Bayesian WTP-space framework. Results show that the triangular distribution can be applied, but is outperformed by the lognormal distribution in terms of model fit. The Gibbs Sampler reveals that simple distributional forms are not likely to be consistent with the underlying distribution of preferences.
ABSTRACT In this paper we extend the behavioural scope of discrete choice models for leisure acti... more ABSTRACT In this paper we extend the behavioural scope of discrete choice models for leisure activity-travel choices. More specifically, we investigate to what extent choices for leisure activities and related travels are driven by the satisfaction of needs. In addition to conventional attributes (such as activity costs), our regret based discrete choice model incorporates latent variables representing the anticipated level of individual needs-satisfaction by a particular leisure activity. The latent variables are calibrated with the help of subjective indicators of needs-satisfaction associated with the leisure activities. Results show that needs-satisfaction allows us to decompose a substantial share of the unobserved heterogeneity in leisure activity-travel decisions across respondents. Identifying the structural drivers of anticipated needs-satisfaction also enables a better prediction of leisure activity choice.
Stated choice (SC) surveys are a key tool for studying travel behaviour and are used to inform po... more Stated choice (SC) surveys are a key tool for studying travel behaviour and are used to inform policy decisions in many countries. Recently, the best-worst (BW) variant of SC has rapidly increased in popularity in fields as diverse as transport, marketing and health research. Key arguments for their use have been that they provide more information per choice task, and are superior to standard ranking due to a perceived propensity by respondents to identify and respond more consistently to extreme options. Despite numerous applications of different variants of BW, the stability of preferences across stages has not been tested in any rigorous manner to date, with analysts relying on an assumption that only differences in scale exist between stages. This paper fills the gap by investigating the stability of response patterns across stages for sequential and simultaneous variants of BW SC. Using datasets from three different studies, we show that, regardless of the dataset used, the obtained estimates are not stable across stages, going beyond the differences in scale commonly taken into account. In all our datasets, we found significant and substantial differences between marginal utility estimates and implied monetary valuations across stages. These results thus question the simple use of BW data as a data augmentation approach, especially in applied work which is aimed primarily at forecasting choices rather than preference orderings. Our findings also support recent discussions about framing effects in BW surveys. The unique presence of corresponding data from ranking and BW surveys allows us to give added weight to these conclusions, while they also show that choice task composition alone can play a key role in leading to differences in sensitivities.
Australian Journal of Agricultural and Resource Economics, 2015
ABSTRACT This study tests the transferability of the nonmarket values of water conservation for d... more ABSTRACT This study tests the transferability of the nonmarket values of water conservation for domestic and environmental purposes across three south European countries and Australia applying a common choice experiment design. Different approaches are followed to test the transferability of the estimated values, aiming to minimise transfer errors for use in policy analysis, comparing both single- and multicountry transfers, with and without socio-economic adjustments. Within Europe, significant differences are found between implicit prices for environmental water use, but not for domestic water use. In the Australian case study, alleviating restrictions on domestic water use has no significant value. Pooling the three European samples improves the transferability of the environmental flow values between Europe and Australia. Results show that a reduction in transfer error is achieved when controlling for unobserved and observed preference heterogeneity in the single- and multicountry transfers, providing additional support for the superiority of socio-economic adjustment procedures in value transfer.
Journal of Environmental Economics and Management, 2012
This paper studies patenting decisions by firms in relation to the negotiation and signing of the... more This paper studies patenting decisions by firms in relation to the negotiation and signing of the Helsinki and Oslo protocol as part of the Convention on Long-Range Transboundary Air Pollution. We use a uniquely constructed patent data set on SO2 abatement technologies filed in 15 signatory and non-signatory countries in the period 1970–1997. The data distinguish between so-called ‘mother’ patents,
ABSTRACT This paper introduces to the field of marketing a regret-based discrete choice model for... more ABSTRACT This paper introduces to the field of marketing a regret-based discrete choice model for the analysis of multi-attribute consumer choices from multinomial choice sets. This random regret minimization (RRM) model, which has recently been introduced in the field of transport, forms a regret-based counterpart of the canonical random utility maximization (RUM) paradigm. This paper assesses empirical results based on 43 comparisons reported in peer-reviewed journal articles and book chapters, with the aim of finding out to what extent, when, and how RRM can form a viable addition to the consumer choice modeler's toolkit. The paper shows that RRM and hybrid RRM–RUM models outperform RUM counterparts in a majority of cases, in terms of model fit and predictive ability. Although differences in performance are quite small, the two paradigms often result in markedly different managerial implications due to considerable differences in, for example, market share forecasts.
Climate change may cause most harm to countries that have historically contributed the least to g... more Climate change may cause most harm to countries that have historically contributed the least to greenhouse gas emissions and land-use change. This paper identifies consequentialist and nonconsequentialist ethical principles to guide a fair international burden-sharing scheme of climate change adaptation costs. We use these ethical principles to derive political principles -historical responsibility and capacity to pay -that can be applied in assigning a share of the financial burden to individual countries. We then propose a hybrid 'common but differentiated responsibilities and respective capabilities' approach as a promising starting point for international negotiations on the design of burden-sharing schemes. A numerical assessment of seven scenarios shows that the countries of Annex I of the United Nations Framework Convention on Climate Change would bear the bulk of the costs of adaptation, but contributions differ substantially subject to the choice of a capacity to pay indicator. The contributions are less sensitive to choices related to responsibility calculations, apart from those associated with land-use-related emissions. Assuming costs of climate adaptation of USD 100 billion per year, the total financial contribution by the Annex I countries would be in the range of USD 65-70 billion per year. Expressed as a per capita basis, this gives a range of USD 43-82 per capita per year. ß
The main objective of this study is to examine how repeated choice affects preference learning in... more The main objective of this study is to examine how repeated choice affects preference learning in stated preference experiments. We test different hypotheses related to preference learning by analyzing response patterns and asking respondents in a choice experiment to report their experienced certainty when going through the choice tasks. In a split-sample test, we show that follow-up choice certainty questions are procedural invariant. The self-reported certainty results indicate that learning occurs, but econometric testing procedures do not identify any significant impact of learning effects on parameter estimates or variance across choice tasks. Additional tests of choice consistency suggest that preferences in the choice experiment are stable and coherent.
This paper presents the first value of statistical life (VSL) meta-analysis that empirically esti... more This paper presents the first value of statistical life (VSL) meta-analysis that empirically estimates correction factors for 'out of context' benefits transfer (BT) purposes. In the field of mortality risk reductions elicited willingness to pay values in one risk context, say road safety, are frequently applied in other risk contexts like air pollution. However, differences in risk perception and the population at risk across contexts are likely to result in diverging VSL estimates. In a meta-analysis of 26 international stated preference studies, a Bayesian model is estimated regressing contingent values for mortality risk reductions, originating from three different risk contexts, on the characteristics of the risk reduction itself and additional variables characterizing the underlying studies. A willingness to pay (WTP) premium for mortality risk reductions in the air pollution and general mortality risk context relative to improving road safety is observed. Evaluated at the mean, road safety VSL estimates should be multiplied by a factor 1.8 before being applicable in the air pollution context. Moreover, in an illustrative BT exercise we find limited overlap in the set of context specific predictive VSL distributions. Consequently, 'out of context' BT results in a substantial overor underestimation of the VSL.
Climate change may cause most harm to countries that contribute least to greenhouse gas emissions... more Climate change may cause most harm to countries that contribute least to greenhouse gas emissions. This paper identifies deontology, solidarity and consequentialism as the principles that can serve as a basis for a fair international burden sharing scheme of adaptation costs. We translate these principles into criteria that can be applied in assigning contributions of individual countries, namely historical responsibility,
Uploads
Papers by Thijs Dekker