|
on Discrete Choice Models |
| By: | André De Palma (ENS Cachan - Ecole Normale Supérieure de Cachan - Ecole Normale Supérieure de Cachan, Department of Economics, Ecole Polytechnique - CNRS : UMR7176 - Polytechnique - X); Karim Kilani (CNAM - Conservatoire National des Arts et Métiers-Paris - CNAM) |
| Abstract: | We study the descriptive and the normative consequences of price and/or other attributes changes in additive random utility models. We first derive expressions for the transition choice probabilities associated to these changes. A closed-form formula is obtained for the logit. We then use these expressions to compute the cumulative distribution functions of the compensating variation conditional on ex-ante and/or ex-post choices. The unconditional distribution is also provided. The conditional moments of the compensating variation are obtained as a one-dimensional integral of the transition choice probabilities. This framework allows us to derive a stochastic version of Shephard's lemma, which relates the expected conditional compensating variation and the transition choice probabilities. We compute the compensating variation for a simple binary linear in income choice model and show that the information on the transitions leads to better estimates of the compensating variation than those obtained when only ex-ante or ex-post information on individual choices is observed. For the additive in income logit, we compute the conditional distribution of compensating variation, which generalizes the logsum formula. Finally, we derive a new welfare formula for the disaggregated version of the represen- tative consumer CES model. |
| Keywords: | Additive random utility models (ARUM), Logit, Transition choice probabilities, Compensating variation, Shephard's Lemma, Logsum, CES |
| Date: | 2009–09–16 |
| URL: | http://d.repec.org/n?u=RePEc:hal:wpaper:hal-00417493_v1&r=dcm |
| By: | André Lapied (GREQAM - Groupement de Recherche en Économie Quantitative d'Aix-Marseille - Université de la Méditerranée - Aix-Marseille II - Université Paul Cézanne - Aix-Marseille III - Ecole des Hautes Etudes en Sciences Sociales - CNRS : UMR6579); Pascal Toquebeuf (GREQAM - Groupement de Recherche en Économie Quantitative d'Aix-Marseille - Université de la Méditerranée - Aix-Marseille II - Université Paul Cézanne - Aix-Marseille III - Ecole des Hautes Etudes en Sciences Sociales - CNRS : UMR6579) |
| Abstract: | This paper studies the application of the two most popular non-expected utility (NEU) models -Choquet Expected Utility (CEU) and Maximin Expected Utility (CEU)- to dynamic choice situations in a purely subjective framework. We give an appropriate version of the reduction of compound acts axiom, that states the equivalence between a static and a dynamic choice situation. We show that if consequentialism -only those consequences that can be reached do matter- is additionally assumed, then a monotonic constant linear representation degenerate into expected utility. We envisage two different ways to resolve this problem for the cases where the representation is a CEU or a MEU one. One way consists to weaken the reduction of compound acts axiom, which does not hold on all events. Another way is to relax consequentialism. Then we axiomatically characterize an updating rule for both approaches allowing recursion in several cases. |
| Keywords: | Choquet expected utility; Maximin expected utility; Consequentialism; Reduction of compound acts; Dynamic choice; Updating |
| Date: | 2009–09–13 |
| URL: | http://d.repec.org/n?u=RePEc:hal:wpaper:hal-00416214_v1&r=dcm |
| By: | Debdulal Mallick |
| Abstract: | In discrete choice models the marginal effect of a variable of interest that is interacted with another variable differs from the marginal effect of a variable that is not interacted with any variable. The magnitude of the interaction effect is also not equal to the marginal effect of the interaction term. I present consistent estimators of both marginal and interaction effects in ordered response models. This procedure is general and can easily be extended to other discrete choice models. I also provide an example using household survey data on food security in Bangladesh. Results show that marginal effects of interaction terms are estimated by standard statistical software (STATA® 10) with very large error and even with wrong sign. |
| Keywords: | Marginal effect, interaction effect, ordered probit, discrete choice. |
| JEL: | C12 C25 |
| Date: | 2009–09–22 |
| URL: | http://d.repec.org/n?u=RePEc:eei:rpaper:eeri_rp_2009_22&r=dcm |
| By: | Xuehong Wang (Centre for Environmental Management, Central Queensland University, Australia); John Rolfe (Regional Development Economics, Faculty of Business and Informatics, Central Queensland University, Australia) |
| Abstract: | Many policy issues, as well as policy funding and management choices, have elements of risk and uncertainty. This means that choice experiments, such as those used in choice modelling (CM), may need to frame trade-offs so that risk and uncertainty are included.This research aims to explore some methodological approaches to identify and treat uncertainty in CM experiments. A review of theoretical models, as well as a case study application in the CM technique reported by Roberts et al. (2008), suggests that including uncertainty information in the choice sets should influence responses significantly. However, key challenges remain to define and describe the elements of risk and uncertainty that are to be included in a choice experiment, to communicate the issues to respondents, and to develop appropriate forms of analysis. |
| Date: | 2009–01 |
| URL: | http://d.repec.org/n?u=RePEc:een:eenhrr:0912&r=dcm |
| By: | Kasia Mazur (Crawford School of Economics and Government, the Australian National University, Australia); Jeff Bennett (Crawford School of Economics and Government, the Australian National University, Australia) |
| Abstract: | The survey was designed to estimate environmental values suitable for integration into MOSAIC, a bio-economic model for catchment and farm level planning. Local residents, as well as distant rural and distant urban communities, were surveyed in three NSW catchments (Lachlan, Namoi and Hawkesbury-Nepean) using choice modelling (CM). The survey aimed to find out respondents’ attitudes about, and preferences for, potential natural resource management (NRM) improvements. In total, 3,997 responses were collected from seven different locations in NSW. Fourteen split samples were established to allow for testing of incentive compatibility in CM, the impact of respondent location on values held, and scale effects. This research report describes the development of the CM questionnaires, the survey design and the data collection process. |
| Keywords: | Nonmarket valuation, choice modelling, survey, questionnaire design |
| Date: | 2009–01 |
| URL: | http://d.repec.org/n?u=RePEc:een:eenhrr:0913&r=dcm |
| By: | Sonia Akter (Crawford School of Economics and Government, the Australian National University, Australia); Jeff Bennett (Crawford School of Economics and Government, the Australian National University, Australia) |
| Abstract: | This study aims to show how Australian households perceive climate change and what they are prepared to do to reduce the harmful effects of climate change. A web-based survey in November 2008 asked approximately 600 New South Wales households about their willingness to pay additional household expenses caused by the Carbon Pollution Reduction Scheme (CPRS) proposed by the Australian government. The Contingent Valuation Method (CVM), a widely used non-market valuation technique, was applied. Results of the study show there is a positive demand to mitigate climate change in Australia resulting from a wish to avoid climate change. Households’ willingness to pay (WTP) for climate change was, however, significantly curbed as households were uncertain about the extent of climate change and whether climate change policies are effective. Australian household support for the CPRS is influenced by schemes of other major greenhouse gas emitting countries (global co-operation). Only when people who didn’t answer the survey are assumed to value climate change mitigation the same as people who did answer the survey, do the benefits of the CPRS, as estimated by respondents’ WTP, exceed its costs. Key words: Contingent valuation, climate change, Carbon Pollution Reduction Scheme, willingness to pay, uncertainty, Australia |
| Date: | 2009–02 |
| URL: | http://d.repec.org/n?u=RePEc:een:eenhrr:0919&r=dcm |
| By: | Marit E Kragt (Crawford School of Economics and Government, the Australian National University, Australia); Jeff Bennett (Crawford School of Economics and Government, the Australian National University, Australia) |
| Abstract: | Choice experiments (CE), also known as choice modelling (CM), are now used widely in environmental valuation in Australia. Many examples assess the trade-offs between river catchment management and socio-economic impacts. There is, however, limited information about the values of Australian estuaries and none of the existing valuation studies addresses catchment management changes in Tasmania. The CE study described in this report aims to elicit community preferences for protecting the rivers and estuary of the George catchment in north-eastern Tasmania. Results from conditional and mixed logit models show that respondents are, on average, willing to pay between $2.47 and $4.46 for a one kilometre increase in native riverside vegetation, and between $9.35 and $10.97 per species for the protection of rare native plants and animals, ceteris paribus. The study results are ambiguous about respondents’ preferences for estuary seagrass area. It also shows significant differences between logit models when accounting for individual heterogeneity, and repeated choices made by individual respondents. Key words: Choice experiments, Preference heterogeneity, Mixed Logit models, River health, Estuary health, Tasmania, Environmental valuation |
| Date: | 2009–02 |
| URL: | http://d.repec.org/n?u=RePEc:een:eenhrr:0916&r=dcm |
| By: | Marit E Kragt (Crawford School of Economics and Government, the Australian National University, Australia); Jeff Bennett (Crawford School of Economics and Government, the Australian National University, Australia) |
| Abstract: | Choice experiments (CE) are increasingly used to estimate the values of non-market goods and services. A cost attribute is typically included in a CE questionnaire to estimate monetary values for changes in the non-market attributes presented. Although the cost attribute is centrally important, there has been limited research into the impacts of varying cost attribute levels on respondents’ choices in CE surveys. The context in which non-market attributes are presented to respondents (the ‘attribute frame’) may also affect value estimates. The challenge for CE practitioners is to identify the ‘appropriate’ attribute frames and cost level range. Results from a CE study in Tasmania show that respondents’ preferences are not impacted by describing an attribute in ‘presence’ versus ‘loss’. The absolute attribute levels, therefore, were most important in this study. Comparisons between different split samples are evidence that changing the cost attribute level does affect respondents’ preferences – higher levels lead to significantly higher estimates of willingness to pay for one of the three environmental attributes. |
| Keywords: | Choice experiments, Mixed Logit models, Environmental valuation, Attribute framing, Cost bias |
| Date: | 2009–02 |
| URL: | http://d.repec.org/n?u=RePEc:een:eenhrr:0917&r=dcm |
| By: | Sonia Akter (Crawford School of Economics and Government, the Australian National University, Australia); Jeff Bennett (Crawford School of Economics and Government, the Australian National University, Australia) |
| Abstract: | This paper proposes an extension to existing models of non-expected utility (NEU) in the stated preference (SP) literature. The extension incorporates the impact of multiple sources of ambiguity in individual decision making behavior. Empirical testing of the proposed decision model was carried out in Australia using a dichotomous choice contingent valuation study of a national ‘Carbon Pollution Reduction Scheme (CPRS)’. The results of the study demonstrate that subjective expectations of the context scenario and subjective policy expectations are important determinants of individual decision making in a SP framework. Furthermore, the results of the study demonstrate that decision weight functions are non-linear (quadratic) in subjective scenario expectations and subjective policy expectation. Although evidence was found to link willingness to pay to scenario ambiguity, policy ambiguity was found to have no statistically significant influence on individual decision making |
| JEL: | C93 D81 |
| Date: | 2009–06 |
| URL: | http://d.repec.org/n?u=RePEc:een:eenhrr:0932&r=dcm |
| By: | Marit E Kragt (Crawford School of Economics and Government, the Australian National University, Australia); Jeff Bennett (Crawford School of Economics and Government, the Australian National University, Australia) |
| Abstract: | Integrated catchment policies are widely used to manage natural resources in Australian catchments. Integration of environmental processes with socio-economic systems is often difficult due to the limitations of decision support tools. To support assessments of the environmental and economic trade-offs of changes in catchment management, fully integrated models are needed. This research demonstrates a Bayesian Network (BN) approach to integrating environmental modelling with economic valuation. The model incorporates hydrological, ecological and economic models for the George catchment in Tasmania. Choice experiments were used to elicit information about the non-market costs and benefits of environmental changes. This allows the efficiency of alternative management scenarios to be assessed. |
| Keywords: | Hydro-economic modelling, Integrated catchment modelling, Ecological modelling, Valuation, Bayesian networks, Water quality |
| Date: | 2009–02 |
| URL: | http://d.repec.org/n?u=RePEc:een:eenhrr:0915&r=dcm |
| By: | Jill Windle (Centre for Environmental Management, Central Queensland University, Australia); John Rolfe (Faculty of Business and Informatics, Queensland University, Australia); Roy Brouwer (Institute of Environmental Studies (IVM)) |
| Abstract: | Metrics for evaluating environmental trade-offs can be developed with varying levels of consistency across case study sites. A key issue is whether standard evaluation experiments can be conducted over multiple sites, or whether experiments have to be tailored to each case study application. To test how useful a consistent approach is, choice modeling (CM) has been used in a number of countries. Choice modeling assess the trade-offs households are prepared to make between water use restrictions, maintaining environmental conditions in waterways, and increased water costs. This research paper reports the results of the Queensland survey. The results show that it is not possible to downplay case study framing issues and that it is not appropriate to standardise applications across case studies that have different characteristics. |
| Keywords: | choice modelling, water, environment, framing |
| Date: | 2009–01 |
| URL: | http://d.repec.org/n?u=RePEc:een:eenhrr:0918&r=dcm |
| By: | Kasia Mazur (Crawford School of Economics and Government, the Australian National University and Australian Bureau of Agricultural and Resource Economics, Australia); Jeff Bennett (Crawford School of Economics and Government, the Australian National University, Australia) |
| Abstract: | To elicit household willingness to pay (WTP) for improvements in environmental quality in three NSW catchments (Lachlan, Namoi and Hawkesbury-Nepean), a choice modelling (CM) study was conducted.This report presents results of research designed to investigate variations in WTP across different communities. The communities included local residents, distant/urban and distant/rural residents. Nine split samples were established to test for ‘location effects’. The analysis involved both conditional logit and random-parameters logit models. Natural resource management (NRM), including Catchment Management Authorities (CMAs), can use the non-market values obtained from this study to guide their investment decisions. Key words: Choice modelling, location effects, non-market valuation, catchment planning, environment |
| Date: | 2009–03 |
| URL: | http://d.repec.org/n?u=RePEc:een:eenhrr:0921&r=dcm |
| By: | Constant Tra (Department of Economics, University of Nevada, Las Vegas) |
| Abstract: | This study develops a discrete choice locational equilibrium model to evaluate the benefits of the air quality improvements that occurred in the Los Angeles area following the 1990 Clean Air Act Amendments (CAAA). The discrete choice equilibrium approach accounts for the fact that air quality improvements brought about by the 1990 CAAA will change housing choices and prices. The study also provides new evidence for the distributional welfare impacts of the 1990 CAAA in the Los Angeles area. Findings suggest that the air quality improvements that occurred in the Los Angeles area between 1990 and 2000 provided substantial general equilibrium benefits to households. The analysis reveals noticeable differences between partial and general equilibrium welfare gains, demonstrating that ignoring equilibrium effects will likely misrepresent the benefits of large environmental changes. In addition, we find that the equilibrium welfare impacts of the 1990 CAAA in the Los Angeles area varied significantly across income groups. |
| Keywords: | benefit analysis, ozone improvement, locational equilibrium, discrete choice |
| JEL: | H0 Q28 R13 R21 |
| Date: | 2009–04 |
| URL: | http://d.repec.org/n?u=RePEc:nlv:wpaper:0922&r=dcm |
| By: | Constant Tra (Department of Economics, University of Nevada, Las Vegas) |
| Abstract: | The paper investigates the implications of nonlinear income effects in random utility models (RUM) for measuring non-marginal welfare impacts. A popular approach in applied welfare analysis is to approximate the expected compensating variation (cv) for an amenity change as the cv of a representative consumer whose indirect utility is given by the expected maximum utility. However, this approach can be misleading in the case of non-marginal changes as it implies that changes in income do not affect the consumer’s choice. In this case the true expected cv can be obtained via simulation. Empirical applications to recreational demand find that the bias from the representative approach is small. This paper re-evaluates the accuracy of the representative consumer approximation in the context of measuring the general equilibrium welfare impacts of large environmental changes. Our findings suggest that, though the representative consumer approximation could lead to biased point estimates of the expected cv, this bias is overwhelmed by the size of the confidence intervals that result from the empirical estimation of household preferences. |
| Keywords: | compensating variation, nonlinear income effects, discrete choice |
| JEL: | Q51 R21 |
| Date: | 2009–03 |
| URL: | http://d.repec.org/n?u=RePEc:nlv:wpaper:0924&r=dcm |