The following documentation outlines the methodology used to produce cost estimates for selected snack foods previously published within the USDA, Economic Research Service’s (ERS) Fruit and Vegetable Prices data product. These archived files were last updated in November 2012.

Substituting fruits and vegetables for other snacks 

Because the consumption of snacks among children has increased markedly over the last three to four decades, USDA, ERS examined the effect of replacing one energy-dense snack a day with a fruit or vegetable to determine the likely impact on both households' food budgets and children's caloric intakes. The study used data from Nielsen's 2010 Homescan panel. Households participating in Nielsen's Homescan panel recorded their food purchases at retail stores, including quantities bought, amount of money paid, and date of purchase. USDA, ERS estimated average retail prices, following the same method used in estimating the retail prices and per cup equivalents.   

The study was published in 2012. Results showed that replacing a calorie-dense snack food with a fruit or vegetable could reduce calorie intake without compromising a household’s food budget. More information is available in the 2012 Amber Waves article, Gobbling Up Snacks: Cause or Potential Cure for Childhood Obesity?


Selecting snack foods to price 

Retail stores in the United States offer a wide variety of snack foods. USDA, ERS selected 20 snack foods from among those that children 6–13 years of age reported eating in the 2005–08 National Health and Nutrition Examination Surveys (NHANES). The National Center for Health Statistics, an agency within U.S. Department of Health and Human Services, is responsible for the NHANES. Survey participants report the types and quantities of foods that they eat over 2 nonconsecutive days. The 20 snack foods selected by USDA, ERS based on NHANES data include salty chips and crackers, baked and sweet goods, and frozen treats. These are commonly consumed foods that require little or no preparation. Most are also high in calories, added sugars, fat, and/or sodium, and can be considered less healthy relative to fruits and vegetables.

USDA, ERS also identified and priced 20 fruits and vegetables (both fresh and processed) that are possible replacements for these snack foods. Some of these items, such as fresh apples and bananas, are already commonly consumed by children; others are not. For 12 of the 20 fruits and vegetables, children reported, on average, eating less than 1/2-cup equivalent; this amount approximates a serving as defined in the Dietary Guidelines for Americans and is also similar to the size of many fruit cups sold in supermarkets for snacks and lunch boxes. It was particularly difficult to find vegetable options, as these tend to be consumed by children infrequently and in small amounts. USDA, ERS researchers assumed that sweet potatoes (not commonly consumed by children) might be an acceptable snack alternative for children, as sweet potatoes are easy to microwave and have a sweet taste. Similar reasoning was used to complete the list of fruit and vegetable snack substitutions.


Estimating the retail price of selected snack foods

The next step in USDA, ERS' price analysis was to estimate the average national per-pound price of selected snack foods at retail stores (or per-count, for popsicles and bars) using the 2010 Nielsen Homescan data. Participating households used a scanner at home to record retail food purchases after shopping. These scanners recorded items purchased, quantities bought, amount of money paid, and date of purchase. Purchases at supermarkets, supercenters, club stores, convenience stores, drugstores, farmers' markets, and other types of retail facilities are all included.

The 2010 Homescan data provided limited information about random-weight foods such as individual apples and store-baked muffins. Thus, average retail prices are estimated only for foods such as prepackaged apples and muffins that are sold with a Universal Product Code (UPC), a type of bar code. The 2010 Homescan data used for this analysis provided information on the purchases of 60,648 households in 2010. Sample weights were applied to derive nationally representative estimates of retail food purchases for all households across the contiguous United States in 2010.

National average retail prices were estimated by dividing total expenditures for each snack food by total quantities purchased. Total expenditures were calculated by aggregating data on all brands and package sizes for closely related products across all stores for an entire year. For example, muffins include small, medium, and large blueberry, cranberry, bran, and other sizes and flavors of muffins sold with a UPC. Similarly, apples include prepackaged bags of small and large Red Delicious, Golden Delicious, Granny Smith, Gala, Fuji, and others. This method gives a greater weight to more frequently purchased varieties of a food product.

Calculating aggregate quantities of snack foods purchased by households required converting some quantities into pounds. For example, the Homescan data prices cantaloupes, watermelon, and frozen treats such as popsicles and bars on a count basis, whereas ice cream is priced per fluid ounce. To convert count data on cantaloupes to weight, USDA, ERS used the USDA National Nutrient Database for Standard Reference (SR) to estimate the average weight of a medium cantaloupe at roughly 2.4 pounds, including the weight of the rind and inedible cavity contents. For watermelons, USDA, ERS used data from the USDA, Agricultural Research Service’s (ARS) SR, USDA's Food and Nutrient Database for Dietary Studies, the relative shares of mini and other watermelons from the Homescan data, and a 2005 HortScience journal article, Lycopene Content of Mini Watermelon Varieties Grown at Four Locations to estimate the average weight of a watermelon at about 16.7 pounds. USDA, ERS chose not to convert popsicles and bars from counts to weight since a count seemed like a more reasonable consumption unit for popsicles and bars. For ice cream, USDA, ERS used data from SR showing that a 1/2-cup of ice cream weighed 66 grams (2.3 ounces), yielding a conversion factor of 1 fluid ounce = 0.58 ounces.

Next, average retail prices were calculated as the ratio of total expenditures to total quantities. For example, USDA, ERS estimated that households living in the contiguous United States spent $620.8 million to purchase 627.4 million pounds of apples, yielding an average cost of $0.99 per pound ($620.8 million/627.4 million pounds).


Estimating the price of eating selected snack foods

Some retail food products, such as potato chips and cookies, can be eaten as is. Other foods require the consumer to remove inedible parts or cook the food before eating, resulting in different edible and retail weights. Also, the number of snacks in a pound of food differs across foods. For example, a 16-ounce bag of potato chips might provide 16 snacks, whereas a pound of watermelon might provide 3 to 4 snacks after removing the inedible rind. Comparing retail prices per pound of watermelon (at $0.24/lb) versus a pound of cookies (at $2.73/lb) will not help consumers determine the impact on their food budget.

To convert average retail prices to prices per edible ounce, USDA, ERS used the method described in the February 2011 USDA, ERS report, How Much Do Fruits and Vegetables Cost?, to account for inedible parts such as watermelon rind and cooking yields (weight lost in cooking a sweet potato or a frozen pizza). The data for making these adjustments are available in the USDA, ARS’ SR and USDA report, Food Yields Summarized by Different Stages of Preparation, Agriculture Handbook 102 (AH102). In making these conversions, USDA, ERS defines a food's retail-equivalent weight as:

Retail-equivalent weight = (1 / (1 - inedible share)) / (cooking yield).

According to the AH102 report, a baked sweet potato weighs 78 percent of its raw weight and has an additional refuse of 22 percent upon removal of the skin. To consume 1 ounce of peeled, cooked sweet potatoes, a consumer would have to purchase (1/(1-0.22))/0.78 = 1.64 ounces of sweet potatoes at retail. Similarly, the AH102 report shows that the cooking yield for a frozen pizza is 93 percent. Thus, to consume one ounce of pizza (from frozen to cooked), a consumer would have to purchase 1/0.93 = 1.08 ounces of frozen pizza at retail.

After determining the price per edible ounce, it was necessary to determine the portion size to compare the cost of replacing snacks with fruits or vegetables. USDA, ERS decided to base portion sizes on current consumption patterns using average amounts consumed by children ages 6–13 in the 2005–08 NHANES. Based on the assumption that younger children would consume smaller quantities and older children larger quantities, the analysis is limited to average amounts consumed by children ages 6–13 because differences in quantities consumed would affect the portion size, and therefore the cost per portion.

To determine whether the average amounts consumed were reasonable, USDA, ERS compared them with common portion sizes. For fruits and vegetables, USDA, ERS used 1/2-cup equivalents in USDA's Survey Foods, 2003–04, Food Surveys Research Group as the comparison, since this is similar to a serving in the 2010 Dietary Guidelines for Americans. For 12 of the 20 fruits and vegetables, the average amount consumed was smaller than the 1/2-cup equivalent, resulting in a low price per portion. Since it was assumed that consumers would replace a less-healthy snack with a reasonable amount of the fruit or vegetable, USDA, ERS used the 1/2-cup equivalent as the portion size for the 12 fruits and vegetables consumed in small amounts (that is, whenever the 1/2-cup serving was larger than the average amount consumed). This would safeguard against underestimating the budgetary effect of replacing less-healthy snack foods with fruits and vegetables. For other snacks, average amounts consumed were similar or larger than common portion sizes in the USDA, ARS’ SR.


Estimating the cost of replacing a snack with a fruit or vegetable

Replacing each of the 20 snacks with one of the 20 fruits or vegetables yields 400 possible substitutions. For example, it would cost a household an additional 20 cents to replace a 1.0-ounce portion of cookies with a 5.2-ounce portion of apples. Conversely, a household would save 11 cents if the 5.2-ounce portion of apples replaced a 2.6-ounce portion of fruit Danish pastry. It is not surprising that some substitutions would cost more, while other substitutions would cost less. A household making each of the 400 possible substitutions would save a net total of $7.00 in food costs.


Estimating the caloric effect of replacing a snack with a fruit or vegetable

One of the potential benefits of replacing a calorie-dense snack with a fruit or vegetable is that it could reduce calories consumed. For example, replacing a 1.0-ounce portion of a soft chocolate-chip cookie for a 5.2-ounce portion of apples would reduce caloric intake by 46 calories; replacing the 2.6-ounce apple Danish pastry would reduce intake by 194 calories. Although some substitutions could save more calories than others, a child making each of the 400 possible substitutions would save an average of 126 calories per substitution. 

