Beverage Choices of U.S. Households: Determinants and Obesity Conseqences
Rigoberto Lopez, Project Director
Overview
The mix of U.S. beverage consumption has changed dramatically in the last few decades and is now often blamed for contributing to the obesity epidemic, particularly among low-income households. The objectives of this project are to examine the determinants of low-income household beverage choices (including milk, bottled water, juices and carbonated soft drinks) and link them to the incidence of obesity. Taste parameters of low-income households with respect to nutritional characteristics of beverage products will be estimated using the random coefficients logit demand model (Berry, Levinsohn and Pakes, 1995) and in a second-stage regression linked to body mass indexes. Data will come from Information Resources, Inc., for market data and the Behavioral Risk Factor Surveillance System Survey for obesity. Results will provide information on price and cross-price elasticities at the product brand level for low-income households, what nutritional characteristics of beverages low-income households value most, how acquired tastes affect the propensity to be overweight and/or obese. In addition, policy simulation results (i.e., taxes on sugar drinks or subsidies on healthier alternatives) will be conducted.
Objectives
The fundamental inquiry of this research is to examine U.S. consumer preferences in carbonated soft drinks, fruit juices, bottled water and milk among low-income populations and how these preferences are linked to an apparent higher incidence of obesity in this population. The specific objectives are to: 1. Estimate the taste parameters of U.S. households embedded in their beverage choices with respect to nutritional characteristics, particularly those linked to the obesity epidemic. 2. Assess the link between the taste parameters for nutritional characteristics and the probability of being obese and/or overweight. The first objective will be accomplished by estimating a random coefficients logit demand model (Berry, Levinsohn and Pakes, 1995; herein referred as BLP) with scanner data from observed purchases of a sample of U.S. cities to assess the importance of product and consumer characteristics on consumption of carbonated soft drinks (CSDs), fruit juices, bottled water and various types of milk among low-income households. The empirical goals are to estimate preferences for characteristics such as price and calories of low-income populations and to estimate own- and cross-price elasticities with respect to beverages and characteristics. The second objective will be accomplished by rerunning the BLP model with a consumer characteristics set that includes body mass indexes (BMI) so that the estimated taste parameters can be linked to BMI. As the BLP model estimates individual consumer taste parameters, a regression of BMI on these taste parameters and other control variables such as exercise will shed light on preferences underlying beverage choicesand obesity. Expected outputs include a technical report, a graduate student thesis, and a refereed journal article.
This summary of the Beverage Choices of U.S. Households: Determinants and Obesity Conseqences was written by Rigoberto A. Lopez.