When people make decisions about what to eat, they may consider such factors as taste, ease of preparation, nutrition, what others enjoy eating, and price. Many researchers seek to understand what makes some people choose healthful diets while others choose less healthful diets. To conduct these studies, researchers must understand as much of the context of the decisions as possible: What foods are normally enjoyed by an individual and his or her co-diners? How much time does an individual or household have to prepare food? Are there health conditions or other factors that might prompt an individual to value health and nutrition more so than others do? What are an individual’s or household’s budget restrictions?
To conduct these studies, researchers need data to describe those factors that might influence the eating decision. The National Health and Nutrition Examination Survey (NHANES) provides a wealth of information on the dietary intake, medical history, and current health status of its respondents. The 2007-08 cycle of NHANES also includes a module with questions addressing consumer behavior, which will allow a deeper under-standing of the food choices when these data are available. However, the National Research Council points out in its report, Improving Data to Analyze Food and Nutrition Policies, that there are limiting factors when using NHANES for economic analysis. For example, income and asset data as well as time resources that might affect food preparation are not included. Detailed information on neighborhood characteristics and specific levels of participation in food assistance programs are also missing, as are price data (National Research Council, 2005).
The U.S. Department of Agriculture’s (USDA) Center for Nutrition Policy and Promotion (CNPP) has taken the first step in attaching price data to NHANES dietary recall data. CNPP has completed an estimation of the expenditures on foods that were reported in the NHANES 2001-02 as having been consumed by participants. This estimate uses scanner data and assumes all food is prepared at home, but uses convenience products available in the grocery store. CNPP is working with the USDA’s Economic Research Service (ERS) to develop the next phase of attaching price data to NHANES dietary recall data: estimating the prices of food eaten away from home, as well as estimating regional and seasonal variation for some foods. This first step—attaching price data for home-prepared foods to NHANES dietary recall data—has three main goals:
- To attach price data to dietary intake data in order to update the USDA Food Plans (discussed later) and to provide a basis for other price- and nutrient-related research.
- To attach nutrient information to products reported as purchased in the Nielsen Homescan™ consumer panel to provide more information to researchers using these panel data to examine household food purchase behavior.
- To develop an automated process that will allow for updates as new dietary intake, food composition, and price data become available.
The last goal is important because these data are updated regularly. USDA receives price and food composition data annually and dietary intake data every other year. In addition, other researchers could use the process to create prices for their own dietary recall data. This report will focus mainly on the first goal, but the other goals will be briefly discussed.
This release of the price data is the version used in estimating the most recent release of the USDA Food Plans and meets the specific needs of the USDA Food Plans. The methodology was originally developed by Shanthy Bowman for the 1999 Thrifty Food Plan (TFP) (Bowman, 1997). To understand the 2001-02 CNPP Prices Database, it is important to understand how the USDA Food Plans are used and why the data needs may differ from other research projects.
USDA estimates the cost of consuming a nutritious diet, as defined by the 2005 Dietary Guidelines for Americans (U.S. Department of Health and Human Services & USDA, 2005), at four expenditure levels: the Thrifty Food Plan (TFP) and the Low-Cost, Moderate-Cost, and Liberal Food Plans. Each plan is a representative and healthful diet and provides a cost estimate for food. The TFP cost level serves as the basis for the maximum food stamp allotment, and the market basket produced from the TFP is used to adjust the allotment for inflation, using the Consumer Price Indexes for food. The Low-Cost Plan is used by bankruptcy courts to set food budgets and by the U.S. Department of State to set a food allowance for visiting students. The U.S. Department of Defense uses the Liberal Plan to set the Basic Allowance for Subsistence rate for all service members. The Low-Cost, Moderate-Cost, and Liberal Plans are used by divorce courts to set alimony and by the CNPP report Expenditures on Children by Families (Lino, 2007), which is used to set State-level child support guidelines and foster care payments. Because the maximum Food Stamp Program allotment does not vary among the 48 coterminous States, a single set of national average prices must be used in the TFP to estimate the cost of food. In addition, under the regulations of the Food Stamp Program, hot food cannot be purchased and food stamps cannot be used in restaurants. The Basic Allowance for Subsistence is also meant to cover food prepared at home or in the communal dining facility, not food purchased in restaurants. For more information on the Thrifty Food Plan, please see, Thrifty Food Plan, 2006 (Carlson, Lino, Juan, Hanson, & Basiotis, 2007). Information on the other three plans can be found in the report The Low-Cost, Moderate-Cost, and Liberal Food Plans, 2007 (Carlson, Lino, & Fungwe, 2007).
In summary, this release of the prices database was developed for estimating the USDA Food Plans for the 48 coterminous States. As a result, these prices assume one national average price for each food that was reported as having been consumed. We also assume that foods are prepared at home, although convenience foods are considered, including those used to reduce time in food preparation, such as bottled sauces, packaged mixes, canned soups, and frozen foods (including frozen fruits and vegetables). It should be noted that no individual faces national average prices, and unlike some information associated with consumer choice, the price is readily available as the consumer is putting food into the grocery cart. Economic theory suggests that variation in price will have a much larger effect on choices than information that is not readily available to the consumer. For example, there are differences in nutrient quantity between two tomatoes that appear to be identical to the average food shopper. To arrive at the amount of vitamin A in a tomato, several tomatoes are tested and an average is considered. Researchers may be tempted to assume this average vitamin A content could be treated the same as the average price. It cannot, because most consumers are not aware of the differences, and those that are aware cannot tell in the grocery store which tomato is better. However, the consumer can determine which tomato costs less and will recognize that there is variation in the price of tomatoes. Even with these limitations, these data do allow researchers to include prices in an analysis of NHANES data.