Recipient Organization
MONTANA STATE UNIVERSITY
(N/A)
BOZEMAN,MT 59717
Performing Department
(N/A)
Non Technical Summary
In the U.S., about 32% of food is wasted at the consumer level, amounting to an estimated $160- $240 billion value annually. The massive amounts of food waste cause challenges for both the sustainability of the food supply chain and the implementation of consumer-oriented food policies. Due to the lack of appropriate empirical methods, existing research on consumer food waste measurement does not offer comprehensive food waste estimates that are both household-specific and commodity-specific for a nationally representative sample of households. Consequently, food waste has been mostly absent in the broader field of food policy studies that implicate consumer behavior issues. The overall goal of the proposed project is to achieve accurate, comprehensive measures of consumer food waste and apply them to analyze a wide array of consumer-oriented food policies. To realize this goal, the project team will (1) develop and apply novel empirical methods to estimate household-level and food category-specific food waste, based on a household utility maximization framework, (2) assess the overall trend in U.S. consumer food waste and link food waste to household and food product characteristics, (3) provide an easy-to-implement calibration tool that can be used by researchers to calculate food waste using widely available consumer purchase data sets, (4) demonstrate how to use the food-waste estimates to adjust the household- level calculations of the Healthy Eating Index, and (5) investigate whether food-waste adjustments to prior research of food access and environment would lead to different policy conclusions or recommendations.
Animal Health Component
75%
Research Effort Categories
Basic
25%
Applied
75%
Developmental
(N/A)
Goals / Objectives
The overall goal of the project is to achieve accurate, comprehensive measures of consumer food waste and apply them to analyze a wide array of consumer-oriented food policies. Specifically, the project team will (1) develop an empirical framework of household utility maximization and incorporate food waste as a choice variable; design appropriate estimation procedures such as partial identification to tackle the estimation of unobserved food waste;(2) estimate household-level and food category-specific food waste in the IRI Consumer Network and the IRI MedProfiler sample annually from 2009- 2020; assess the overall trend in U.S. consumer food waste and link food waste to household and food product characteristics;(3) calculate the implied actual household food intake in the IRI data and a waste-adjusted Healthy Eating Index, a frequently used dietary healthfulness measure; explain the differences between the indices calculated from waste-adjusted intake quantities and from the purchase quantities and articulate policy implications;and (4) assess the effects of food access and environment on consumer behavior after adjusting for waste; investigate whether better access to retail stores reduces food waste and improves dietary quality measured by actual intake.
Project Methods
This project addresses the methodological challenges in estimating consumer food waste by developing novel empirical methods based on the household utility maximization framework. We consider food waste as a rational outcome that responds to both household-specific factors and product-specific characteristics, and we exploit the heterogeneity of these variables to simultaneously identify food waste for different households and food categories. Moreover, we incorporate a caloric production process that considers food waste as inefficiency in household food production, further enhancing the identification power of our methods. To illustrate how waste can be incorporated into other food policy research, we will apply the newly created methods to analyze two policy issues relating to dietary healthfulness and food access. We also provide an easy-to-implement calibration tool that can be used by researchers to calculate food waste using widely available consumer purchase data sets. Specifically, the project includes the following five tasks: (1) Acquire and prepare the data. The project team will prepare and organize the IRI Consumer Network, IRI MedProfiler consumer health data, and the National Household Food Acquisition and Purchase Survey (FoodAPS), for empirical analysis. The team will also conduct and analyze a nationally representative household survey on food waste behaviors to identify relevant household and product factors that influence the generation of food waste. (2) Develop the empirical estimation procedures. The project team will develop an empirical framework of household utility maximization and incorporate food waste as a choice variable. Two estimation procedures will be performed. The first procedure utilizes partial identification method, which exploits the fact that food waste is bounded between 0% and 100%. The second procedure adopts specific cost functions that describe the cost of reducing food waste and solves for reduced-form solutions of food waste. (3) Estimate and analyze annual food waste. By applying the developed estimation procedures to the IRI Consumer Network and MedProfiler data, the project team will study the trend in the national average food waste in the past decade. We will also perform a reduced-form post-estimation analysis of food waste and its association with major household and contextual factors. The results from this reduced-form analysis can be used to calibrate or impute the amount of food waste for any typical household, given its food purchase information, demographic characteristics, household structure, and the retail environment of its location. (4) Conduct first policy application in evaluating Healthy Eating Index (HEI). We first illustrate how to measure HEI from purchase data by subtracting food waste in the IRI datasets and compare the HEI calculated from our method to the one calculated directly from purchases. We will also generalize the use of our food waste estimates and the implied food intakes in calculating HEI to other datasets such as FoodAPS. With the estimates from the post-estimation analysis, we will impute food waste for each household in the target dataset based on the household's food basket composition and demographic characteristics. We then calculate the amounts of actual intakes for the household and its corresponding HEI. (5) Conduct second policy application in consumer behavior studies. The project team will assess the effects of food access and environment on consumer behavior after adjusting for waste. Specifically, we will incorporate our methods into several existing consumer studies and re-estimate their effects. By incorporating food waste, we will investigate the extent to which food waste is reduced under improved food access and the resulting changes in dietary quality measured by actual intake. The outputs of this project will be evaluated by academic researchers through peer-reviewed publications and conference presentations. The findigs will also be assessed by policymakers and stakeholders nationwide via webinars and ERS reports.