5
EVALUATION OF COST
Amajor consideration in the redesign of the WIC food packages was the requirement to achieve cost neutrality in proposing recommended changes. According to Public Laws 101-147 and 105-336 (U.S. Congress, 1989, 1998), “States [i.e., WIC state agencies] must undertake cost containment measures, including contracts for the purchase of infant formula and, if possible, other WIC foods.”1 The importance of considering cost also was stated explicitly in the September, 2003 Advance Notice of Proposed Rulemaking (FNS, 2003a, p. 53907). For the purposes of this report, the term cost neutrality means that the average cost per participant of the complete set of revised WIC food packages (Food Packages I through VII) proposed in this report does not exceed that cost of the current WIC food packages using identical methods for estimating costs. This chapter explains the methods used to estimate the costs of the current and revised food packages and the results of these estimations on food package costs for the program as a whole. This chapter also presents comparisons of the market value of current and revised food packages for the three types of mother/infant pairs—fully breastfeeding, partially breastfeeding, and fully formula-feeding mother/infant pairs. These comparisons show changes in the potential monetary value of the packages for breastfeeding and non-breastfeeding postpartum women. Lastly, because current trends in the
prices of milk and infant formula indicate the potential for large increases in the future costs of the WIC food packages with or without revisions, the chapter addresses the sensitivity of estimates to changes in the prices of these foods.
OVERVIEW
In the process of redesigning the food packages, the committee estimated the cost of a number of possible sets of food packages. At each iteration, possible adjustments were considered in the types and amounts of foods needed to achieve cost neutrality while meeting the criteria shown in Chapter 1—Introduction and Background (Box 1-1). In following this approach, the committee initially worked with the basic food packages for women and children—that is, the food packages without substitutions. Depending on the package, these basic food packages include fluid milk, cheese, peanut butter, dried beans, whole wheat bread, eggs, tuna, and fresh fruits and vegetables. Because the committee strived to allow for flexibility in the revised food packages, the costs (and nutrient content) of food packages that incorporated substitutions at specified rates were also estimated (see Appendix E). The final cost estimates for the set of revised food packages include the cost of making selected substitutions at specified rates (see Appendix E)2 to the basic set of food packages. The specified substitution rates are based on assumptions; differences in assumptions would lead to a range in estimated average participant cost per month. Since most of the substitutions are higher-cost food items, the estimated cost of the set of revised food packages with substitutions is higher than the cost of less flexible food packages.
Within regulatory parameters, WIC state agencies currently can control costs by specifying a food item in lower-cost forms, varieties, brands, or container sizes. In estimating cost, the committee did not consider additional state or local agency discretion. Instead, costs were calculated using various forms, varieties, brands, and container sizes of food items that are representative of current practice or common use (i.e., based on the average share of household market purchases in national survey data) (ACNielsen Homescan; ACNielsen, 2001).
In evaluating the cost neutrality of proposed changes, the committee estimated the food costs to the WIC program based on the estimated costs
2 |
Bases of Assumptions Used in Nutrient and Cost Analyses of Food Packages can be found in Tables E-1 (for infants) and E-2 (for children and women) in Appendix E—Cost Calculations. Calculated Costs of Representative Amounts of Foods in Revised Packages can be found in Tables E-3A (for infants) and E-3B (for children and women). |
of each food package and the number of participants in the relevant participant category for the year 2002.3 In all cases, it was assumed that the revised food packages or the allowed substitutions had no effect on rates of participation in the WIC program. The cost of each of the current and revised food packages was estimated using the maximum monthly allowance for each food and a nationally representative price for the specified food items. For revised food packages, assumptions were made about the substitution rates for selected higher cost substitutions in the package. This process yields an estimate of the cost of the maximum package per month. Although changes in the prescription rates4 or redemption rates5 have the potential to change program costs, data are unavailable on which to adjust for the current or future prescription rates or redemption rates.
METHODS
Data
General Data Considerations
The base year for analysis was 2002, a recent year with a reasonably complete set of program and participant data available. The quantities for food items were based on the maximum allowances specified for the current and revised packages (for current Food Packages I–VII, see Table 1-1 in Chapter 1—Introduction and Background; for the revised food packages for infants see Table 4-1 and for children and women see Table 4-2 in Chapter 4—Revised Food Packages).
Prices
Data for the prices of each of the foods were identified from sources considered nationally representative of the food items likely to be selected and available to the participants. No single best source was available for all of the food prices. The committee used the sources indicated for the following types of foods:
-
Infant Formula—Retail prices for infant formula were obtained from a report released by the Economic Research Service (ERS) (Oliveira et al., 2001) that calculated the average market price of infant formula using 2000 retail-scanner data (scanner-based retail sales tracking data from Information Resources, Inc. [IRI, Chicago, IL], InfoScan database). The data were designed to be representative of the United States using 64 market areas. The ERS report included price data on all types of infant formula (i.e., standard6 and specialized).7 The committee used only data on prices for standard milk-based formulas for this analysis.
-
Fruits and Vegetables—Estimated prices for fruits and vegetables, including prices by form of the produce (i.e., fresh, canned, frozen, dried), were obtained from Reed et al. (2004) and a recently released data set (ERS, 2004b) on fruit and vegetable purchases and prices. These prices are based on ACNielsen Homescan 1999 price data (ACNielsen, 2001).
-
Eggs—The source of the egg price was monthly average price data for 2002, Grade A, large eggs from the Consumer Price Index—Average Price Data of the Bureau of Labor Statistics (BLS, 2004a).8
-
Other Groceries—Prices for groceries (including milk and milk products) were calculated based on 2001 ACNielsen Homescan purchase-price data (ACNielsen, 2001). Specifications for each food item reflected the allowed product characteristics for the current food packages obtained from the FNS website (FNS, 2004d, 2004e) or for the revised food packages as described in Table 4-3 (Proposed Specifications for Foods, Chapter 4—Revised Food Packages). The pricing data used also reflect container sizes allowed, if known for the current packages or if applicable for the
-
revised packages. The average unit value (expenditure divided by quantity) used purchase data from all sample respondents in the ACNielsen Homescan panel (i.e., regardless of income). This method yields a market purchase-weighted price.
The price data come from different years (1999–2002), depending on the source of data. However, adjustment of the price data to the base year 2002 (BLS, 2004b, Consumer Price Index—Food at Home) showed that, during this period, the adjustment for overall price changes made a small and similar difference in overall costs for both the current and revised set of packages (that is, less than 1 percent difference for either set of packages). Hence, unadjusted price data were used in the analyses presented here.
Infant Formula Rebate Assumption
Under cost control requirements, WIC state agencies must negotiate rebate contracts with infant formula companies. All of the cost estimates make use of the following assumption on the rebate.
Data provided by FNS (public communication during open session, February 2004, J. Hirschman, Office of Analysis, Nutrition and Evaluation, Food and Nutrition Service, USDA; FNS, 2004c, 2002 data) indicate that the average monthly pre-rebate cost for Food Package I (for infants ages birth through 3 months) was $94.03, and the average monthly post-rebate food package cost was $30.17. Because the current Food Package I comprises infant formula only, the committee used the ratio of the two costs to estimate the post-rebate cost as 32.1 percent of the pre-rebate cost of formula. Therefore, the committee adjusted the actual cost of the infant formula (obtained as described above) by a factor of 0.321 to obtain the post-rebate cost for the formula included in both the current and revised food packages. By holding the rebate level constant, prices are held constant for the purposes of comparing costs between current and revised food packages.
Numbers of Participants
The numbers of individuals in each participant category were from WIC Participant and Program Characteristics: PC2002 (Bartlett et al., 2003, tables; Kresge, 2003, executive summary). Further delineation of participant groups by breastfeeding status was based on estimates of breastfeeding practices across the United States and among WIC participants by the Center for Disease Control and Prevention (CDC, 2004b, 2004c, 2004e).
Estimating Food Package Costs
Estimated package costs for the current and revised food packages are based, respectively, on the current or revised amounts of each food item and an estimated cost per unit of the food item.
The contents of the current and revised packages can be described in terms of general food categories (e.g., breakfast cereals) or representative food items (e.g., instant oatmeal). In many cases, the price for a general food category included in the package is the weighted average of several food items, estimated using a series of assumptions. The specific assumptions used in the cost analysis are presented in Tables E-1 and E-2 in Appendix E—Cost Calculations. For example, a weighted average for the cost of breakfast cereals using market share data (ACNielsen Homescan; ACNielsen, 2001) was used to determine the proportion of total cereal products purchased as cooked cereal (10 percent) and as ready-to-eat cereal (90 percent). The weighting done to estimate package costs is the same weighting that was done for the nutritional analyses except for some selected food items;9 details of the weighting are presented in Tables E-1 and E-2. When the package included a cash-value voucher for fresh fruits and vegetables, the value of the voucher was included in the cost of the package. That is, the total package cost for each participant category was calculated as the sum of the costs of component food items plus the cash value of the voucher for fresh produce, as applicable. See Table 5-1 for a comparison of the estimated costs of the current and revised food packages. See Tables E-3A and E-3B in Appendix E—Cost Calculations for the cost of representative amounts of component food items used in the revised food packages.
Estimating Program Costs for Food
To estimate program costs for the sets of current and revised food packages, the estimated number of participants receiving each package in 2002 was multiplied by the estimated cost of the respective package. The committee assumed that there would be no change in WIC participation rates and no shifts among applicable participant categories. Although some
TABLE 5-1 Comparison of Estimated Costs of Current and Revised Food Packages (2002)a
shifting in WIC participation rates and among participant categories might be expected in response to changes in WIC food packages and policies relating to them, the basic comparison of costs assumed no change in participation rates. The potential effect of participation rate changes on costs was explored through sensitivity analysis (see section on Cost Neutrality).
The current and revised sets of food packages do not include estimates of the costs of the package for participants with special dietary needs, that is, for infants receiving special formulas and children and women receiving Food Package III. The committee’s assumption is that there would be no differences in the cost of special formulas and medical foods in the sets of current and revised food packages. As an example, the committee used the assumption that the prescription rate for soy infant formula stays constant for the current and revised food packages; therefore, this parameter was not included in the cost analyses.
Current Package Cost (post-rebate, if applicable) |
Revised Package No. |
Revised Package Cost (post-rebate, if applicable) |
$29.75 |
I-FF-A |
$29.75 |
$29.75 |
— |
— |
$29.75 |
I-BF/FF-A |
$11.96 |
0 |
— |
0 |
$37.43 |
I-FF-B |
$32.63 |
$37.43 |
I-BF/FF-B |
$16.32 |
$ 7.68 |
I-BF-B |
0 |
$37.43 |
II-FF |
$42.30 |
$37.43 |
II-BF/FF |
$30.78 |
$ 7.68 |
II-BF |
$57.10 |
$39.29 |
IV-A |
$38.98 |
$39.29 |
IV-B |
$38.49 |
$41.23 |
V |
$48.45 |
$41.23 |
V |
$48.45 |
$34.39 |
VI |
$37.41 |
$50.61 |
VII |
$57.05 |
DATA SOURCES: Price data are from Economic Research Service, USDA (ERS, 2004b, 1999 price data; Oliveira et al., 2001, 2000 infant formula price data) and ACNielsen Homescan (ACNielsen, 2001, price data for 2001obtained through ERS, USDA); and the Bureau of Labor Statistics, U.S. Department of Labor (BLS, 2004a, 2002 price data). |
RESULTS AND DISCUSSION
The estimates of the total program cost for food in 2002 are reported in Tables 5-2 and 5-3 (Estimated Program Costs for Food per Month Using Current Packages and Revised Packages). In these tables, the average post-rebate costs are the “program participant” weighted average monthly food package costs. The current program cost for food (Table 5-2) is estimated to be an average 2002 cost per participant of $34.76 per month. In comparison, FNS estimated the average monthly post-rebate food package cost for FY02 to be $34.84.10 The FNS estimate is based on participation,
TABLE 5-2 Estimated Program Costs for Food per Participant per Month Using Current Packages (2002)a
Group |
Age/Participant Category |
Description |
Package |
|
Infants |
0–3.9 mo |
|
Fully formula-fed |
I |
Partially breast-fedc |
I |
|||
Fully breast-fedc |
— |
|||
Subtotalsd |
|
|
||
4–5.9 mo |
|
Fully formula-fed |
II |
|
Partially breast-fede |
II |
|||
Fully breast-fede |
II |
|||
Subtotalsd |
|
|
||
6–11.9 mo |
|
Fully formula-fed |
II |
|
Partially breast-fedf |
II |
|||
Fully breast-fedf |
II |
|||
|
Subtotalsd |
|
|
|
Totals for infantd |
|
|
||
Children |
1–4.9 yg |
|
IV |
|
Totals for childrend |
|
|
||
Women |
Pregnantd |
|
V |
|
Partially breastfeedingf |
|
V |
||
Non-breastfeeding postpartumd |
|
VI |
||
Fully breastfeedingh |
|
VII |
||
Totals for womend |
|
|
||
Total for program |
||||
Average cost per participant (per month) |
||||
aAll costs use market purchase-weighted prices estimated using 1999–2002 price data (see data sources). Data on rates of participation are for 2002 (Bartlett et al., 2003). bData on the total pre- and post-rebate cost of the infant packages was provided to the committee by FNS (public communication during open session, February 2004, J. Hirschman, Office of Analysis, Nutrition and Evaluation, Food and Nutrition Service, USDA). The committee used these data to estimate that the average post-rebate cost of infant formula was 32.1%of the pre-rebate cost using 2002 data. The data for FY2002 can be obtained on the FNS website (FNS, 2004c); data for other years are also available and verify the recent range of post-rebate costs as 29.2 to 32.5% (1999 through 2003) (FNS, 2000, 2001, 2003b, 2005). cPercentage of infants fully breast-fed at 3 mo of age was reported (CDC, 2004c). Percentage of partially breast-fed infants was calculated from these data and data on the percentage of infants who had ever been breast-fed at 3 mo of age (CDC, 2004c). dNumber of participants was calculated using data Exhibit 3.1 from USDA’s WIC Participant and Program Characteristics, 2002 (Bartlett et al., 2003), recognizing that some discrepancies exist in these data. An infant is defined as a participant who, at certification, is under 1y of age and who would be classified as a child at the age of 366 d. However, in 2002, about 2.84% of WIC participants categorized as 1-y-old children are, in fact, 11-mo-old infants who have been recertified as 1-y-old children; additionally, about 0.38% of WIC participants who are classified as infants are participants who are older than 366 d. |
Percentage within Age/ Participant Category |
Number of Participants |
Cost (pre-rebate, if applicable) |
Post-Rebate Costb |
36 |
668,309 |
$ 92.69 |
$29.75 |
28 |
519,796 |
$ 92.69 |
$29.75 |
36 |
668,309 |
0 |
|
100 |
1,856,414 |
|
|
69 |
38,428 |
$100.37 |
$37.43 |
20 |
11,138 |
$100.37 |
$37.43 |
11 |
6,126 |
$ 7.68 |
|
100 |
55,692 |
|
|
79 |
118,955 |
$100.37 |
$37.43 |
16 |
24,092 |
$100.37 |
$37.43 |
5 |
7,529 |
|
$ 7.68 |
100 |
150,576 |
|
|
|
2,062,682 |
|
|
100 |
4,020,032 |
|
$39.29 |
100 |
4,020,032 |
|
|
45 |
878,619 |
|
$41.23 |
11 |
205,559 |
|
$41.23 |
31 |
597,451 |
|
$34.39 |
13 |
252,572 |
|
$50.61 |
100 |
1,934,201 |
|
|
|
8,016,915 |
|
|
|
|
|
$34.76 |
ePercentage of infants fully or partially breast-fed at 4–5.9 mo of age was extrapolated from data for infants at 3 and 6 mo of age (CDC, 2004c; Abbott Labs, 2002, 2003 [2001 data]). fPercentages of infants fully or partially breast-fed at 6–11.9 mo of age were calculated as the average of data reported for infants at 6 mo (CDC, 2004c) and 12 mo of age (CDC, 2004c; Briefel et al., 2004a). gIncludes 0.8% of children, age 1–4.9 y, who were reported as “age not reported.” hPercentage distribution of women as fully breastfeeding (55% of the total) or partially breastfeeding (45%of the total) was calculated according to the distribution of infants identified as fully or partially breast-fed (see notes e and f). NOTES: Additional detail is available in Table E-4 in Appendix E—Cost Calculations. DATA SOURCES: Price data are from Economic Research Service, USDA (ERS, 2004b, 1999 price data; Oliveira et al., 2001, 2000 infant formula price data); ACNielsen Homescan (ACNielsen, 2001, price data for 2001 obtained through ERS, USDA); and Bureau of Labor Statistics, U.S. Department of Labor (BLS, 2004a, 2002 price data). Data on rates of participation are from resources published by USDA (Bartlett et al., 2003, 2002 data; Kresge, 2003, 2002 data). Data on percentages of infants breast-fed were obtained from the 2003 National Immunization Survey (CDC, 2004b, 2004c) and published resources (Abbott Labs, 2002, 2003; Briefel et al., 2004a). |
program total food expenditures, and total rebates from the WIC program administrative databases, adjusted for price changes. The committee’s estimates are based on participation, food package quantities, and food prices. The committee’s estimates (1) for program participation are described in Tables 5-2 and 5-3, (2) for quantities of food are described in Table 1-1 (Chapter 1—Introduction and Background) for the current food packages and in Tables 4-1 and 4-2 (Chapter 4—Revised Food Packages) for the revised food packages, and (3) for food prices are described in Tables E-3A and E-3B (Appendix E—Cost Calculations). The committee’s methods are expected to affect the estimated 2002 cost of the current set of food packages as follows.
-
The assumption of a full prescription rate11 and the selection of some high-priced food items in the packages for children and women (e.g., not restricting selection to store brands), could lead to cost estimates higher than those obtained from administrative data on program total food expenditures.
-
The committee’s cost estimates do not include any separate costs for participants with special dietary needs (Food Package III). Because the committee assumed these costs would not change, the comparison of the cost of the current and revised food packages are valid. However, the committee’s assumption should lead to an estimated cost that is lower than the cost obtained from administrative cost data that include the costs of Food Package III.
-
Using nationally representative prices rather than the prices available to WIC participants in local stores or in certain areas served by the WIC program could lead to cost estimates lower than those faced by WIC state agencies (Davis and Leibtag, 2005). For example, some high cost areas such as Alaska, Hawaii, territories, and reservations that are served by the WIC program were not represented in the price data used by the committee.12
Nonetheless, the use of the same method for estimating the costs of the current and of the revised packages minimizes the effects of these assump-
tions used in the cost analyses. This approach produces valid estimates to use in determining whether or not the revised packages are cost-neutral.
Cost Neutrality
For the purposes of evaluating whether the set of revised packages is consistent with controlling food costs, the committee compared the average cost per participant for the current set of packages (as estimated by the committee) to the average cost per participant for the set of revised packages (estimated by the committee in the same manner). The comparison of costs between the two sets of packages is made on the basis of “average post-rebate food package cost.” A cost-neutral set of proposed changes would be such that the post-rebate average cost per participant of the set of revised packages is close to that of the current average post-rebate average cost per participant. Thus, the basis of comparison is the committee’s estimate of an average 2002 cost per participant for the current food packages of $34.76 per month.
The average 2002 cost of the revised food package is estimated to be $34.57 per participant per month—approximately equal to the current set of packages ($0.19 less than that of the current set of food packages, a difference of less than 1 percent of the average participant cost of the set of current packages). See Tables 5-2 and 5-3. Therefore the set of revised food packages meets the cost neutrality constraint. Varying some of the rates of substitution of product forms in ways that test the sensitivity of the estimates to some of the assumptions leads to estimates of the average cost per participant that lie in the range of $34.03 to 34.95 per participant per month. Even though the estimated cost of a revised food package is higher than the cost of the corresponding current food package for some participant groups, costs for other revised packages are lower than those of the corresponding current package. On average, the cost per participant is no higher. Since the methods and sources used for determining costs were the same for the sets of current and revised food packages, the similarity in costs indicates that the proposed changes in components of the packages would have little effect on overall food costs to the WIC program assuming no changes in participation by category or in redemption rates.
Costs of Substitutions
Table 5-4 shows the effects of selected substitutions on costs. For example, one quart of yogurt costs $1.58 more, on average, than does one quart of fat-reduced milk. Buying two quarts (one-half gallon) of calcium-and vitamin D-rich soy beverage (“soy milk”) in place of two quarts of lowfat milk would cost an extra $1.88 (2 qt at $0.94 per qt). Use of canned
TABLE 5-3 Estimated Program Costs for Food per Participant per Month Using Revised Packages (2002)a
Group |
Age/Participant Category |
Description |
Package |
|
Infants |
0–3.9 mo |
Fully formula-fed |
I |
|
—(0–0.9 mo) |
||||
I (1–3.9 mo) |
||||
Fully breast-fedc |
— |
|||
|
Subtotalsf |
|
|
|
4–5.9 mo |
Fully formula-fed |
II |
||
Partially breast-fedg |
II |
|||
Fully breast-fedg |
II |
|||
|
Subtotalsf |
|
|
|
6–11.9 mo |
Fully formula-fed |
II |
||
Partially breast-fedh |
II |
|||
Fully breast-fedh |
II |
|||
|
Subtotalsf |
|
|
|
Totals for infantsf |
|
|
||
Children |
1–1.9 yi |
|
IV-A |
|
2–4.9 yi |
|
IV-B |
||
Totals for childrenf |
|
|
||
Women |
Pregnantf |
|
V |
|
Partially breastfeedingj |
|
V |
||
Non-breastfeeding postpartumf |
|
VI |
||
Fully breastfeedingj |
|
VII |
||
Totals for womenf |
|
|
||
Totals for program |
||||
Average cost per participant (per month) |
||||
aAll costs use market purchase-weighted prices estimated using 1999–2002 price data (see data sources). Data on rates of participation are for 2002 (Bartlett et al., 2003). bData on the total pre- and post-rebate cost of the infant packages was provided to the committee by FNS (public communication during open session, February 2004, J. Hirschman, Office of Analysis, Nutrition and Evaluation, Food and Nutrition Service, USDA). The committee used these data to estimate that the average post-rebate cost of infant formula was 32.1%of the pre-rebate cost using 2002 data. The data for FY2002 can be obtained on the FNS website (FNS, 2004c); data for other years are also available and verify the recent range of post-rebate costs as 29.2 to 32.5% (1999 through 2003) (FNS, 2000, 2001, 2003b, 2005). cPercentage of infants fully breast-fed at 3 mo of age was reported (CDC, 2004c). Percentage of partially breast-fed infants was calculated from these data and data on the percentage of infants who had ever been breast-fed at 3 mo of age (CDC, 2004c). dFor the category of partially breast-fed infants 0–3.9 mo, the committee estimated that the number of infants aged 0–0.9 mo was 25% of the category total and the number of infants aged 1–3.9 mo was 75% of the total. In the absence of data on the proportion of infants to anticipate in each of the first 4 mo of life, the committee assumed the distribution would be approximately equal in each month, using the census data for children under the age of 5 y as a model (20.0% ± 0.3%, mean ± SD) (U.S. Census Bureau, 2004). |
Percentage within Age/ Participant Category |
Number of Participants |
Cost (pre-rebate, if applicable) |
Cost Post-Rebateb |
36 |
668,309 |
$92.69 |
$29.75 |
7 |
129,949 |
$ 4.65e |
$ 1.49 |
21 |
389,847 |
$37.25 |
$11.96 |
36 |
668,309 |
0 |
|
100 |
1,856,414 |
|
|
69 |
38,428 |
$101.66 |
$32.63 |
20 |
11,138 |
$50.83 |
$16.32 |
11 |
6,126 |
0 |
|
100 |
55,692 |
|
|
79 |
118,955 |
$91.02 |
$42.30 |
16 |
24,092 |
$55.14 |
$30.78 |
5 |
7,529 |
$57.10 |
|
100 |
150,576 |
|
|
|
2,062,682 |
|
|
36 |
1,447,212 |
|
$38.98 |
64 |
2,572,820 |
|
$38.49 |
100 |
4,020,032 |
|
|
45 |
878,619 |
|
$48.45 |
11 |
205,559 |
|
$48.45 |
31 |
597,451 |
|
$37.41 |
13 |
252,572 |
|
$57.05 |
100 |
1,934,201 |
|
|
|
8,016,915 |
|
|
|
|
|
$34.57 |
eOne alternative is to provide one small can (up to 15 oz) of powdered formula to breastfed infants during the first month if requested by the mother. The committee used the assumption that the number of breastfeeding mothers requesting formula in the first months would approximate 50% of the current number of partially breastfeeding mother/infants pairs. The additional monthly cost per participant who choose this option would be $9.30 in pre-rebate costs and $2.98 in post-rebate costs. Using the estimate of 50% of the current partially breastfeeding participants (0.5 × 129,949 = 64,747) for the first month postpartum, the additional monthly program cost would be an additional 2.4¢ in the average cost per participant. fNumber of participants was calculated using data Exhibit 3.1 from USDA’s WIC Participant and Program Characteristics, 2002 (Bartlett et al., 2003), recognizing that some discrepancies exist in these data. An infant is defined as a participant who, at certification, is under 1 year of age and who would be classified as a child at the age of 366 days. However, in 2002, about 2.84% of WIC participants categorized as 1-y-old children are, in fact, 11-mo-old infants who have been recertified as 1-y-old children; additionally, about 0.38% of WIC participants who are classified as infants are participants who are older than 366 d. |
gPercentage of infants fully or partially breast-fed at 4–5.9 mo of age was extrapolated from data for infants at 3 and 6 mo of age (CDC, 2004b; Abbott Labs, 2002, 2003 [2001 data]). hPercentages of infants fully or partially breast-fed at 6–11.9 mo of age were calculated as the average of data reported for infants at 6 mo (CDC, 2004c) and 12 mo of age (CDC, 2004c; Briefel et al., 2004a). iThe committee calculated the number of participants in each category using data from the USDA sponsored WIC Participant and Program Characteristics 2002 (Bartlett et al., 2003); data from Exhibit 3.1 (Bartlett et al., 2003) were used to estimate the number of participants ages 1–1.9 y and 2–4.9 y. jPercentage distribution of women as fully breastfeeding (55% of the total) or partially breastfeeding (45% of the total) was calculated according to the distribution of infants identified as fully or partially breast-fed (see notes g and h). NOTES: Additional detail is available in Table E-4 in Appendix E—Cost Calculations. DATA SOURCES: Price data are from Economic Research Service, USDA (ERS, 2004b, 1999 price data; Oliveira et al., 2001, 2000 infant formula price data); ACNielsen Homescan (ACNielsen, 2001, price data for 2001 obtained through ERS, USDA); and Bureau of Labor Statistics, U.S. Department of Labor (BLS, 2004a, 2002 price data). Data on rates of participation are from resources published by USDA (Bartlett et al., 2003, 2002 data; Kresge, 2003, 2002 data). Data on percentages of infants breast-fed were obtained from the 2003 National Immunization Survey (CDC, 2004b, 2004c) and published resources (Abbott Labs, 2002, 2003; Briefel et al., 2004a). |
fruits and vegetables instead of a cash-value voucher for fresh fruits and vegetable would lead to an estimated $1.73 increase in costs per month for an adult and $1.21 increase in costs per month for a child. Buying canned beans (4 16-oz cans) rather than one pound of dried beans would increase the monthly cost by $1.79.
In estimating the average cost per participant of the set of revised food packages (shown in Table 5-3), the committee allowed for the types and rates of substitutions specified in Tables E-1 and E-2 in Appendix E—Cost Calculations. For example, costs for the food package for pregnant and partially breastfeeding women assume that 50 percent of participants will obtain the canned form of legumes rather than the dried form; 50 percent of participants will select processed forms of fruits and vegetables rather than the fresh form; 10 percent of women will obtain calcium- and vitamin D-rich soy beverage (“soy milk”) as an alternate for milk; and 50 percent of women will choose the maximum amount of allowed substitutions for milk (see Tables E-1 and E-2 for detailed information). Since virtually all of the proposed substitutions increase the cost of the package, the costs shown in Table 5-4, which include all of the allowed substitutions in the revised food packages, are higher than the cost would be if less flexibility were allowed. Nonetheless, even allowing this flexibility, the estimated average monthly
cost of the set of revised packages is approximately equal to that of the set of current food packages; it is lower by less than 1 percent of the average monthly cost of the current package.
Sensitivity analysis that considered the various choices indicated the estimated average 2002 cost would range from $34.03 to $34.95. The lower end of the range uses the cost of only the fresh option for fruits and vegetables and the higher end of the range assumes that 75 percent rather than 50 percent of participants will choose canned rather than dried legumes. The committee encourages the fresh produce option wherever feasible for several reasons: greater participant choice and therefore higher acceptability by participants of widely varied cultural backgrounds, lower estimated costs, and lower sodium content. However, the committee recognizes that the cost of fresh fruits and vegetables is likely to increase over time and, as a result, the amount of fresh produce that could be obtained with the cash-value voucher would decrease. Because this would lead to a reduction in the nutrient content of the package, the value of the cash-value voucher (both monetary and nutritive value) would need to be reviewed every 1 to 3 years. One index to use as the basis of the adjustment is the Bureau of Labor Statistics Consumer Price Index series for fresh fruits and vegetables (BLS, 2005).
COMPARING COST INCENTIVES FOR BREASTFEEDING
Using data from Tables 5-2 and 5-3, Table 5-5 was constructed to compare the market (pre-rebate) value of the maximum allowances for current and revised food packages for the three types of mother/infant pairs—fully breastfeeding, partially breastfeeding, and fully formula-feeding. The data in Table 5-5 consider the full length of time that the mother and infant are eligible for the food packages. Table 5-5 shows that the market value for the revised packages for the fully breastfeeding mother/ infant pair is substantially higher than that of the current package. The three revised food packages for mothers/infants pairs are more similar in value than are the current food packages. The market value of the set of revised food packages for the fully formula-feeding mother/infant pair remains higher than that for the fully breastfeeding pair, but the difference is greatly diminished. The market value of the current packages for the fully formula-feeding pair is more than two times higher than that for the fully breastfeeding pair; but, for the revised packages, the market value for the fully formula-feeding pair is less than a third higher. The revised food packages for the three possible feeding methods have similar market values.
The changes in the relative value of the mother/infant pairs, when combined with appropriate nutrition education efforts, are designed to encourage more breastfeeding. A switch to more breastfeeding (both full
TABLE 5-4 Estimated Costs of Basic Foods, Selected Substitutions, and Net Cost Changes Resulting from Selected Substitutions in WIC Food Packages (2002)a
and partial) would have an impact on costs. The direction and magnitude of the change was investigated by simulating possible shifts in participation rates. As expected, shifting mother/infant pairs from the fully formula-feeding option to breastfeeding options has the effect of moving mother/ infant pairs from the most expensive set of packages to less expensive ones (Table 5-5). In simulations with fewer fully formula-feeding mother/infant
Selected Substitution |
|
Food |
Net Change in Cost of Food Item ($) |
Yogurt (1 qt) for whole milk (1 qt) |
+1.55 |
Cheese (1 lb) for milk (3 qt) |
+1.11 |
Yogurt (1 qt) for milk (1 qt) |
+1.58 |
Soy beverage(1 qt) for milk (1 qt) |
+0.94 |
Tofu (1 lb) for milk (1 qt) |
+1.06 |
Cheese (1 lb) for milk (3 qt) |
+1.23 |
Peanut butter for dried beans |
+1.03 |
Canned beans (4 16-oz cans) for dried beans (1 lb) |
+1.79 |
Brown rice for whole wheat bread (1:1) |
–0.03 |
Salmon for tuna (~30 oz:~30 oz) |
+0.53 |
Canned fruits and vegetables only |
+1.21 |
Canned fruits and vegetables only |
+1.73 |
for fresh fruit is $0.69 per lb; estimated average price for fresh vegetables is $0.94 per lb. Details on costs of food items in the revised packages are presented in Tables E-3A and E-3B in Appendix E—Cost Calculations. DATA SOURCES: Price data are from Economic Research Service, USDA (ERS, 2004b, 1999 price data); ACNielsen Homescan (ACNielsen, 2001, price data for 2001 obtained through ERS, USDA); and the Bureau of Labor Statistics, U.S. Department of Labor (BLS, 2004a, 2002 price data). |
pairs and more fully and partially breastfeeding mother/infant pairs (using a constant shift of 30 percent for 1 to 11 months of age from partial to full breastfeeding, and a smaller range of shifts from full formula feeding to full breastfeeding), the average 2002 cost fell from $34.57 to $33.93, a decrease of nearly 2 percent (for further detail, see Appendix E—Cost Calculations, section on Possible Shifts in Participation Rates).
TABLE 5-5 Comparison of the Market (Pre-Rebate) Value of Maximum Allowances for Current and Revised Food Packages for Mother/Infant Pairs (2002)a
|
Fully Breastfeeding |
Partially Breastfeeding |
|||
Participant Category |
Cost for Month |
Cost per First Year Months |
Cost per Postpartum |
Month |
Cost for First Year Months |
Current Food Packages |
|||||
Mother |
$50.61 |
12 |
$ 607 |
$ 41.23 |
12 |
Infant, 0–3.9 mo |
— |
4 |
— |
$92.69 |
4 |
Infant, 4–11.9 mo |
$ 7.68 |
8 |
$61 |
$100.37 |
8 |
Total Cost |
|
|
$ 668 |
|
|
Revised Food Packages |
|||||
Mother |
$57.05 |
12 |
$ 685 |
$ 57.05 |
1b |
|
— |
— |
— |
$48.45 |
11 |
Infant, 0–3.9 mo |
0 |
4 |
— |
$37.25 |
3b |
Infant, 4–5.9 mo |
0 |
2 |
— |
$50.83 |
2 |
Infant, 6–11.9 mo |
$57.10 |
6 |
$ 343 |
$ 55.14 |
6 |
Total Cost |
|
|
$1,028 |
|
|
aAll costs use market purchase-weighted prices estimated using 1999–2002 price data (see data sources). Data on rates of participation are for 2002 (Bartlett et al., 2003). bThere are no packages for the partially breastfeeding mother/infant pair in the first month postpartum. In an attempt to promote successful breastfeeding, the mother and infant would be considered as fully breastfeeding for the first month. These estimates do not include the cost of one can of powdered formula available to mothers who are listed as partially breastfeeding in the first month. |
PROJECTING THE EFFECTS OF CHANGES IN INFANT FORMULA AND MILK PRICES
Of course, any evaluation of costs based on prices of the base 2002 period (or 1999–2002 prices) is sensitive to changes in prices. Greater variability in prices for fruits and vegetables and increases in prices over time, for example, may mean that the amount of food obtained from a fixed valued instrument may change. From the perspective of the program costs, however, the fixed valued instrument for fresh fruits and vegetables isolates the program from the price variation. Current trends, such as relatively large price increases for milk products in 2004, changes in the types of infant formulas marketed, and unfavorable changes in the rebates for infant formulas have made the WIC program vulnerable to increased food costs. Because of changes in the relative amounts of formula and of
|
Fully Formula Feeding |
||
Cost for Postpartum |
Month |
Cost per First Year Months |
Postpartum |
$ 495 |
$ 34.39 |
6 |
$ 206 |
$ 371 |
$ 92.69 |
4 |
$ 371 |
$ 803 |
$100.37 |
8 |
$ 803 |
$1,669 |
|
|
$1,380 |
$57 |
$37.41 |
6 |
$ 224 |
$ 533 |
|
|
|
$ 112 |
$ 92.69 |
4 |
$ 371 |
$ 102 |
$101.66 |
2 |
$ 203 |
$ 331 |
$ 91.02 |
6 |
$ 546 |
$1,135 |
|
|
$1,344 |
DATA SOURCES: Price data are from Economic Research Service, USDA (ERS, 2004b, 1999 price data; Oliveira et al., 2001, 2000 infant formula price data); ACNielsen Homescan (ACNielsen, 2001, price data for 2001 obtained through ERS, USDA); and Bureau of Labor Statistics, U.S. Department of Labor (BLS, 2004a, 2002 price data). |
milk in the current and proposed packages, there may be some cost differences between the current and revised program costs due to shifts of the prices of these foods that represent a relatively large component of program package costs.
In order to evaluate the sensitivity of program costs to price changes, the committee considered (1) a decreased rebate rate for infant formula, shifting the post-rebate cost from 0.321 to 0.50 times the market price, and (2) an increase in milk prices of 20 percent. The effect of the reduced rebate rate on overall program food costs leads to an 8.2 percent increase for the current food packages and a 6.1 percent increase for the revised food packages. An increase of 20 percent in milk prices increases the program food cost by 5.6 percent for the current set of food packages and 4.5 percent for the set of revised food packages. These changes illustrate that, compared with the current set of food packages, the set of revised food
packages is less sensitive to changes in price for these two food components. The greater diversity of food items included in the revised food packages will tend to reduce the sensitivity of the food package cost to a change in the price of any single food item.
SUMMARY
The revised food packages are cost-neutral. Using identical methods to estimate the average cost per participant of the current and revised WIC food packages, the committee found essentially no change. In particular, the estimated average 2002 cost per participant for the current set of food packages was $34.76 per month, and for the set of revised food packages was $34.57 per month (and in the range of $34.03–$34.95), approximately equal to the estimated cost of the current package. Thus, given the same methods and prices for comparison, and assuming no shifts in participation by program categories, the changes proposed are likely to have little effect on program food costs. Furthermore, compared to the cost of the current food packages, the cost of the revised food packages would change less in response to changes in the costs of dairy products and infant formula.
The changes in the food packages greatly increase the relative market value (i.e., pre-rebate price) of the combined packages for the fully breastfeeding mother/infant pair; this change in the set of revised food packages could serve as an increased incentive for breastfeeding.
The costing method used includes a cash-value voucher that can be used to obtain a variety of fresh fruits and vegetables of the participants’ choosing; the addition of the cash-value voucher could increase the cultural acceptability of the WIC food packages. Because an increase in the cost of fresh produce would lead to a reduced amount of fruits and vegetables that could be obtained with the cash-value voucher and this, in turn, would reduce the nutrient content of the packages, the committee recommends review and revision of the total value of the cash-value vouchers for fresh fruits and vegetables every 1 to 3 years.
The cost evaluation of the revised food packages encompassed major changes directed toward allowing healthier choices (e.g., the addition of fruits and vegetables, reduced-fat rather than whole milk for participants 2 years of age and older; allowed breakfast cereals are whole grain). These changes could serve to improve the diets of WIC participants. The cost evaluation also included specific amounts of substitutions that were requested by participants (e.g., allowing yogurt, tofu, and soy beverage [“soy milk”] as a substitute for milk); these substitutions could increase the incentive value of the food packages for families to participate in the WIC program.