Estimation of consumers' demand function

Publication Type:
Issue Date:
Full metadata record
In this thesis, we test the existence of the behavioral component in the consumer’s decision-making process that captures the direct influence of other available products and their characteristics on the consumer’s utility. We introduce this behavioral component to the empirical demand model and show that it plays an important role in the widely used approach of employing rival products’ characteristics as instruments to overcome the price endogeneity problem in demand estimation. To do so, we use a dataset on individuals’ choices of the red wines from an experiment. The obtained results show that the exclusion condition is not satisfied for some of the rival products’ characteristics, but is satisfied for other rival products’ characteristics. We extend the choice model by allowing the subjective evaluations of the products’ quality in the consumer’s utility function. We exploit the unique survey design of the discrete choice experiment on wine choice with random prices to estimate the coefficients of consumers’ demand function for wines. The consumers form their subjective evaluations of the quality of the new wine from the bottle design and label information. Consumers’ subjective evaluations of the wine’s quality may be correlated with unobserved product characteristics. To solve the endogeneity problem of the subjective evaluations, we use characteristics of other wines from the randomly formed choice set as instruments. The existence of the individuals’ behavioral bias allows us to use other product characteristics as instruments. Additionally, we study how the purpose of consumption affects individuals’ choices of the wines.
Please use this identifier to cite or link to this item: