|
on Discrete Choice Models |
| By: | Benoît Chèze (IFPEN - IFP Energies nouvelles - IFPEN - IFP Energies nouvelles, EconomiX - UPN - Université Paris Nanterre - CNRS - Centre National de la Recherche Scientifique, ECO-PUB - Economie Publique - INRA - Institut National de la Recherche Agronomique - AgroParisTech); Maia David (ECO-PUB - Economie Publique - INRA - Institut National de la Recherche Agronomique - AgroParisTech); Vincent Martinet (ECO-PUB - Economie Publique - INRA - Institut National de la Recherche Agronomique - AgroParisTech, EconomiX - UPN - Université Paris Nanterre - CNRS - Centre National de la Recherche Scientifique) |
| Abstract: | Despite reducing the use of pesticides being a major challenge in developed countries, dedicated agri-environmental policies have not yet proven successful in doing so. We analyze conventional farmers' willingness to reduce their use of synthetic pesticides. To do so, we conduct a discrete choice experiment that includes the risk of large production losses due to pests. Our results indicate that this risk strongly limits farmers' willingness to change their practices, regardless of the consequences on average profit. Furthermore, the administrative burden has a significant effect on farmers' decisions. Reducing the negative health and environmental impacts of pesticides is a significant motivator only when respondents believe that pesticides affect the environment. Farmers who earn revenue from outside their farms and/or believe that yields can be maintained while reducing the use of pesticides are significantly more willing to adopt low-pesticide practices. Policy recommendations are derived from our results. |
| Abstract: | La réduction de l'usage de produits phytosanitaires est devenu un défi majeur en France et dans la plupart des pays développés. L'essor récent du Bio en France (+20% de ventes en 2016, BioBaromètre 2017) montre une réelle prise de conscience du grand public et de nouvelles exigences des consommateurs. Plusieurs pratiques permettant de réduire le recours aux produits phytosanitaires tout en maintenant les profits semblent avoir émergé dans des réseaux pilotes (Lerchenet et al 2017) mais demeurent peu appliquées par la majorité des agriculteurs. Les politiques publiques mises en place jusqu'à présent ont abouti à des résultats décevants et l'utilisation de produits phytosanitaires continue d'augmenter (hausse d'environ 12% en France entre 2009 et 2014). Notre étude s'interroge sur les principaux freins et principales motivations des agriculteurs français à réduire leur utilisation de pesticides. Elle repose sur la méthode d'expérimentation par les choix (Discrete Choice Experiment) et permet de mesurer le poids relatif de différents facteurs de décision dans les choix de pratiques des agriculteurs. Elle permet également d'estimer des consentements et payer / à recevoir pour les éléments non monétaires associés à ces choix de pratiques agricoles. Nous montrons notamment que le rôle du risque-récolte est prépondérant dans la réticence des agriculteurs à réduire leur usage de produits phytosanitaires. |
| Keywords: | Q18,Q51,Q57,C35,industrial buildings,willingness to pay,Pesticides,Agricultural practices,Production risk,Discrete choice experiment JEL Classification: Q12,phytosanitaire,consentement à payer,risque,pesticide,pratique agricole,choix discret |
| Date: | 2020–01 |
| URL: | http://d.repec.org/n?u=RePEc:hal:journl:hal-02317175&r=all |
| By: | Melvin Wong; Bilal Farooq |
| Abstract: | We present a Residual Logit (ResLogit) model for seamlessly integrating a data-driven Deep Neural Network (DNN) architecture in the random utility maximization paradigm. DNN models such as the Multi-layer Perceptron (MLP) have shown remarkable success in modelling complex data accurately, but recent studies have consistently demonstrated that their black-box properties are incompatible with discrete choice analysis for the purpose of interpreting decision making behaviour. Our proposed machine learning choice model is a departure from the conventional feed-forward MLP framework by using a dynamic residual neural network learning based approach. Our proposed method can be formulated as a Generalized Extreme Value (GEV) random utility maximization model for greater flexibility in capturing unobserved heterogeneity. It can generate choice model structures where the covariance between random utilities is estimated and incorporated into the random error terms, allowing for a richer set of higher-order substitution patterns than a standard logit might be able to achieve. We describe the process of our model estimation and examine the relative empirical performance and econometric implications on two mode choice experiments. We analyzed the behavioural and theoretical properties of our methodology. We showed how model interpretability is possible, while also capturing the underlying complex and unobserved behavioural heterogeneity effects in the residual covariance matrices. |
| Date: | 2019–12 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:1912.10058&r=all |
| By: | Karun Adusumilli; Dita Eckardt |
| Abstract: | We propose a new algorithm to estimate the structural parameters in dynamic discrete choice models. The algorithm is based on the conditional choice probability approach, but uses the idea of Temporal-Difference learning from the Reinforcement Learning literature to estimate the different terms in the value functions. In estimating these terms with functional approximations using basis functions, our approach has the advantage of naturally allowing for continuous state spaces. Furthermore, it does not require specification of transition probabilities, and even estimation of choice probabilities can be avoided using a recursive procedure. Computationally, our algorithm only requires solving a low dimensional linear equation. We find that it is substantially faster than existing approaches when the finite dependence property does not hold, and comparable in speed to approaches that exploit this property. For the estimation of dynamic games, our procedure does not require integrating over the actions of other players, which further heightens the computational advantage. We show that our estimator is consistent, and efficient under discrete state spaces. In settings with continuous states, we propose easy to implement locally robust corrections in order to achieve parametric rates of convergence. Preliminary Monte Carlo simulations confirm the workings of our algorithm. |
| Date: | 2019–12 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:1912.09509&r=all |
| By: | Cowling, Michael Leith (University of Melbourne); Wooden, Mark (Melbourne Institute of Applied Economic and Social Research) |
| Abstract: | This paper examines the extent to which solo self-employment serves as a vehicle for job creation. Using panel data from the Household, Income and Labour Dynamics in Australia (HILDA) Survey, a dynamic multinomial logit model of transitions between labour market states is estimated. The empirical strategy closely follows that used in a previous study employing household data from Germany by Lechmann and Wunder (2017). Estimates of true cross-state dependence between solo self-employment and employership are obtained that are relatively small. Further, our results imply that the probability of a male remaining an employer just two years after transitioning out of solo self-employment is only 2% (and among women, it is virtually zero). The extent of both true cross-state dependence and true state dependence in employership is, however, much greater among individuals who have demonstrated a preference for self-employment in the past. This implies that pro-entrepreneurial policies that target more 'entrepreneurial' individuals will have more pronounced and long-term effects in stimulating job creation. |
| Keywords: | dynamic multinomial logit, HILDA Survey, solo self-employment, state dependence, stepping stones |
| JEL: | L26 |
| Date: | 2019–12 |
| URL: | http://d.repec.org/n?u=RePEc:iza:izadps:dp12841&r=all |
| By: | Youssef M Aboutaleb; Mazen Danaf; Yifei Xie; Moshe Ben-Akiva |
| Abstract: | This paper introduces a new data-driven methodology for estimating sparse covariance matrices of the random coefficients in logit mixture models. Researchers typically specify covariance matrices in logit mixture models under one of two extreme assumptions: either an unrestricted full covariance matrix (allowing correlations between all random coefficients), or a restricted diagonal matrix (allowing no correlations at all). Our objective is to find optimal subsets of correlated coefficients for which we estimate covariances. We propose a new estimator, called MISC, that uses a mixed-integer optimization (MIO) program to find an optimal block diagonal structure specification for the covariance matrix, corresponding to subsets of correlated coefficients, for any desired sparsity level using Markov Chain Monte Carlo (MCMC) posterior draws from the unrestricted full covariance matrix. The optimal sparsity level of the covariance matrix is determined using out-of-sample validation. We demonstrate the ability of MISC to correctly recover the true covariance structure from synthetic data. In an empirical illustration using a stated preference survey on modes of transportation, we use MISC to obtain a sparse covariance matrix indicating how preferences for attributes are related to one another. |
| Date: | 2020–01 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2001.05034&r=all |
| By: | Tien Mai; Patrick Jaillet |
| Abstract: | We study robust versions of pricing problems where customers choose products according to a general extreme value (GEV) choice model, and the choice parameters are not given exactly but lie in an uncertainty set. We show that, when the robust problem is unconstrained and the price sensitivity parameters are homogeneous, the robust optimal prices have a constant markup over products and we provide formulas that allow to compute this constant markup by binary search. We also show that, in the case that the price sensitivity parameters are only homogeneous in each subset of the products and the uncertainty set is rectangular, the robust problem can be converted into a deterministic pricing problem and the robust optimal prices have a constant markup in each subset, and we also provide explicit formulas to compute them. For constrained pricing problems, we propose a formulation where, instead of requiring that the expected sale constraints be satisfied, we add a penalty cost to the objective function for violated constraints. We then show that the robust pricing problem with over-expected-sale penalties can be reformulated as a convex optimization program where the purchase probabilities are the decision variables. We provide numerical results for the logit and nested logit model to illustrate the advantages of our approach. Our results generally hold for any arbitrary GEV model, including the multinomial logit, nested or cross-nested logit. |
| Date: | 2019–12 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:1912.09552&r=all |
| By: | Tamara Sheldon; Rubal Dua (King Abdullah Petroleum Studies and Research Center) |
| Abstract: | Subsidies for promoting plug-in electric vehicle (PEV) adoption are a key component of China’s overall plan for reducing local air pollution and greenhouse gas (GHG) emissions from its light-duty vehicle sector. This paper explores the impact and cost-effectiveness of the Chinese PEV subsidy program. A vehicle choice model is estimated using a large random sample of individual-level data for new vehicle purchases in China for model year 2017. |
| Keywords: | China Electric Car Market, China New Energy Vehicle Policy, Subsidies for Electric Vehicle |
| Date: | 2019–12–29 |
| URL: | http://d.repec.org/n?u=RePEc:prc:dpaper:ks--2019-dp77&r=all |
| By: | Martinovici, A. (Tilburg University, School of Economics and Management) |
| Abstract: | This dissertation contains three empirical essays on the role of attention in consumer choice. The models developed in this dissertation propose that attention reveals moment-to-moment utility accumulation processes that take place during choice. The first essay investigates which fundamental attention processes contribute to the accumulation of utility and brand choice. The results show that certain types of attention (e.g. attention for integration) are better able to reflect brand utilities, and brand loyalty manifests itself via attention. Essay 2 looks into the link between attention, brand choice, and moment of choice and proposes a model where both brand and search utilities change from moment to moment as more eye movements are observed. This provides insights into consumer heterogeneity in decision thresholds and implicitly decision duration, and test different drivers of brand choice and moment of choice. In essay 3, brand utilities are decomposed into two components that capture the importance of the attributes that describe the brands and the subjective value that the consumer attaches to the attribute levels corresponding to each of the brands on display. The results show that eye movements reflect not only how the consumer evaluates the brands, but also why some brands are preferred by identifying which attributes are considered more important over time. |
| Date: | 2019 |
| URL: | http://d.repec.org/n?u=RePEc:tiu:tiutis:7dca38a5-9f78-4aee-bd81-cc643aa8d499&r=all |
| By: | David Ennnen (Institute of Transport Economics, Muenster) |
| Abstract: | In the policy debate on ride-hailing services such as Uber, the impacts on traffic, emissions, and public transport are hotly discussed. The regulatory framework in Germany has so far prevented a widespread entry of ride-hailing providers. In this paper, we use a mode choice model and trip data to determine the likely impacts of ride-hailing services for a representative region in Germany. We find that the significantly lower fares compared to taxis lead to strong substitution of public transport, cycling, and walking. As a consequence, motorized traffic increases, despite the pooling of individual rides by ride-hailing providers. However, the total impact on mode choice and traffic remains modest, and a widespread displacement of public transport is not to be expected. The final welfare analysis shows that the emergence of ride-hailing services is beneficial for society as a whole. In particular, the benefits from lower fares exceed the external costs arising from additional motorized traffic. |
| Keywords: | Ride-hailing, Transportation Network Company, TNC, Taxi, Regulation, Germany |
| JEL: | L92 L98 |
| Date: | 2020–01 |
| URL: | http://d.repec.org/n?u=RePEc:mut:wpaper:29&r=all |