Progress 09/17/01 to 09/30/06
Outputs
Dr. Mojduszka left the University in 2006.
Impacts
This research focuses on the important determinants of consumer demand for food products and how they have been changing over the last decade. Understanding these factors is key information for assessing the competitiveness of U.S. agricultural producers and food processors as they choose product designs and marketing strategies; for understanding the benefits and costs of government regulations, such as labeling, intended to influence consumer food choice and, ultimately, public health; and for evaluating the impact of changing consumer demand for food on the agricultural and food sectors of the U.S. economy.
Publications
- No publications reported this period
|
Progress 01/01/04 to 12/31/04
Outputs
During the last calendar year, we have successfully completed the data phase of our research work. We have compiled one comprehensive data set that consists of all the data information necessary to complete the project. We have also started the estimation phase of our work that we expect to complete by 04/30/2005. Producing this very comprehensive analysis for our project requires the development and integration of multiple data sources that provide information relevant to the determinants of consumer food choice. These include IRI Info-scanTM Data for quantities, prices, and in-store promotion levels; Nutritional Quality Change Data at the University of Massachusetts and the USDA National Nutrient Database for nutrient content information; National Leading Advertisers Data for advertising expenditures; and USDA Diet and Health Knowledge Survey Data and Consumer Demographics Data for consumer characteristics. We obtained data on market shares, prices, brand offerings,
and in-store marketing efforts for all the products in selected food categories from the IRI Info-scan database. We matched the IRI Info-scan quarterly market shares, prices, and other data for each product with five other data sources. First, we matched it to the nutrition composition data included in the Nutritional Quality Change Data developed at the University of Massachusetts and to the USDA National Nutrient Database available at the USDA web site. The IRI Info-scan data do not provide information on the amounts of nutrients contained in food products. Thus, information on market shares and prices had to be matched with information on the nutritional content of the respective products from the other two data sources. In order to obtain accurate information on the nutritional content of the products that are included in our empirical analysis, we compared, evaluated, and complemented two data sources. The USDA National Nutrient Database was used to complement the Nutritional
Quality Change Data, which provides exact information on the nutritional composition of all products offered in a large super-store but does not contain information on all products offered at the national level. Nonetheless, the latter set includes a complete census of all products in the most popular package size offered in 33 food product categories in a representative super-store in New England for the years 1992 through 1999. Second, we obtained information on the distribution of consumer knowledge about nutrition and nutrition label use by sampling individuals from the Diet and Health Knowledge Survey (DHKS) conducted by the U.S. Department of Agriculture. Third, we obtained information on the distribution of consumer demographic variables by sampling individuals from the Current Population Survey (CPS) carried out each year by the U.S. Bureau of the Census. Fourth, we matched the IRI Info-scan data with the quarterly expenditures on advertising for each of the products in the
four product categories taken from the Leading National Advertisers database for 1993-2002.
Impacts
This research focuses on the important determinants of consumer demand for food products and how they have been changing over the last decade. Understanding these factors is key information for assessing the competitiveness of U.S. agricultural producers and food processors as they choose product designs and marketing strategies; for understanding the benefits and costs of government regulations, such as labeling, intended to influence consumer food choice and, ultimately, public health; and for evaluating the impact of changing consumer demand for food on the agricultural and food sectors of the U.S. economy.
Publications
- No publications reported this period
|
Progress 01/01/03 to 12/31/03
Outputs
During the last year, our work concentrated on developing and estimating a model of consumer food product choice and nutrition information/knowledge. The objective of this research is to integrate a latent variable model of consumer information about nutrition, which aims to capture unobservable concepts, with a discrete food product choice model. The methodology incorporates indicators of the latent variables provided by responses to survey questions to aid in estimating the model. A simulated estimator is used, which results in latent variables that provide the best fit to both the choice and the latent variable indicators. The integrated model consists of two parts: a discrete choice model and a latent variable model. Each part consists of one or more structural equations and one or more measurement equations. To estimate the integrated model it is necessary to use simulation approaches. In addition, we utilize the two-stage sequential estimation method. In the
first stage, we perform factor analysis on the indicators and use the fitted latent variables in the choice model. In the second stage, we integrate the choice probability over the distribution of the latent variables obtained from the factor analysis. The two-stage sequential method with integration gives us consistent estimates (Madanat et al. 1995, Morikawa 1989). We apply simulation techniques to approximate the integrals in the likelihood function and maximize the logarithm of the resulting simulated likelihood function across all individuals with respect to the parameters to be estimated in the model. The procedure to simulate the likelihood function is as follows: (1) for a given value of the parameter vectors, draw a particular realization of the parameter from its distribution, (2) compute the probability of the chosen alternative for each choice occasion, (3) insert the computed probabilities into the log-likelihood function to obtain a simulated individual likelihood
function as the product of the average likelihood functions across all choice occasions of the individual, (4) repeat steps (1)-(3) several times and average the results. This average is the simulated estimator that is unbiased by construction. Its variance decreases as the number of draws increases. It is also strictly positive, so that the logarithm of the simulated probability is defined. We use the Halton sequences to draw individuals from corresponding empirical distributions. The estimations and computations for this paper were carried out using MATLAB programming language on a personal computer. Gradients of the log simulated likelihood function with respect to the parameters were coded.
Impacts
This research focuses on the important determinants of consumer demand for food products and how they have been changing over the last decade. Understanding these factors is key information for assessing the competitiveness of U.S. agricultural producers and food processors as they choose product designs and marketing strategies; for understanding the benefits and costs of government regulations, such as labeling, intended to influence consumer food choice and, ultimately, public health; and for evaluating the impact of changing consumer demand for food on the agricultural and food sectors of the U.S. economy.
Publications
- Eliza M. Mojduszka and Rachel Everett. Endogenous Consumer Preferences and Knowledge About Nutrition. Selected Paper presented at the Annual Meetings of the American Association of Agricultural Economics, Montreal, Quebec, August 3-6, 2003.
|
Progress 01/01/02 to 12/31/02
Outputs
Our work expands and extends a line of research using discrete choice demand and latent variable models that address shortcomings in current approaches to analyzing the determinants of consumer food choices (see Mojduszka, Caswell, and Harris, 2001). This approach combines both aggregate store-level product data and disaggregate individual consumer-level survey data, and it provides a model of individual consumer utility and demand that is explicitly aggregated to obtain product level demands. It therefore already contains a framework for analyzing aggregate and disaggregate data sources. However, consumer choice of food products may be further conditioned by nutrition information. In 2002, our work concentrated on developing new models with endogenous consumer preferences. To account for this possibility, we assume that consumer choice of food products and nutrition information are correlated, implying a simultaneous system of equations. We incorporate nutrition
information measures in an integrated discrete choice model system of product choice and nutrition information (Mojduszka 2001, Ben-Akiva and Bowman 1998). In this new model, the distribution of consumer utilities depends on both measured and unmeasured individual characteristics. These determine preferences for product attributes (some of which are unobserved) and hence determine demand. The changes incorporated into the new model allow us to estimate three sets of parameters using a nested method of moments algorithm. The first quantifies the effect of measured individual characteristics on tastes for product attributes. The second set measures the importance of unmeasured individual characteristics in determining preferences for product attributes. The third set allows us to estimate the effect of product attributes on the mean utility of a product. In other words, the first two sets give direct evidence on the extent to which the demand parameters can be explained by individual
characteristics. The aggregate data then are used to estimate the additional parameters that determine the relationship between product attributes and the mean utility levels of the products.
Impacts
By integrating a product choice model and a latent variable model of nutrition information as well as all of our data sources, we obtain precise estimates of the demand parameters that are crucial in determining consumer food choices over time. Understanding these determinants is central to the design of effective nutrition information programs and to the design of marketing and promotion strategies by producers, manufacturers, and distributors. The results of the work contribute to precise answers to the question of how consumer information about nutrition effects individual food choices in the market place.
Publications
- No publications reported this period
|
Progress 01/01/01 to 12/31/01
Outputs
We concentrated on creating the integrated database for the project. We have been matching the IRI scanner data with the Nutritional Quality data and the Advertising data. We have also started our work on developing the discrete choice demand models for the selected food categories.
Impacts
Our research project focuses on the important determinants of consumer demand for food products and how they have been changing over the last decade. Understanding these factors is key information for assessing the competitiveness of U.S. agricultural producers and food processors as they choose product designs and marketing strategies; for understanding the benefits and costs of government regulations, such as labeling, intended to influence consumer food choice and, ultimately, public health; and for evaluating the impact of changing consumer demand for food on the agricultural and food sectors of the U.S. economy. The most innovative and unique aspect of the project is that it moves beyond the existing studies to integrate several behavioral models and data sources. As a result, our conclusions will provide important insights into the economic forces that tend to limit the efficient provision and use of nutrition information in consumer choices of foods.
Publications
- No publications reported this period
|
|