Appendix U

The Regulatory Impact Analysis (Complete)

This appendix presents a regulatory impact analysis (RIA) for the proposed revisions to the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) food packages. This RIA was guided by the overall structure and analytic approach presented in the 2007 Revisions to the WIC Food Package Interim Rule and associated appendix (USDA/FNS, 2007a,b), along with those prepared for the Proposed Rule and Final Rule (USDA/FNS, 2006, 2014a). Unlike WIC RIAs found in the Federal Register, this analysis is not part of the federal rulemaking process nor has it been reviewed by the U.S. Department of Agriculture (USDA), the Office of Budget and Program Analysis, the Office of Management and Budget, or any other federal entity. Accordingly, all estimates should be considered provisional.

An overview of the types of estimates in this RIA is presented in Box U-1. Throughout this appendix, the committee compares the projected costs of the proposed revised food packages (assuming all recommendations are fully implemented) to the projected costs of the current set of food packages (assuming the current regulations remain intact). Estimated costs are based on the forecasted program participation levels, calculated redemption of each food package item, and inflated prices of each food package item. Unless otherwise noted, cost differences describe how much more the revised food packages would save or cost program-wide, as compared to the current food package.

A number of factors affect the total food costs to the WIC program, including interstate variation in food prices, caseload composition, and cost-containment practices (USDA/ERS, 2005). Accurate estimates of total food costs, therefore, need to be based on data that capture this variability. To project the cost effects associated with specific changes to the food packages, data need to provide insight into how each individual food item within each specific food package contributes to the total food costs. At the present, this level of granularity does not exist in data sources representative of the entire WIC program.

The data and assumptions used throughout this analysis are primarily based on WIC agency and administrative data sources. These include the participation data from the WIC Participant and Program Characteristics 2014: Food Package Report (USDA/FNS, 2016a), national participation levels presented on the USDA’s Food and Nutrition Services (USDA-FNS) website (USDA/FNS, 2016b), redemption data from six de-identified state agencies provided to the committee by USDA-FNS, redemption data provided to the committee from six individual state agencies, and a 2014 report detailing redemption in three states as they transitioned to the electronic benefit transfer (EBT) food benefit issuance method (USDA/ERS, 2014).

The committee estimated total food costs and cost differences by integrating the available WIC data sources, some of which were from a relatively small number of state agencies. This analysis assumes the data are representative of the WIC population at large. The committee, however, recognizes that the cost estimates calculated are not identical to costs derived from FNS administrative data. For FY2015, the committee estimated the per-participant food cost to be $37.09¹ per month, based on assumptions about prices per unit, substitutions of allowable options, and redemption of each food package item. USDA-FNS, however, reported the FY2015 per-participant food costs to be $43.37 per month. The committee was unable to discern what component(s) of its analysis led to this difference, because some of the data available to the committee were de-identified and the majority of the data represented only a small portion of states and territories. The estimates presented in this RIA should be interpreted in context of these limitations.

__________________

¹ A per participant cost of $37.27 per month was presented previously in this report. This higher estimate reflects the $1 increase in women’s cash value voucher, which was implemented in FY2016 (from $10 to $11). The committee used this monthly per participant cost to assess the cost neutrality of the revised food packages, because it reflects current regulations. The $37.09 estimate, in contrast, reflects the regulations existed in FY2015.

The magnitude of the cost effects and stability of the projections in this analysis are largely defined by the timeframe evaluated. The exact timing of when the revised food packages will be implemented, however, depends on a number of factors, including when the federal regulations are put forth and how quickly state agencies can operationalize the changes. For the purposes of this analysis, the committee assumes that the earliest implementation of the revised food packages could occur would be April 1, 2018, approximately 15 months after the release of this consensus report. Food cost estimates for both the current and revised food packages are projected through FY2022. This RIA, therefore, encompasses a 54-month period.2

The committee assumes within each state agency all proposed food package changes will be fully implemented at one time, but that across the WIC program, state agencies may begin offering the revised food packages at different times. As such, two sets of cost projections were generated: unadjusted and phased in. Unadjusted estimates assume that all agencies fully implement the revised food packages on April 1, 2018. Phased-in estimates, in contrast, account for states implementing the revised food packages at different times. The phased-in scenario assumes that states with EBT systems operational as of the August 2016 EBT Detail Status Report (USDA/FNS, 2016c) would be early implementers of the proposed food package changes (i.e., those implementing the changes on April 1, 2018).³ The remaining state agencies would have up to 18 additional months to implement the proposed food package changes (i.e., implemented by October 1, 2019). As evidenced by the final implementation dates from the previous food package revision (USDA/FNS, 2012), most state agencies instituted changes on the regulatory deadline, rather than before or after. This analysis assumes the same will occur for these proposed revisions. Accordingly, the phased-in cost differences between the current food packages and proposed revised food packages are 33.3 percent of the unadjusted cost differences for FY2018 (6 months) and FY2019 (12 months), and 100 percent of the unadjusted differences thereafter. The dates and rates of implementation used in this RIA are not intended to be prescriptive or to be the committee’s recommended timeline for implementation, but rather are the committee’s informed assumptions necessary for this analysis.

__________________

² The fiscal year starts October 1. April 1, 2018, is halfway through FY2018.

³ This estimate was calculated by identifying agencies that have implemented EBT statewide as of August 2016 (USDA/FNS, 2016c) and determining what proportion of participants are served by those agencies from total participation administrative data (USDA/FNS, 2016b). For FY2015, EBT states serve 34.3 percent of WIC participants. Given this, the assumption for the phased-in estimate is that one-third of participants would be served by an “early implementer” state agency. This assumption only affects the phased-in estimates, not the unadjusted estimates.

As Table U-1 highlights, the proposed revisions largely leave the current structure of the food packages unchanged. There are, however, three important considerations:

• Food package V is currently a single food package prescribed to both pregnant women and partially (mostly) breastfeeding women, up to 1 year postpartum. Under the proposed revisions, it would be split into two distinct food packages, V-A for pregnant women and V-B for partially (mostly) breastfeeding women.
• Revised food package I-BF/FF-A for a partially (mostly) breastfed infant, age 0 to less than 1 month, may not ultimately be operationalized as separate from food package I-BF/FF-B for a partially (mostly) breastfed infant, age 1 to less than 4 months. Under the proposed revisions, infants issued food package I-BF/FF-A could potentially be prescribed the same amount of formula as food package I-BF/FF-B, on a case-by-case basis. In implementing the proposed revisions, it may be decided to consolidate the two food packages and align the naming structure across the infant food packages (i.e., I-BF/FF-A for infants ages 0 to less than 4 months, I-BF/FF-B for infants ages 4 to less than 6 months). The committee chose to keep the revised food package I-BF/FF-A separate in this analysis because the proposed revision maintains the provision that issuing infant formula to breastfed infants in the first month of life should not be standard. The committee expects that infants

a Despite revised food packages I-BF/FF-A and I-BF/FF-B having the same maximum (“up to”) amount for infant formula, the two food packages remain separate in this analysis. The committee supports issuance of infant formula to food package I-FF/BF-A recipients only on a case-by-case basis after a breastfeeding assessment. Accordingly, the committee assumes the average quantity of infant formula prescribed in food package I-BF/FF-A to be less than the average amount prescribed to I-BF/FF-B recipients. Food package I-BF/FF-A may not ultimately be operationalized as separate from food package I-BF/FF-B for partially (mostly) breastfed infants, ages 1 to less than 4 months.

^b The parenthetical “mostly” is determined by the amount of formula provided to the infant through the WIC program. Women and their infants are classified as partially (mostly) breastfeeding if the infant’s formula prescription does not exceed the monthly maximum allowance for the age-specific BF/FF food packages.

^c Food package III recipients are prescribed food packages appropriate to their age and physiological state. There are specific food packages for women, infants, and children within food package III that correspond to specific food packages I, II, IV, V, VI, and VII.

^d Participants prescribed food package III have one or more medically documented qualifying conditions that require use of a WIC formula. Recipients are eligible to receive WIC formula (i.e., infant formula, exempt infant formula, and WIC-eligible nutritionals) and supplementary foods.

^e The parenthetical “minimally” is determined by the amount of formula provided to the infant through the WIC program. Women and infants are classified as partially (minimally) breastfeeding if the infant’s formula prescription exceeds the monthly maximum allowance for the age-specific BF/FF food packages but does not exceed the full amount allowed in the age-specific fully formula-fed food packages. Women who are minimally breastfeeding less than 6 months postpartum receive food package VI, along with breastfeeding support. Women who are minimally breastfeeding and with infants more than 6 months of age that receive more formula than is allowed for a partially breastfed infant do not receive a food benefit. These women, however, are eligible to continue to receive other benefits through WIC, such as breastfeeding support, nutrition education, and health and social services referrals (USDA/FNS, 2013). This group of participants is included in one of the committee’s assumption regarding the revised food package cost estimates, and is therefore included in the above table.

issued the revised food package I-BF/FF-A would continue to be prescribed only small quantities of infant formula, markedly less than what is prescribed to food package I-BF/FF-B recipients.

• Food package “N/A” is listed as a food package because it is included in an assumption about how participation is expected to change with the revised food packages. Women who are minimally breastfeeding more than 6-months postpartum do not receive a food benefit (referred to herein as “food package N/A”), but are still eligible to continue to receive breastfeeding support, nutrition education, health and social services referrals, and other program benefits (USDA/FNS, 2013a). These program participants do not contribute to the total food costs of the program. The committee assumes that preventing premature classification of mother–infant dyads as fully formula fed is likely to shift dyads toward the partially (mostly) breastfeeding food packages. This projected participant shift includes women more than 6-months postpartum.

ACTION

Nature

This RIA was conducted by the Committee to Review WIC Food Packages as part of a National Academies of Sciences, Engineering, and Medicine (the National Academies) consensus study.

Need

The WIC program serves low-income, nutritionally at-risk pregnant women, breastfeeding and nonbreastfeeding postpartum women, and children younger than 5 years of age. The program provides participants with nutritious supplemental foods, breastfeeding support, nutrition education, and referrals to health and social services. Supplemental foods are intended to be appropriate for the participants’ life stage and are designed to provide specific nutrients that current nutrition research indicates are lacking in the diets of WIC’s target population groups (7 C.F.R. § 246). Given the nature and funding of the program, the supplemental foods must provide these specific nutrients to participants in a manner that is both effective and economical. To ensure this, regulatory language explicitly defines the specifications and amount of each authorized food item.

In 2006, an Institute of Medicine (IOM) committee offered its first set of recommended changes to the quantity and types of foods included in the WIC food packages (IOM, 2006). The recommended revisions aligned the food benefit with the 2005 Dietary Guidelines for Americans (DGA) and

the American Academy of Pediatrics’ (AAP’s) infant feeding guidelines, and led to the first major revisions to the WIC food packages since the inception of the program. A decade has passed since the IOM’s first set of recommendations and with it has come advances in nutrition research and insight into the effects of WIC generally and the food package changes specifically. Congress asked the USDA, which administers the WIC program, to review and update the food packages to be consistent with the current (2015–2020) DGA.4 To accomplish this, USDA-FNS asked the National Academies to recommend revisions to the food packages to align with current research and dietary guidance, be culturally suitable to the increasingly diverse WIC population, be relatively cost-neutral compared to current food packages, be efficient for nationwide distribution, and be nonburdensome to state and local agencies that administer the program.

Affected Parties

The proposed changes to the food packages affect a broad range of individuals and entities associated with the WIC program, including, but not limited to, USDA-FNS; the 90 agencies that administer WIC and their associated staff; authorized vendors; food producers, manufacturers, and distributors; and program participants.

EFFECT

The proposed changes to the food packages are consistent with the DGA, current recommendations of the AAP, and other recently-available scientific evidence. Revisions also consider the Dietary Reference Intakes (DRIs). The proposed changes take into account the nutrient and food group intake, health status, and cultural needs of the program participant population, while simultaneously considering the efficiency and efficacy of program operations and administration. The analyses and sections that follow describe the potential economic impact of the proposed revisions.

Background

Piloted by the USDA-FNS in 1972, WIC was established as a permanent program in 1975 (P.L. 94-105, 1975). While its mission has remained the same—to “safeguard the health of low-income women, infants, and children up to age 5 who are at nutritional risk”—WIC’s goals have evolved over time. Goals currently include promoting and supporting successful long-term breastfeeding; providing WIC participants with a wider variety

__________________

⁴ References to the DGA in this appendix are specific to 2015-2020 unless otherwise noted.

of foods, including vegetables, fruits, and whole grains; and providing WIC state agencies greater flexibility in prescribing food packages to accommodate cultural food preferences of WIC participants (USDA/FNS, 2014b). WIC supports the national health goals of Healthy People 2020, specifically those related to birth weight, childhood and adult weight, and breastfeeding prevalence (NWA, 2013; HHS, 2015).

WIC is funded through discretionary grants that are appropriated by Congress on an annual basis. Grants are administered to 90 WIC state agencies, which include agencies in each of the 50 states, the District of Columbia, 5 territories, and 34 Indian Tribal Organizations (USDA/FNS, 2015a). WIC is not an entitlement program and is not guaranteed unlimited funds. The maximum number of eligible women, infants, and children who can be served has the potential to be limited by the amount of funding appropriated in a given year. In 2014, the WIC program served an average of approximately 8.2 million women, infants, and children on a monthly basis through its 1,900 local agencies in 10,000 clinic sites (USDA/FNS, 2015a). Approximately 50 percent of infants and 40 percent of pregnant women in the United States benefit from WIC services (USDA/FNS, 2015a; personal communication, J. Hirschman, USDA-FNS, October 15, 2014).

Participants in the WIC program must meet eligibility criteria for life stage, income, and nutritional risk in order to receive benefits. Specifically, applicants must be: (1) women who are pregnant and up to 6-months postpartum, or, if breastfeeding, 1-year postpartum; infants or children up to 5 years of age; (2) at or below 185 percent of federal poverty guidelines or enrolled in Temporary Assistance for Needy Families, Supplemental Nutrition Assistance Program, or Medicaid; and (3) at nutritional risk (e.g., anemia, obesity, underweight, high-risk pregnancy). WIC participants receive a range of benefits, including nutrition education, breastfeeding support, health and social services referrals, and nutritious supplemental foods. The supplemental foods provided in the WIC food packages are designed to provide specific nutrients determined to be lacking in the diets of the WIC target population (7 C.F.R. § 246).

The composition of the food packages offered to participants remained relatively unchanged throughout the first 30 years of the program, aside from enhancing the breastfeeding food package in 1992. In 2006, an IOM committee proposed the first significant revisions to the WIC food packages, at the request of USDA-FNS (IOM, 2006). Most, but not all, of the 2006 IOM report recommendations were fully implemented, which dramatically changed the food benefit offerings. While most were instituted by fall of 2009 in accordance with the Interim Rule (USDA/FNS, 2007c), some changes have taken up to 6 years to be implemented.

A decade has passed since the first set of recommended changes to the WIC food packages. In that time, the WIC population has changed in

ways that reflect demographic changes across the United States, which have included declining birthrates (CDC, 2015) and population growth coming from immigration, temporary and permanent residency, and other population shifts (DHS, 2014). Shifts have also occurred in the public’s attention to food production, sustainability, and food systems. Since 2006, the DGA, which serve as a foundation for the WIC food packages, have undergone two revisions. Changes have also been seen in the administration of the WIC program. During this period, many state agencies have made significant updates to their Management Information Systems, moving toward more efficient, Web-based technologies. Some states have migrated to an EBT card to issue food benefits, which is required of all state agencies by 2020. Research has been conducted that provides insight into the effect of WIC generally and the food package changes specifically. Given the changing landscape of participants, advances in science, and lessons learned from the previous revision to the food packages, an opportunity exists to further refine and enhance the WIC food packages.

The USDA-FNS charged the National Academies’ current Committee to Review WIC Food Packages to conduct a two-phase evaluation of the WIC food packages and develop recommendations for revising the packages to be consistent with the current DGA and to consider the health and cultural needs of a diverse WIC population while remaining cost neutral, efficient for nationwide distribution, and nonburdensome to administration in national, state, and local agencies.

Summary of Proposed Changes and Benefits

In 2014, USDA-FNS contracted with the National Academies to conduct a comprehensive examination of the WIC food packages in relation to the best science available on food and nutrient intake relative to needs, food security and operational factors, and make recommendations for future improvements in the types and quantities of foods provided. The food package revisions recommended in this report reflect the consensus of an expert committee. Compared to the current food packages, the proposed revisions do the following:

• Provide supplemental quantities of foods across food packages. The committee’s concept of the word supplemental, as described in Chapter 6 of this report, is that the amounts of nutrients and food groups in the WIC packages should provide a moderate proportion of an individual’s requirement for a particular nutrient or recommended amount of a food group. The current food packages provide widely varying proportions of required nutrients (5 to 400 percent of DRIs) and DGA food groups (0 to 190 percent

of recommended intake). In developing the revised food packages, the committee sought to balance the food package offerings. While it did not create or apply a strict definition of “supplemental,” the committee considered that a supplemental amount was less than 100 percent of recommended amounts of nutrients and food groups. Therefore, adjustments were made to the extent permitted by marketplace options.5 The primary exception to this is for fully formula-fed infants less than 6 months of age (food packages I-FF-A and I-FF-B), as no other household food is appropriate for substitution in instances of formula insufficiency. For these food packages, the committee emphasizes the need to tailor the food package to meet the specific needs of the infant.

• Align the food packages with the DGA. Analyses presented in the DGA indicate that vegetable consumption across all age groups (ages 2 years and older) and fruit consumption among adolescents and adults do not currently meet the recommended intake ranges. To support the recommended dietary shift toward more vegetables and (mainly whole) fruits put forth in the DGA, the proposed food package revisions provide larger cash value vouchers (CVVs) to children and women and allow for additional CVV value to be substituted for juice.⁶ The DGA also report that average seafood intake falls well below what is recommended for all age groups. To support the DGA’s recommendation to increase seafood intake, the revised food packages include fish in all children’s and women’s food packages. Furthermore, DGA analyses also indicate that while most age groups’ consumption of total grains is within the recommended intake ranges, average refined grain consumptions exceeds

__________________

⁵ One of the challenges the committee faced in developing revised food packages that provided supplemental quantities of foods was the standard container sizes of products currently available on the market. Infant cereal, as will be discussed later in this appendix, serves as an illustrative example. The committee aimed to provide more infant cereal to fully breastfed infants in food package II than to their partially breastfed and fully formula-fed counterparts, because of greater need for certain micronutrients from complementary foods. Infant cereal, however, is typically available in 8- and 16-ounce containers, which would provide infants with 50 and 100 percent of recommended infant cereal intake, respectively. As such, either all food package II recipients would receive the same amount of infant cereal at a supplemental quantity (i.e., 8 ounces per month = 50 percent of recommended intake), or fully breastfed infants would receive 100 percent of the recommended amount of infant cereal. While the revisions better align with the concept of supplemental, it was not always possible to provide all nutrients and food groups in quantities less than 100 percent of the recommended amounts across all food packages.

⁶ The DGA currently apply to individuals ages 2 years and older. The revised food package IV-A for children ages 1 to less than 2 years, was modeled after the revisions to food package IV-B for children ages 2 to less than 5 years. Proposed revisions to the infant food package are discussed in the next section.

Rationale and benefit The current WIC policy guidelines state that the maximum monthly allowance of infant formula should not be used as the standard for issuance unless the mother is not breastfeeding the infant at all (USDA/FNS, 2016d). The committee, however, recommends formula prescriptions be tailored to the individual needs of the infant. This proposed revision would give clinics greater flexibility to tailor the infant formula quantities prescribed to all infants who are not fully breastfed. Aside from food package I-BF/FF-A (previously described), the maximum quantities are unchanged.

Food Package II—Infants Ages 6 to Less Than 12 Months

Proposed revision Allow tailoring of all prescribed infant formula quantities.

• Similar to food package I, across all infant food packages, the maximum formula quantities should be considered “up to” amounts. All formula prescribed to infants should be tailored to the individual needs of the infant.
• The amount of formula that can be provided in food package II-BF/FF is up to 312 fluid ounces (monthly maximum allowances of 315 fluid ounces reconstituted liquid concentrate, 338 fluid ounces ready-to-feed, or 384 fluid ounces reconstituted powder).
• The amount of formula that can be provided in food package II-FF is up to 624 fluid ounces (monthly maximum allowances of 630 fluid ounces reconstituted liquid concentrate, 643 fluid ounces ready-to-feed, or 696 fluid ounces reconstituted powder).

Rationale and benefit Rationale and benefits of allowing tailoring of the prescribed infant formula quantities are the same as for younger infants. See “Food Package I—Infants Under 6 Months” for details.

Proposed revision Reduce the quantity of infant cereal prescribed in food package II.

• Food package II-BF recipients are prescribed 16 ounces of infant cereal per month.
• Food packages II-BF/FF and II-FF recipients are prescribed 8 ounces infant cereal per month.

• The amount of infant cereal currently prescribed in food package II provides approximately 150 percent of the amount recommended by the AAP (AAP, 2014). Reducing the amount of infant cereal prescribed in food package II aligns with the concept of supplemental.
• Infant cereals are a favored first food and are good sources of zinc and iron. Because fully breastfed infants are not likely to receive zinc and iron from infant formula, food package II-BF recipients are prescribed more infant cereal per month compared to their partially (mostly) breastfed and fully formula-fed counterparts.

Proposed revision Reduce the quantity of jarred infant food vegetables and fruits provided to fully breastfed infants; offer CVV substitution option to all food package II recipients.

• Food package II-BF recipients are prescribed 128 ounces jarred infant food vegetables and fruits per month.
• All food package II recipients can substitute a $10 or $20 CVV for half or all of the jarred infant food vegetables and fruits, respectively.
• The CVV can be used to purchased vegetables and fruits in all forms allowed by the state agency (i.e., fresh, frozen, canned) except dried.

Rationale and benefit

• The IOM committee (2006) that put forth the first set of recommended food package changes did not provide nutrient-based rationale for the selected quantity of jarred infant food vegetables and fruits for food package II-BF recipients. Rather, it stated, “to encourage or promote full breastfeeding, the recommended amounts of [infant] food fruits and vegetables are more generous for fully breastfed infants than other infants” (page 103).
• The AAP does not provide specific intake recommendations for infant food vegetables and fruits, but it does acknowledge they are useful for transitioning infants to solid foods (AAP, 2014).
• Public comments indicated that 256 ounces of jarred infant food vegetables and fruits per month can exceed the need of the infant.
• The reduction of jarred infant food vegetables and fruits brings the fully breastfed infant’s food package in better alignment with the concept of supplemental. Other incentives for full breastfeeding have been included in other aspects of the mother–infant dyad’s

• The AAP (2014) recommends provision of 1 to 2 ounces of meat per day, which corresponds to 30 to 60 ounces per month. Currently, food package II-BF provides 130 percent of the maximum of this recommended range.
• Redemption of jarred infant food meat is particularly low; published data and public comments indicate that it is not a preferred infant food.
• The committee considered, but could not identify nutritionally equivalent but preferred alternatives to jarred infant food meat suitable for inclusion in the WIC food packages. For these reasons, jarred infant food meat is retained in food package II-BF, at a reduced quantity.
• The option of substituting canned fish for some of the jarred infant food meat provides participants with a choice that better meets their cultural needs and personal preferences, while still providing key nutrients and within cost-containment parameters. Canned fish

• The role of juice in the food package has historically been to provide a source of vitamin C (which overlaps with one of the roles of the CVV) and is a convenient means of providing fruit to children.
• The amount of juice in the current packages for children provides more than 100 percent of the lower end of the AAP recommended limit of 4 ounces per day (AAP, 2014). The reduction of juice, therefore, better aligns with the concept of supplemental.
• The AAP (2014) recommends that most fruit intake should be from whole fruit because whole fruit also contributes fiber and other plant-based compounds that are removed during processing. The DGA include a recommendation that at least half of fruit intake should be from whole fruit, and state that most individuals in the United States “would benefit from increasing intakes of fruit, mostly whole fruit” (USDA/HHS, 2016). The option for CVV substitution aligns with both the AAP and DGA recommendations and allows recipients to select from options that may better meet their cultural needs and personal preferences.
• All juice offered through the WIC program (across food packages) would be 64 fluid ounces, decreasing vendor burden and streamlining options across food packages.
• The cost savings from the reduction of juice allows for other revisions across the food packages.

Proposed revision Reduce the amount of milk prescribed to children; specify substitution options.

• Food package IV-A recipients are prescribed 12 quarts of milk per month. Food package IV-B recipients are prescribed 14 quarts of milk per month.
• Children can substitute up to 4 quarts of milk per month, choosing either to substitute 2 quarts of yogurt for 2 quarts for milk or to substitute 1 pound of cheese and 1 quart of yogurt for 4 quarts of milk.
• Soy-based yogurt substitute and soy-based cheese substitute are allowable yogurt and cheese substitution options, respectively, for individuals with lactose intolerance, a milk allergy, or who consume a vegan diet. Soy-based yogurt and cheese substitutes must meet calcium and protein specifications (described later).

Rationale and benefit

• Milk in the current food packages provides what the committee considered a greater than supplemental amount of dairy (e.g.,

• The amount of peanut butter prescribed in the current food packages provides 167 percent of recommend amounts of the protein food subgroup of nuts, seeds, and soy, and the amount of legumes provides approximately 177 percent of recommended amounts of beans and peas for children. Both items exceed what the committee considers to be a supplemental amount.
• The proposed revision specifies that state agencies must authorize both canned and dried legumes. According to the 2015 Food Policy Options report, 85 percent of WIC state agencies have already authorized canned legumes (USDA/FNS, 2015b). Therefore, the committee considered this change an administratively feasible strategy with the potential to promote redemption and consumption.
• Authoritative groups recommend consumption of 1.0 to 1.7 ounces of lower-mercury fish per day by children ages 1 to 4 years (AAP, 2014; FDA/EPA, 2014; USDA/HHS, 2016). Intake of seafood was either low or uncommon across subgroups of WIC participants. As such, the committee considered it important to include fish in additional food packages as costs allowed.
• In the revised food package IV, legumes, peanut butter, and fish are rotated on a quarterly basis in order to provide supplemental amounts and amounts that better align with participant preferences. Although the amount of fish offered in the food packages is small compared to DGA recommended amounts of seafood (19 percent for children), it is a significant improvement over the current packages that offer no fish. It also introduces this food into participants’ diets.
• In the revised food packages, canned fish may be water packed, and may include the same sauces and flavorings that are currently allowed as a state option. Oil packed is no longer authorized. Water-packed varieties are more nutrient dense because water-packed fish is lower in energy but contains the same amounts of key nutrients per serving.

Proposed revision Increase the CVV. Food package IV recipients are prescribed a $12 vegetable and fruit CVV per month.

Rationale and benefit

• The CVV offers participants the most flexibility to meet their cultural food preferences and provides them with access to two food

groups (vegetables and fruits) and nutrients (potassium and fiber) for which intakes were inadequate.

• The current amount of the CVV permits participants to purchase less than one serving of fruits or vegetables per day. Based on the committee’s composite cost for vegetables and fruits most commonly purchased among WIC participants,7 $23 would be required for individuals who consumed a 1,300-kcal diet to meet half of the recommended intakes of vegetables and fruits.
• The committee increased the CVV across food packages to the extent possible within cost-neutral restrictions.

Food Package V-A—Pregnant Women

Proposed revision Create separate food packages for pregnant and partially (mostly) breastfeeding women.

• Pregnant women are issued food package V-A, which is different from food package V-B for partially (mostly) breastfeeding women, up to 1-year postpartum.

Rationale and benefit One of the committee’s objectives with the revised food packages was to promote and incentivize any level of breastfeeding. To do this within cost-neutral parameters, food package V was divided into two distinct food packages.

Proposed revision Reduce the amount of juice prescribed; offer a CVV substitution option.

• Pregnant women are prescribed 64 fluid ounces of juice per month.
• A $3 CVV can be substituted for the 64 ounces of juice.

Rationale and benefit The rationale and benefits are comparable to those presented for the children’s food package. See “Food Package IV—Children Ages 1 to Less Than 5 Years” for details.

Proposed revision Reduce the amount of milk prescribed; specify substitution options.

• All women are prescribed 16 quarts of milk per month.

__________________

⁷ Based on the vegetable and fruit composite ($0.55 per cup-equivalent) applied in this report, which considered the vegetables and fruits most commonly purchased by WIC participants in Massachusetts, Texas, and Wyoming.

• Legumes are key sources of many nutrients (potassium, fiber, and folate) for which intakes below recommended amounts were prevalent across participant subgroups. Therefore, legumes are a valuable component of the WIC food packages. Provision of 2 pounds of dry legumes (or eight 15- to 16-ounce cans) every 3 months brings the amount prescribed closer to what is considered supplemental for women: between 47 and 59 percent of the DGA recommended amount.
• The amount of peanut butter in the current food packages for women provides up to 168 percent of recommended amounts. In the proposed revisions to the food packages, the same quantities of peanut butter currently provided every month will still be provided but only once every 3 months, which brings the prescription in better alignment with the concept of supplemental.
• Authoritative groups recommend consumption of 1.0 to 1.7 ounces of lower-mercury fish per day by pregnant and breastfeeding women (FDA/EPA, 2014; AHA, 2015; USDA/HHS, 2016).
• Inasmuch as intake of seafood was either low or uncommon across subgroups of WIC participants, the committee considered it important to include fish in additional food packages as costs allowed. Although the amount of fish offered in the food packages is small compared to DGA recommended amounts of seafood, it is a significant improvement over the current packages that offer no fish.
• Canned fish may be packed in water and may include the same sauces and flavorings that are currently allowed as a state option.

• At present, flavored milk is offered in 6 percent of states and 40 percent of Indian Tribal Organizations (ITOs), which together cover 3 percent of WIC participants (USDA/FNS, 2015b). At least one ITO is removing flavored milk from its food list to align with CACFP policy (personal communication, D. Tipton, Chickasaw Nation WIC, July 2016). Therefore, the recommendation is not expected to cause a significant disruption in administration of food packages nationally.

Proposed revision Increase federal minimum vegetable stocking requirement; increase number of vegetable and fruit forms states must authorize.

• Require WIC vendors to stock a minimum of three varieties of vegetables and two varieties of fruits.
• States must authorize fresh and at least one additional form (frozen, canned, and/or dried) of vegetables and fruits.

Rationale and benefit

• The committee foresees the need to provide participants with a greater variety of options, given the larger CVVs and opportunities to substitute juice and jarred infant food vegetables and fruits with CVVs.
• The proposed revision seeks to enhance the amount and forms of vegetables and fruits available in WIC authorized stores, while minimizing burden on small vendors.
• The increased vegetable requirement is intended to encourage participants to prioritize the use of the CVV for vegetables.

Proposed revision Increase soy substitution options; specify nutrient profiles of soy options.

• Soy-based yogurt and cheese are authorized yogurt and cheese substitution options, respectively, for individuals with lactose intolerance, a milk allergy, or who consume a vegan diet.
• Soy-based yogurt must contain at least 250 mg of calcium and 6.5 grams of protein per 8 ounce serving.
• Soy-based cheese must contain at least 250 mg of calcium and 6.5 grams of protein per 1.5-ounce serving.
• In addition to current standards for authorized soy-based beverages, the total sugars content should be as low as possible, not to exceed 12 grams per 8-ounce serving.

• In addition to meeting current nutrient specifications, all ready-to-eat cereals on a state agency’s authorized food list must adhere to the “whole grain-rich” criteria as outlined by USDA-FNS for the School Lunch Program (USDA/FNS, 2014c).

Rationale and benefit

• The committee’s analysis of NHANES data indicated that intakes of whole grains continue to be poor and intakes of refined-grains

excessive across the WIC subgroups studied. Focusing on whole grains in the WIC food packages may also improve acceptability of whole grains for the longer term. This change increases the amount of whole grains in the food packages. This change also aligns WIC specifications with those of CACFP (USDA/FNS, 2014c).

• Whole grain cereal options have expanded substantially since the last review of the WIC food packages, removing marketplace availability as a barrier to access. Launches of whole grain products including cereal (wheat and other grains) doubled between 2006 and 2011 (Oldways, 2015).
• Two large national manufacturers of RTE breakfast cereals produce at least 14 different types of RTE cereals that meet the current WIC whole grain criteria (which is similar to the whole grain rich criteria), including 4 gluten-free whole grain varieties.
• The previous committee to review WIC food packages (IOM, 2006) recommended that all breakfast cereals provided through WIC be whole grain (that is, the grain component is at least 50 percent whole grain) to align with the 2005 DGA. Since 2005, the DGA have consistently specified that at least half of grain intake should be from whole grains. The DGA specifically state that intake of refined grains should be limited, although individuals (particularly women capable of becoming pregnant) who consume all grains as whole should include some sources fortified with folic acid as a means of preventing neural tube defects. RTE, fortified whole grain cereals are an example such a source (USDA/HHS, 2016).

Proposed revision Modify the specifications for canned fish.

• Authorized canned fish may be packed in water. Pack may include bones or skin. Added sauces and flavorings, such as tomato sauce, mustard, lemon, are authorized at the state agency’s option. May be regular or lower in sodium content. Oil-packed is no longer authorized.

Rationale and benefit

• Water-packed varieties are more nutrient dense because water-packed fish is lower in energy but contains the same levels of key nutrients per serving.
• Few states currently offer oil-packed fish. The effect of this specification, therefore, is expected to have little effect on participant choice.

The anticipated effects of the proposed food package revisions on four broad stakeholder groups—the USDA/federal government, state and local agencies, vendors and industry, and WIC participants—are summarized in Table U-2. The overall projected net cost difference are also noted; a detailed evaluation of the cost effects is presented in the section that follows Table U-2. The direction and magnitude of the presented cost differences are dependent on the committee’s assumptions. The cost effects of alternative assumptions are evaluated in the “Uncertainties” section of this appendix.

COST

Proposed Revisions to the Benefit

Unadjusted Estimates of Food Costs and Total Cost Differences

Table U-3 presents unadjusted food cost estimates for the current and revised food packages. The total unadjusted food costs from FY2018 through FY2022 are estimated to be $17.7 billion (averaging $3.93 billion per year) for the current food packages and $17.4 billion (averaging $3.87 billion per year) for the revised food packages. Over the course of FY2018 through FY2022, the proposed revisions are projected to lead to a total unadjusted cost savings of $263 million, as compared to the current food packages.

Phased-in Cost Differences

Table U-4 presents the phased-in cost differences between the current and revised food packages, from FY2018 through FY2022. Assuming phased-in implementation across the WIC program inherently decreases the projected total cost savings. The total phased-in cost savings for FY2018 through FY2022 are approximately $42 million less than the total unadjusted cost savings ($220.4 million versus $262.8 million). The estimated cost differences not only reflect changes to the type and quantity of items in the specific food packages, but also the proportion of participants who are prescribed each food package. The cost savings in food package I in FY2021 and FY2022, for example, is driven by the anticipated 5 percent shift of fully formula-fed mother–infant dyads to the partially (mostly) breastfeeding participant categories, due to the incentives incorporated in to the revised food packages (see the “Participation” subsection in the section “Cost Estimate Methodology” for additional details about this assumption).

* The projected net cost differences described in this table are contingent on the committee’s assumptions, which are detailed in the “Cost” section of this appendix. The magnitude and even the direction of some of the cost differences can change with different assumptions that would still be considered reasonable. The cost effects of various different assumptions are evaluated in the “Uncertainties” section of this appendix.

To determine the source(s) of the projected cost savings of the revised food package, the committee evaluated the total costs of each food package item category within the current and revised food packages. The total costs of each of the food items presented in Table U-5 include assumptions about substitutions and allowable options within each category and assumptions about changes in redemption rates (see “Cost Estimate Methodology” for additional details). Across the food package items, the total cost savings are larger than the total added costs, resulting in estimated total cost savings of the revised food packages as compared to the current food packages. Two major sources of cost differences between the current and revised food packages are juice and the CVV. In the current food packages, all women and children are prescribed juice. The reduction of total juice in the revised food packages results in a total phased-in cost savings of approximately $627 million over the course of FY2018 through FY2022, as compared to the current food packages. In contrast, increasing the value of the CVV for all women and children in the revised food package leads to an estimated total phased-in cost increase of approximately $780 million over the course of FY2018 through FY2022.

Major Cost Differences of Food Package Items Within Each Food Package

To further explore the sources of the cost differences, the committee evaluated the total phased-in cost differences of each food package item within each food package. Table U-6 presents each food package item revision that resulted in a total phased-in cost difference of at least $25 million in cost saving or increases over the course of FY2018 through FY2022, as compared to the total cost of the corresponding item in the current food packages. The major total cost differences summarized in the table not only reflect the specific revisions to the items and the quantity prescribed, but also the distribution of participants across the different food packages. Because food package IV-B comprises the largest participant group (approximately 36 percent of food package recipients), relatively small changes lead to more drastic cost differences. For example, the CVV in the revised food package IV-B would increase by $3 per month compared to the current food package,⁸ leading to an estimated $246 million increase in

__________________

⁸ Participants do not received adjustments in the CVV until the inflated value inflates crosses a $1 increment. The CVV for food package IV is currently $8 per month. By 2018, it is expected to cross the next $1 increment and be adjusted to $9 per month.

a Phased-in estimates assume full implementation of the proposed revisions in state agencies serving one-third of all WIC participants as of April 1, 2018. All other state agencies are assumed to implement the proposed revisions as of FY2020. Cost differences were calculated by subtracting the estimated cost for the current food package item from the estimated costs corresponding to the food item in the revised food package. Negative values (−) indicate that the revised food packages result in cost savings compared to the current food packages. Positive values (+) indicate that the revised food packages result in cost increases compared to the current food packages.

^b This analysis assumes the incentives in the proposed revisions will result in a 5 percent shift of fully formula-fed mother–infant dyads to the partially (mostly) breastfed participant categories in the year after full implementation of the food package revisions across the entire WIC program under the phased-in assumption. The total amount of formula prescribed in the revised food packages I and II is expected to decrease because of this shift in participants. For food package I, the total cost difference for FY2018 through FY2022 is expected to result in $22.5 million in savings.

^c Because food package III recipients comprise a small proportion of food package recipients, all food package revisions resulted in total phased-in cost differences of less than $25 million over the course of FY2018 through FY2022.

^d The estimated cost differences for milk includes assumptions about the proportion of participants purchasing fluid milk (e.g., whole milk, 2%), milk alternatives (e.g., lactose-free milk, soy-based beverage), and substitutions (e.g., cheese, yogurt).

^e The estimated cost difference for whole wheat bread includes assumptions about the proportion of participants purchasing options available under the current food packages and proposed revisions (e.g., corn tortillas, instant oatmeal).

^f The proposed revisions remove cheese as a separate food package category prescribed to food package VII participants. Cheese remains as substitution option for milk across all children’s and women’s food packages.

estimated total phased-in costs, as compared to the current food package. In contrast, the CVV for fully breastfeeding women (food package VII)—who comprise approximately 3 percent of food package recipients—would increase by $24 per month in the revised food packages, and would only lead to an estimated $200 million increase in estimated phased-in costs.

Cost Differences of Food Package Items Over Time

The preceding sections broadly evaluate the cost implications of each food package item, summed across all projected years. As presented in Table U-7, the phased-in costs differences are not static over the course of FY2018 through FY2022 and reflect assumptions of the analysis. Cost differences in FY2018 and FY2019 are markedly lower than in subsequent years, for example, because they reflect the phased-in implementation of the revised food package. The values for these 2 years are one-third of the unadjusted value. The cost differences for FY2018 are further reduced, because the values only encompass a 6-month period.

the monthly costs of a food package by multiplying the projected number of recipients of the package, the average proportion of each food package item projected to be redeemed, and the estimated prices of each item in the food package and then summing these values within the food package. This process was repeated for each food package. Monthly costs were multiplied by 12 to arrive at annual estimates for each fiscal year in the analysis (FY20189 through FY2022). The sections that follow detail how the committee projected participation, redemption, and prices.

The presented estimates only encompass the cost of supplemental foods to the state agency. Administrative costs are not included. Although state and local WIC providers will likely incur additional administrative burden associated with the proposed revisions initially, such as updating their Management Information Systems and retraining staff, the committee was not in a position to quantify the costs of such changes.

Participation

The WIC Participation and Program Characteristics 2014: Food Package Report (USDA/FNS, 2016a) and average participation by participant category from administrative data posted on FNS’s website¹⁰ (USDA/FNS, 2016b) both describe WIC program participation by food package. While each provides a slightly different portrait of participation, neither directly provides the number of individuals who claim each specific food package type (outlined in Table U-1). The committee used both data sources to estimate the average number of participants who are actually issued a food instrument each month.

The 2014 Food Package Report is the only data source that provides insight into the distribution¹¹ of participants across the 27 specific types of food packages. While the data used for the Food Package Report are the most comprehensive available on the topic of specific food package assignments, there are limitations including, but are not limited to:

• The dataset captured all individuals certified to receive benefits during the month of April of the assessment year, regardless of whether the food benefits were claimed. The total number of participants in the Food Package Report’s analyses, therefore, is higher than average monthly participation counts derived from other sources

__________________

⁹ FY2018 in this analysis only encompasses 6 months. As such, monthly estimates were multiplied by 6 rather than 12.

¹⁰ Administrative data includes all state agencies, Indian Tribal Organizations, and territories.

¹¹ Throughout this appendix, the phrase “food package distributions” or simply “distributions” refers to the proportions of a WIC participant group assigned specific food packages. It does not refer to the physical distribution or claiming of the food benefit.

of administrative data, which generally reflect those that claimed benefits from the WIC agency.

• The food package type variable in the dataset is still considered to be in the testing phase. Because of this, the Food Package Report notes: “Findings should be treated as suggestive rather than definitive” (USDA/FNS, 2016a).
• The dataset only includes WIC participants in Puerto Rico; Washington, DC; and 48 states.¹² Louisiana and New Mexico did not provide information on specific food package assigned to participants. Furthermore, other territories and Indian Tribal Organizations were not included in the dataset. As such, the distributions of the food packages within the participant categories approximates, but may not necessarily represent, that which exists across all participating agencies.
• The dataset had variable levels of missing data. The Food Package Report notes that “the information provided by state agencies on food package types is not entirely complete or reliable” (USDA/FNS, 2016a).
• In the data received from states, participants were recorded as being part of one of five certification categories: pregnant women, breastfeeding women, postpartum women, infants, and children. The breastfeeding category encompasses fully breastfeeding women (intended to be food package VII); partially (mostly) breastfeeding women up to 1-year postpartum (intended to be food package V); partially (minimally) breastfeeding women up to 6-months postpartum (intended to be food package VI); and partially (minimally) breastfeeding women, 6- to 12-months postpartum (no food package). In the analyses performed in the Food Package Report, the breastfeeding certification category was separated into fully and partially breastfeeding women by dichotomizing the participants based on prescription of food package VII. Food package VII, however, can also be assigned to women pregnant with or partially (mostly) breastfeeding multiples, so this approach to subgrouping can only closely approximate the number of fully breastfeeding women in the dataset. Furthermore, partially (minimally) breastfeeding women more than 6-months postpartum are eligible for WIC services but do not receive a food package. Agencies differed in their approach to reporting the food package associated with these participants. Some reported a null food package (food package “N/A”) while most did not report a food package.

__________________

¹² This statement specifically pertains to Table C.1 in the Food Package Report, “Food Packages Assigned by Participant Category (2014).”

To extrapolate WIC participation into the future, the committee quantified the relationship between WIC participation levels and economic conditions, using regression analysis. The number of unemployed persons served as a measure of economic conditions. Regression analysis showed a strong correlation between the number of unemployed persons and WIC participation. This correlation was used to predict WIC participation based on the Federal Reserve’s forecast of unemployment rates and the Bureau of Labor Statistics’s forecast of the civilian labor force (BLS, 2013). Based on these data and the relationship between WIC participation and unemployment, the committee forecasts WIC participation to decline by 2.2 percent between FY2015 and FY2018. From FY2018 to FY2022, the committee forecasts WIC participation to increase by 1.5 percent. The FY2022 participation levels are forecasted to be 0.7 percent lower than the FY2015 levels. The relative changes in participation used in this RIA are summarized in Table U-8. The committee assumed the relative change would be experienced the same across all participant categories. Projected participation for the current food packages, by food package type, is presented in Table U-9. These values served as the participant multipliers for estimating costs of the current food packages.

Anticipating a Shift in Fully Formula-Fed Mother–Infant Dyads Under the Proposed Revisions

Participation projections are identical between the current and proposed revised food packages, with one exception. The committee anticipates the

NOTES: The base year of this analysis is FY2015. The committee estimated FY2015 participation from the 2014 Food Packages Report and 2015 average annual program participation from data posted on FNS’s website (USDA/FNS, 2016a,b).

^a The relative change describes the percent by which total participation increased or decreased, relative to the total participation in the preceding fiscal year. Total participation in FY2016, for example, was projected to be 1.1 percent lower than total participation in FY2015.

^b Year not included in the cost estimates presented in this RIA, but it was necessary to arrive at total participation projections for FY2018 through FY2022.

proposed revision to the categorization of partial (mostly) breastfeeding dyads in the first month will result in a shift of 5 percent of formula-fed infants (food packages I-FF-A, I-FF-B, and II-FF), 5 percent of postpartum and partially (minimally) breastfeeding women less than 6-months postpartum (food package VI recipients), and 5 percent of women classified as receiving the N/A food package to their respective partially (mostly) breastfeeding categories. The shift of women assigned the food package N/A to V-B slightly increases the total number of participants issued food benefits under the proposed revisions, as compared to the projections for the current food packages. The 5-percent shift was selected based on data presented in USDA/FNS (2011) that indicated the 2009 food package resulted in an approximately 8 to 11 percent shift of women out of the partially breastfeeding food package. The shifts were largely due to the fully formula-feeding food packages. The 5 percent value used in this analysis was therefore considered a conservative estimate for the number of participants that would shift back to the partially (mostly) breastfeeding category. The committee anticipates that the shift will take place after all the revisions have been implemented in all states and will be sustained but not recur in subsequent years. Accordingly, the shift has been incorporated in FY2021 participation estimates, which corresponds to the year after full implementation under the phased-in assumption. Projected participation under the proposed revisions by specific food package is presented in

Table U-10. These values served as the participant multipliers for estimating food package costs under the proposed revisions.

Redemption

Three sources of data were used to estimate and project redemption for each item in the current and revised food packages: (1) anonymized redemption data provided by FNS (hereafter referenced as the FNS redemption dataset), (2) redemption data provided by six individual WIC agencies, and (3) a 2014 Altarum report detailing redemption in Kentucky, Michigan, and Nevada as they transitioned to EBT (USDA/ERS, 2014).

Redemption data provided by FNS USDA-FNS provided the committee with 12 months (August 2013 through July 2014) of WIC redemption data from a convenience sample of 6 WIC state agencies, representing five of the seven regions of the country (the FNS redemption dataset). The identities of the states were not known to the committee. The states were diverse in terms of size and did not include Indian Tribal Organizations or territories.¹³

For each state, each month’s worth of data included the following information: a description of the main food package item category (e.g., “legume”); the subcategory of food (e.g., “canned beans,” “dry beans”); size and measure of the container purchased (e.g., 16 ounces); number of containers redeemed; average price for the container; and the total amount paid by WIC for that specific item (see the “Prices” section for use of the cost data). Because states agencies varied in their reporting approach—some providing specific descriptions of the items purchased (e.g., “soft corn tortilla”), while others only provided general descriptions (e.g., “whole grains, all types”)—redemption estimates collectively describe the main food package item category, rather than individual food items. To arrive at redemption estimates from the FNS redemption dataset, the following steps were taken for each state:

• The number of units purchased in a given month was standardized by multiplying the container size by the number of containers redeemed, summing the total amount purchased across the various container sizes, and, where applicable, dividing by a food package-appropriate unit size. For example, ounces of fish purchased in the month was determined across all container sizes (3.75 to 20 ounces), summed, and divided by 30 to arrive at number of 30-ounce units purchased.

__________________

¹³ As characterized by USDA-FNS.

a Food package III recipients are prescribed food packages appropriate for their age and physiological state. There are specific food packages within food package III for women, infants, and children that correspond to specific food packages I, II, IV, V, VI, and VII.

^b Women who are minimally breastfeeding and with infants more than 6 months of age that receive more formula than is allowed for a partially breastfed infant do not receive a food benefit. These women, however, are eligible to continue to receive other benefits through WIC, such as breastfeeding support, nutrition education, and health and social services referrals (USDA/FNS, 2013a). This group of participants is included in one of the committee’s assumption regarding the revised food package cost estimates, and is therefore included in the above table.

^c Total number of food package recipients does not include women assigned food package N/A.

a This analysis assumes the proposed revisions will incentivize partially (mostly) breastfeeding food packages for mother–infant dyads. Accordingly, the participation projections for the revised food package anticipates a 5 percent shift of fully formula-fed dyads to the partially (mostly) breastfed participant categories in the year after full implementation of the food package revisions. The participant shift is assumed to take place in FY2021. The shift is expected to be sustained, but to not recur in FY2022.

^b Food package III recipients are prescribed food packages appropriate to their age and physiological state. There are specific food packages for women, infants, and children within food package III that correspond to specific food packages I, II, IV, V, VI, and VII.

^c Women who are minimally breastfeeding and with infants more than 6 months of age that receive more formula than is allowed for a partially breastfed infant do not receive a food benefit. These women, however, are eligible to continue to receive other benefits through WIC, such as breastfeeding support, nutrition education, and health and social services referrals (USDA/FNS, 2013a). The anticipated shift in fully formula-fed dyads to the partially (mostly) breastfeeding categories includes women who are more than 6-months postpartum. Accordingly, 5 percent of women who did not qualify for a food package under the current food packages are anticipated to qualify for food package V-B under the proposed revisions.

^d Total number of food package recipients does not include women assigned food package N/A.

• The number of standardized units purchased in a given month was then averaged across the 12 months for each food package item category.
• The average number of standardized units purchased in a month was divided by the number of participants expected to receive the food item and accounted for differences in monthly maximum allowances by food package (e.g., the equation used to estimate juice redemption accounted for the different maximum amounts prescribed children and the various food packages for women). To keep the identity of the state agencies anonymous, FNS inputted average monthly participation by participant category for each state into a spreadsheet that contained the redemption equations created by the committee, and returned the overall unweighted average redemption across the state agencies per food package item.14 Through this process, FNS identified one of the six states as a clear outlier, and removed it from the averages (personal communication, K. Castellanos-Brown, USDA-FNS, June 22, 2016). As such, redemption estimates are calculated from five of the six state agency datasets provided by FNS.

Redemption estimates based on the FNS data have both strengths and limitations for the purposes of estimating and projecting costs associated with the food benefit. One of the strengths is that the data reflect actual purchases made by WIC participants with their food instrument. The data capture the range of container sizes that participants chose to purchase, and also include state-specific options for substitutions. Using redemption data, rather than making assumptions based on average quantity prescribed to participants, has the potential to better reflect the actual food costs. Limitations of the existing data, however, affect generalizability. The data only provide insight into the total number of units purchased in a given month or over the course of the year and were not currently reported in a way that could be used to see how redemption differed across specific food packages when an item was not unique to a single food package (e.g., breakfast cereal redemption may not be the same for children as it is for women). Furthermore, the FNS data could also not be used to arrive at redemption estimates for certain food package items. Redemption of milk and cheese, for example, could not be determined because cheese is offered not only as a separate food item provided in food package VII, but also as a substitution option for milk across children’s and women’s food packages. The data also predated the authorization of yogurt as a milk substitution option.

__________________

¹⁴ The committee was provided with unweighted averages to preserve the anonymity of the states included in the FNS dataset.

The redemption rates derived from the data provided by FNS and the data provided by individual state agencies were applied to the current food packages and served as the base for the redemption projections under the proposed revised food packages. The committee used a 2014 report from Altarum (USDA/ERS, 2014), which provided insight into three types of redemption practices (full-, partial-, and nonredemption) to develop data-based assumptions for the redemption behaviors in the revised food packages (detailed in Appendix R). Redemption estimates that were based only on assumptions about full-, partial-, and nonredemption were considered “implied redemption rates.” Implied redemption rates, however, do not account for behavioral changes, such as those that might result from the offering of new substitution options. Accordingly, the committee made slight adjustments to account for such changes. Table U-11 outlines the redemption rates applied to the current and revised food packages.

Applying redemption rates to the food packages The committee applied the calculated and projected redemption rates to the current and proposed monthly maximum allowances) for all food packages. Tables U-12 and U-13 summarize the estimated average amount redeemed for each food item for the current and revised food packages, respectively. The projected average amounts redeemed for each food package item serves as the quantity multipliers used in the estimation of total food costs.

a The U.S. Department of Agriculture’s Food and Nutrition Service (FNS) provided the committee with 12 months (August 2013 through July 2014) of price and redemption data from a convenience sample of six WIC state agencies, representing five of the seven regions of the country. The identity of the agencies was not known to the committee. The states were diverse in terms of size and did not include Indian Tribal Organizations or territories. FNS identified one of the states as an outlier and removed its data from the average redemption estimates. Accordingly, estimates presented in the table are based on data from five of the six states.

^b Cheese as its own food package item.

^c Based on redemption data provided by the Chickasaw Nation, Kentucky, Massachusetts, and Texas.

^d The committee used the redemption rates in the current food packages and the distribution of redemption shares (full, partial, nonredeemers) provided in a 2014 report from Altarum to calculate an implied redemption rate under the proposed revisions. The implied redemption rate accounts for the change in quantity prescribed to participants, primarily driven by the assumed average amount redeemed by partial redeemers. Although these values account for changes in amounts prescribed, they do not account for changes in composition or substitution options. The implied redemption rate does not account for behavioral changes.

^e The revised redemption rate is approximately the average of the implied redemption rate (Altarum) and the revised redemption rate for the CVV; it is also the actual average redemption rate from Altarum so represents a reasonable redemption rate.

^f Based on redemption data provided by the Chickasaw Nation, Kentucky, Massachusetts, and Wyoming.

^g Used the Altarum report average to account for bimodal distribution of redemption. Bimodal distribution refers to cases where the tails of the distribution (none and full redemption) had a relatively large share of the redemptions and the “partial redemptions” were a small share (USDA/ERS, 2014).

^h Redemption projection is or closely approximates the average redemption from the Altarum report, and therefore represents a reasonable projection.

ⁱ Based on redemption data provided by the California, Kentucky, and Massachusetts.

^j Approximately a 5 percent increase from the Altarum implied redemption rate.

^k Based on redemption data provided by the Chickasaw Nation, Kentucky, and Massachusetts.

SOURCES: Personal communication, K. Castellanos-Brown, USDA-FNS, redemption data provided to the committee, June 30, 2016; USDA/ERS, 2014; redemption data provided to the committee by individual states. Data are available in the public access file for this study (Email: paro@nas.edu).

a Calculated by multiplying the monthly maximum allowance by the redemption rate for the applicable food package and food package item. Value represents the estimated average quantity purchased by participants issued the specific food package.

^b Weighted average of the monthly maximum allowances of the three different forms of infant formula. Weights were 0.78 for reconstituted powder, 0.12 for reconstituted liquid concentrate, and 0.10 for ready-to-feed, and were derived from the 2014 Food Package Report (USDA/FNS, 2016a).

^c Food package III recipients were not uniquely identified in the redemption data. Accordingly, the committee assumed the redemption estimates in the current food packages and corresponding redemption projections applied to all participants categorically eligible to be prescribed the food item (i.e., a child prescribed food package III was assumed to be prescribed all of food package IV foods). With the exception of WIC formulas, the committee used both the monthly maximum allowances and the overall redemption rates for the standard food packages to arrive at corresponding food package III estimates. Owing to a lack of data on the reimbursement and costs to the program associated with exempt infant formulas and WIC-eligible nutritionals, estimates for food package III only encompass supplementary foods. The committee did, however, make one exception regarding WIC formula. The 2014 Food Package Report indicated that at least 40 percent of infant formula prescribed to partially (mostly) breastfed and fully formula fed infants is nonexempt (USDA/FNS, 2016a). Some nonexempt specialty formulas, such as thickened formulas, are reportedly covered by rebate contracts (USDA/ERS, 2004). Accordingly, the committee applied the proportion of nonexempt formula prescribed to the monthly maximum allowances for each of the infant food packages in food package III. The 94-percent redemption rate was then applied to those quantities and used in cost estimates. Because the quantity of infant formula did not change in the revised food package, this assumption is not expected to affect the estimated cost difference between the food packages.

^d Food package IV and VI recipients can select from either peanut butter or legumes each month. They do not receive both in a given month.

^e The 16 oz corresponds to the size of the dried legumes. States may also authorize four 16 oz cans (64 oz total) in lieu of the dried legumes. The 18 oz corresponds to size of the peanut butter. Although states may authorize smaller containers of canned legumes (15- to 16-oz) and peanut butter (16- to 18-oz), this analysis uses maximum container size for each.

^f Food package V and VII recipients are prescribed both legumes and peanut butter each month.

Prices

As described in the preceding “Redemption” section, FNS provided the committee with 12 months’ worth of anonymized redemption data from six state agencies (August 2013 through July 2014).¹⁵ The data included the

__________________

¹⁵ As was previously discussed, one of the states included in the six states dataset produced outlier redemption estimates when average participation was inputted into the redemption equations. The outlier state was omitted from the average redemption estimates. In looking at price per unit across the six states, however, there were no clear outliers. As such, all six states were included in price estimates.