Skip to main content
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Dec 1.
Published in final edited form as: Am J Prev Med. 2024 Jul 19;67(6):906–915. doi: 10.1016/j.amepre.2024.07.013

Examining the 2021 Cash Value Benefit Increase and WIC Participant Food Purchases

Emily W Duffy 1,2, Shu Wen Ng 1,2, Maxime Bercholz 2, Cassandra R Davis 2,3, Molly De Marco 1, Marissa G Hall 2,4,5, Joanna Maselko 2,6, Lindsey Smith Taillie 1,2
PMCID: PMC11585448  NIHMSID: NIHMS2016451  PMID: 39032521

Abstract

Introduction:

The Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) provides nutritious foods to more than six million low-income families. In June 2021, WIC increased the amount of money provided for fruits and vegetables (FV), but studies have not investigated whether this increase changed WIC participant FV purchases. The objective was to estimate the association between the FV funding increase and WIC shopper FV purchases.

Methods:

This evaluation uses longitudinal food transaction data from 496 stores in a grocery store chain in North Carolina between June 2020 and April 2022 and a propensity score weighted difference-in-differences (DID) approach to estimate the association between the FV funding increase and FV and other food group purchases. Analyses were conducted in 2023 and 2024.

Results:

WIC shoppers’ WIC-eligible FV purchases increased by $12.4 per shopper per month (Adjusted Confidence Interval (ACI), $12.0 to $12.9) after the FV benefit increase, which was $9.3 (ACI, $8.7 to $10.0) more than non-WIC shoppers. The monthly volume and unique varieties of FV purchased also increased more among WIC shoppers relative to non-WIC shoppers (DID 67.1 ounces (ACI, 61.9 to 72.3) and DID 2.1 varieties (ACI, 2.0 to 2.3)). There were also increases in the volume of processed foods (DID 22.0 ounces (ACI, 17.2, 26.9)) and sugar-sweetened beverages (DID 49.1 ounces (ACI, 33.4, 64.9)) purchased among WIC shoppers relative to non-WIC shoppers.

Conclusions:

Increased WIC FV funding was associated with greater FV purchases. Research is needed to understand the effects of this policy on total dietary intake.

Introduction

Adequate fruit and vegetable (FV) consumption is critical for preventing diet-related chronic diseases throughout the lifecourse.1–3 However, most pregnant and postpartum people and young children in the US do not consume enough FV.3,4 There are also disparities in FV intake, with individuals with low incomes being less likely to consume adequate FV.5–7

The Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) provides access to nutritious foods (i.e., food packages) such as FV, healthcare referrals, and nutrition education to more than six million families with low incomes at nutritional risk.8 Beginning in 2009, WIC provided $9–11/person/month for FV through a component of the food package called the Cash Value Benefit (CVB), which equates to less than one cup of FV per person per day.9

Increasing the CVB has been proposed as an important policy measure to reduce FV consumption disparities;9 however, there has been little opportunity to understand whether a CVB increase would affect FV purchases of WIC families. Additionally, considering that young children more often under-consume vegetables than fruits, but both are vital for healthy development,3,4 it is important to understand if both fruit purchases and vegetable purchases would increase with a higher CVB. Finally, early exposure to a variety of FV is critical for establishing healthy dietary behaviors,10,11 so it is also worth examining whether families would use a higher CVB to increase the diversity of FV purchased.

Recent changes to the WIC food packages provide a unique opportunity to address these research questions. In response to the COVID-19 pandemic, the United States Department of Agriculture (USDA) increased the CVB in June 2021, initially to $35/month/person and eventually to $24–47/month/person (Appendix Figure 1).12 Studies have examined the effects of the 2021 CVB increase and found high satisfaction with the CVB increase and increases in FV consumption.13–17 These data suggest this policy change has the potential to mitigate income-based FV consumption disparities; however, in addition to self-report data among WIC participants, we need objective measures of food purchasing behavior and studies with non-WIC comparison groups to differentiate policy effects from secular trends. Given that WIC participation is not randomized and all WIC participants received the CVB increase at the same time, quasi-experimental methods provide a useful approach to assess associations between the CVB increase and FV purchases.18

The aim of this study is to estimate the association between the 2021 CVB increase and WIC shoppers’ purchases of CVB-eligible FV using longitudinal food purchasing data from a large grocery store retailer in North Carolina using a propensity score weighted difference-in-differences (DID) approach. A secondary aim was to investigate the association between the CVB increase and purchases of processed foods and sugar sweetened beverages (SSBs).

Methods

Longitudinal loyalty card food purchasing data19,20 from June 2020 to April 2022 from 496 stores belonging to one of the top two food retailers in North Carolina located in 86 of North Carolina’s 100 counties was used. Most WIC benefits are redeemed at large grocery stores, and this retailer is the preferred retailer by NC WIC participants due to clear shelf labeling of WIC-approved products.16,21 These data include every item purchased in each shopping episode including barcode or item number, item size, item description, price, unit of measure, quantity sold, payment types used, store location, date of sale, and loyalty-card ID for the transaction. If multiple payment types (e.g., debit card, WIC Electronic Benefit Transfer (EBT) card, SNAP EBT card) were used in a transaction, the specific food items that were purchased with each payment type cannot be determined, only that certain payment type(s) were used in a given transaction. Online purchases made directly with the retailer are included, though they comprise 1% or less of all transactions. Using each product’s unique product code (UPC) or product look up (PLU) code, products have been linked to nutrition and product data from the Mintel Global New Product Database, Label Insight, and USDA’s Food and Nutrient Database for Dietary Studies.22–24 These data do not include food purchases made without loyalty cards, though the majority of grocery store transactions involve the use of a loyalty card (90–95% based on some estimates).25,26

These analyses focus on five food groups of public health interest and directly related to the CVB: CVB-eligible fruits, CVB-eligible nutrient dense vegetables (e.g., carrots, spinach), CVB-eligible starchy vegetables (e.g., potatoes, corn), non-essential packaged, processed foods (processed foods) (e.g., salty snacks, desserts), and SSBs (Appendix Table 1). Documentation from the North Carolina Department of Health and Human Services and nutrition label and product data were used to determine CVB-eligible fruits and vegetables.27,28 Vegetables were categorized as starchy or nutrient-dense based on existing USDA definitions.4,23 In North Carolina, fresh, frozen, and canned FV without added sugar, fats, salt, or artificial sweeteners are CVB-eligible.27,28

Study Sample

First, WIC shoppers were identified by the payment type used in a transaction. If a shopper paid with a WIC EBT card at least once a month, they were deemed a WIC shopper for that month. In the initial sample, all observations from shoppers that paid with WIC at least one time in the study period were included (n= 261,174 shoppers, 4,333,495 shopper-month observations). Then, a random sample of non-WIC shoppers equal to the number shoppers that ever used WIC during this time period was selected and initially all observations from this sample were included (n=261,174 shoppers, 2,768,596 shopper-month observations). These non-WIC shoppers may include individuals that are and are not WIC-eligible. Non-WIC comparison shoppers were used to, in-part, control for secular trends in purchases.

Several additional exclusion criteria were applied to these initial samples of WIC and non-WIC shopper observations. Observations from loyalty cards (in WIC and non-WIC groups) that were in the top 1% of expenditures in a given month were excluded because these are likely ‘store cards’ used by cashiers for many shoppers. Observations from shoppers participating in other, active FV incentive programs were also excluded (Appendix Figure 2). The WIC sample was then restricted to shoppers that used WIC EBT at this retailer at least once every three months (n=536,349 shopper-month observations) and the sample of non-WIC shoppers was restricted to those that shopped at this retailer at least once every three months (n=1,894,056 shopper-month observations). Every three months was selected because we wanted to focus the analyses on individuals using WIC benefits regularly, but did not want to exclude ‘real’ WIC shoppers that redeemed their benefits at a different retailer in a given month.

A comparison of shopper characteristics of the unweighted initial and analytical samples of observations from WIC and non-WIC shoppers can be found in Appendix Table 2.

Measures

Primary outcomes were expenditures ($) and volume (ounces) purchased per shopper per month of total CVB-eligible FV, CVB-eligible fruits, CVB-eligible nutrient-dense vegetables, and CVB-eligible starchy vegetables. The CVB is the only component of the WIC food packages that is disbursed in dollars as opposed to a certain number of packages (e.g., 1 quart yogurt), so it is important to measure how expenditures on FV changed over time. Prices of fruits and vegetables vary across form (e.g., fresh vs canned) and variety (e.g., organic vs. conventional) and FV recommendations are in terms of cups per day, so it is also important to measure volume of FV purchased. Also considering far fewer children achieve vegetable recommendations than fruit recommendations and differences in nutrition guidance and nutrient composition in starchy and nutrient-dense vegetables, analyses estimated the association between the CVB increase and changes in fruit, nutrient-dense vegetable and starchy vegetable purchases separately.3,4

Secondary outcomes included number of unique varieties of total CVB-eligible FV, CVB-eligible fruits, CVB-eligible nutrient-dense vegetables, and CVB-eligible starchy vegetables given self-reported increases in the variety of FV consumed after the CVB increase and the importance of eating a variety of FV for healthy taste preference development.10,11,13,16 Secondary outcomes also included volume of processed foods and SSBs purchased per month as other studies have reported WIC participants may have used the increased CVB to reduce their out-of-pocket FV spending and used the saved funds to purchase other food categories.13–16

The exposure of interest is the policy change that led to the CVB increase that occurred in June 2021 in NC. The exposure is operationalized by using a binary indicator variable for months before the CVB increase (June 2020-May 2021) and months after the increase (June 2021-April 2022).

Statistical Analysis

Unadjusted mean purchases per month for all outcomes before and after the CVB increase were compared by regressing each outcome on an indicator variable for pre/post policy, an indicator variable for WIC or non-WIC shopper and their interaction with standard errors clustered at the shopper level. An overlap weighted DID approach was used to estimate the average effect of the CVB increase among WIC shoppers relative to non-WIC comparison shoppers.19,29 Appendix Figures 3 and 4 demonstrate parallel trends in the pre-period. Overlap weights are propensity-score based weights that perfectly balance observed covariates in the pre-period between our WIC and non-WIC group using pre-policy data while limiting the influence of observations with extreme propensity score values.29 To estimate shopper-level propensity scores, logistic regression models where the outcome was a binary variable for WIC vs non-WIC shopper and the predictors were the shopping characteristics listed in Appendix File 1 were used.30 Propensity scores were used to create shopper-level overlap weights, which assign weights proportional to the shopper belonging to the opposite group. Overlap weights were used to simulate randomization and strengthen the internal validity as it relates to the parallel trends assumption. The overlap weights also limit the potential bias introduced by secular events or other policy changes that may differentially influence the WIC and non-WIC groups. As with all loyalty card data, the dataset does not include shopper demographic characteristics, so those characteristics were not used in weighting.29

Due to a relatively large number of non-purchasers of FV (25–58% of observations depending on the outcome (e.g., starchy vegetables, nutrient-dense vegetables), multivariable overlap weighted two-part models were used for expenditure and volume outcomes with the first part being a probit model and the second part being a generalized linear model with a gamma distribution and log link with standard errors clustered at the shopper level to account for repeated observations within shoppers over time.31 Select coefficients from the first part, probit models, can be found in Appendix Table 3. Multivariable overlap weighted Poisson models with standards errors clustered at the shopper-level were used for the variety outcomes since they are overdispersed counts.32 All models included an indicator for month to control for seasonality33 and an indicator for each shopper’s monthly top store, or the store where they spent the most money each month, to control for time-invariant store environment characteristics. To aid with interpretation, marginal effects were estimated from the two-part and Poisson models. The Bonferroni correction was used to adjust for multiple comparisons, grouping primary outcomes (8 outcomes, two-sided alpha of 0.00625) and secondary outcomes (6 outcomes, two-sided alpha of 0.00833) together for adjustment. All analyses were conducted in 2023 and 2024. This study was reviewed by the University of North Carolina at Chapel Hill Institutional Review Board and deemed non-human subjects research.

In North Carolina, there were implementation challenges when the federal CVB amount changed from $35/person/month to $24–47/person/month starting October 1, 2021. North Carolina was not able to implement this change quickly enough, so the CVB amount returned to $9–11/person/month for October 2021 before increasing to $24–47/person/month in November 2021. Therefore, in sensitivity analyses to understand the effect of the policy as intended, October 2020 (for seasonality) and October 2021 were removed from the dataset and all models were re-run as described above.

Results

In the unweighted sample pre-policy, WIC shoppers had higher numbers of shopping episodes and episodes using SNAP benefits each month as well as larger total food expenditures and absolute expenditures on each food group compared to non-WIC shoppers (Table 1). WIC shoppers had lower percentages of their total food purchases coming from the five food groups examined compared to non-WIC shoppers. Upon applying overlap weights, all measured shopper characteristics were balanced in the pre-policy period as expected (Table 1).

Table 1.

Unweighted and weighteda means of WIC and non-WIC shoppers in the pre-CVB increase period (n 1,273,443 shopper month observations)

Characteristic Unweighted Weighted
Non-WIC Shoppers WIC Shoppers Non-WIC Shoppers WIC Shoppers
Mean(SD)
Number of shopping episodes/month 4.7 (3.6) 6.5 (4.3) 6.3 (4.6) 6.2 (4.1)
Number of episodes using SNAP/month 0.5 (1.6) 2.7 (3.3) 1.8 (3.1) 1.8 (2.8)
Percent of transactions using cash 20.5 (30.9) 14.1 (18.4) 16.1 (23.3) 16.0 (21.0)
Percent of transactions using debit 25.3 (39.3) 6.3 (15.6) 9.2 (22.8) 9.2 (19.8)
Percent of transactions using credit 42.3 (41.0) 25.2 (24.7) 31.2 (34.6) 31.2 (28.2)
Food expenditures ($)/month 189.8 (176.2) 336.7 (233.2) 301.4 (225.7) 297.1 (214.0)
Processed food expenditures ($)/month 46.2 (47.9) 68.4 (60.0) 67.7 (58.9) 66.6 (58.7)
SSB expenditures ($)/month 22.3 (32.6) 36.5 (40.8) 35.0 (41.7) 34.5 (39.0)
Fruit Expenditures ($)/month 8.5 (14.3) 15.4 (18.2) 13.4 (21.6) 13.3 (16.6)
Nutrient Dense Vegetable Expenditures ($)/month 10.1 (14.4) 13.4 (14.9) 13.5 (15.9) 13.3 (15.7)
Starchy Vegetable Expenditures ($) 2.2 (4.0) 3.2 (4.8) 3.2 (4.8) 3.1 (4.9)
% of Volume Purchased from Processed Foods 14.3 (14.7) 9.8 (7.9) 11.2 (8.8) 11.2 (9.6)
% of Volume Purchased from SSBs 25.0 (24.3) 23.1 (18.2) 25.1 (20.4) 25.0 (19.6)
% of Volume Purchased from Fruit 5.2 (9.1) 5.0 (6.1) 4.8 (7.4) 4.8 (6.3)
% of Volume Purchased from Nutrient Dense Vegetables 4.9 (8.2) 3.5 (4.7) 3.9 (5.3) 3.9 (5.6)
% of Volume Purchased from Starchy Vegetables 1.8 (4.6) 1.4 (2.9) 1.6 (3.2) 1.6 (3.3)
Fruit varieties 2.1 (2.6) 3.6 (3.3) 3.0 (3.2) 3.2 (3.1)
Nutrient Dense Vegetable Varieties 3.9 (4.6) 5.2 (4.8) 5.2 (5.0) 5.1 (4.9)
Starchy Vegetable Varieties 0.7 (1.1) 1.0 (1.3) 1.0 (1.3) 1.0 (1.3)

SD: standard deviation; SNAP: Supplemental Nutrition Assistance Program; SSB: sugar sweetened beverages

a

Weighted means are not identical between WIC and non-WIC groups because overlap weights were created and applied at the shopper-level and the dataset is at the shopper-month level

After the CVB increase, WIC shopper purchases of CVB-eligible FV at this retailer increased by $12.4 per month (99.4% CI 12.0, 12.9), which was $9.3 (99.4% CI 8.7,10.0) greater than the increase in expenditures among non-WIC shoppers (Figure 1, Table 2). Most of this increase came from increases in fruit expenditures, followed by nutrient-dense vegetables (Figure 2, Table 2). WIC shoppers’ purchases of FV increased by 68.8 ounces per month (99.4% CI 65.0, 72.5) after the CVB increase. This increase was 67.1 ounces (99.4% CI 61.9, 72.3) greater than the change in FV purchases among non-WIC shoppers (Figure 1, Table 2). WIC shoppers purchased 1.9 more varieties of FV per month (99.2% CI 1.8, 2.0) after the CVB increase, which was 2.1 (99.2% CI 2.0, 2.3) more than non-WIC shoppers who experienced a decrease in the varieties of FV purchased (Figure 1, Table 2). After the CVB increase, WIC shoppers also purchased 22.0 more ounces (99.2% CI 17.2, 26.9) of processed foods than non-WIC shoppers. WIC shoppers purchased 49.1 more ounces of SSBs (99.2% CI 33.4, 64.9) than non-WIC shoppers who experienced a 17.9 ounce (99.2% CI −29.1, −6.7) decrease in SSB purchases.

Figure 1.

Figure 1.

Changes in (A) dollar expenditures (B) ounces and (C) unique varieties of total CVB-eligible fruits and vegetables per month per shopper after the CVB increase among WIC shoppers, non-WIC shoppers, and the difference-in-differences

*Statistically significant after Bonferroni adjustment for multiple comparisons

Estimates are marginal effects of the pre/post variable holding the group variable at WIC and non-WIC, and their difference, from overlap weighted and multivariable two-part (dollars and ounces) and Poisson (varieties) models

Table 2.

Unweighted mean purchases pre and post CVB increase, unweighted and weighted and adjusted pre-post differences and difference-in-differences

WIC Shoppers (n=536,349 shopper-month observations) Non-WIC Shoppers (n=1,894,056 shopper-month observations) Difference-in-Differences
Outcome Pre-Policy Post-Policy Unweighted Difference Weighted Differencea Pre-Policy Post-Policy Unweighted Difference Weighted Differencea Unweighted DID Weighted DIDa
Total CVB FV dollars 32.0
(31.7, 32.3)
44.8
(44.4,45.1)
12.8
(12.6, 13.0)
$12.4
(12.0, 12.9)
20.7
(20.6, 20.8)
23.1
(22.9, 23.3)
2.4
(2.3, 2.5)
$3.1
(2.7, 3.6)
10.4
(10.1, 10.6)
$9.3
(8.7,10.0)
Fruit dollars 15.4
(15.2, 15.5)
23.8
(23.6, 24.1)
8.5
(8.3, 8.6)
$8.3
(8.0, 8.6)
8.5
(8.4, 8.6)
10.0
(9.9,10.1)
1.5
(1.5, 1.6)
$2.2
(1.9 ,2.4)
6.9
(6.8, 7.1)
$6.1
(5.7, 6.5)
Nutrient dense vegetable dollars 13.4
(13.2, 13.5)
17.2
(17.0, 17.3)
3.8
(3.7, 3.9)
$3.7
(3.4, 3.9)
10.1
(10.0, 10.1)
10.9
(10.8, 11.0)
0.8
(0.8, 0.9)
$1.0
(0.8, 1.2)
3.0
(2.9, 3.1)
$2.7
(2.4, 3.0)
Starchy vegetable dollars 3.2
(3.2, 3.3)
3.8
(3.7, 3.8)
0.5
(0.5, 0.6)
0.5
(0.5,0.6)
2.2
(2.1, 2.2)
2.2
(2.2, 2.2)
0.0
(0.0, 0.1)
0.0
(−0.06,0.06)
0.5
(0.4, 0.5)
$0.5
(0.5, 0.6)
Total CVB FV oz 280.0
(277.6, 282.5)
352.9
(359.1, 355.7)
72.8
(71.0, 74.7)
68.8
(65.0,72.5)
175.6
(174.5, 176.7)
183.3
(182.1, 184.5)
7.7
(6.9,8.4)
1.7
(−2.2, 5.5)
65.2
(63.2, 67.1)
67.1
(61.9, 72.3)
Fruit oz 137.0
(135.6, 138.5)
184.9
(183.2, 186.6)
47.8
(46.7, 49.0)
45.6
(43.5,47.8)
74.7
(74.1, 75.3)
79.8
(79.1, 80.4)
5.1
(4.6, 5.5)
2.2
(−0.01, 4.5)
42.8
(41.6, 44.0)
43.4
(40.3, 46.5)
Nutrient dense vegetable oz 98.5
(97.5, 99.5)
115.0
(113.8, 116.1)
16.5
(15.7, 17.2)
15.0
(13.4, 16.6)
70.6
(70.1, 71.1)
71.6
(71.1, 72.1)
1.0
(0.7, 1.3)
−1.2
(−2.8, 0.4)
15.5
(14.7, 16.3)
16.2
(14.0, 18.4)
Starchy vegetable oz 44.5
(44.0, 45.1)
53.0
(52.4, 53.6)
8.5
(8.0, 9.0)
7.8
(6.8, 8.8)
30.3
(30.1, 30.6)
31.9
(31.6, 32.2)
1.6
(1.4, 1.8)
0.7
(−0.2, 1.6)
6.9
(6.4, 7.4)
7.1
(5.7, 8.4)
Processed food oz 304.2
(301.6, 306.8)
333.9
(331.1, 336.6)
29.7
(28.0, 31.3)
30.0
(26.5, 33.6)
204.5
(203.4, 205.6)
214.4
(213.2, 215.7)
10.0
(9.2, 10.7)
8.0
(4.6, 11.5)
19.7
(17.9, 21.5)
22.0
(17.2, 26.9)
SSB oz 842.7
(833.1, 852.2)
864.0
(854.1, 873.9)
21.4
(15.6, 27.1)
31.2
(19.7, 42.7)
520.9
(516.6, 525.2)
521.6
(517.2, 526.0)
0.7
(−1.7, 3.2)
−17.9
(−29.1, −6.7)
20.6
(14.4, 26.8)
49.1
(33.4, 64.9)
Total CVB FV varieties 9.8
(9.7, 9.8)
11.7
(11.6, 11.8)
1.9
(1.9, 2.0)
1.9
(1.8, 2.0)
6.7
(6.7, 6.7)
6.5
(6.5, 6.6)
−0.2
(−0.2, −0.1)
−0.3
(−0.4, −0.2)
2.1
(2.0,2.1)
2.1
(2.0, 2.3)
Fruit varieties 3.6
(3.6, 3.6)
4.8
(4.8, 4.8)
1.2
(1.2, 1.2)
1.2
(1.1, 1.2)
2.1
(2.1, 2.1)
2.1
(2.1, 2.2)
0.1
(0.1, 0.1)
0.07
(0.03, 0.1)
1.1
(1.1, 1.2)
1.1
(1.1, 1.2)
Nutrient dense vegetable varieties 5.2
(5.1, 5.2)
5.8
(5.8, 5.9)
0.7
(0.6, 0.7)
0.6
(0.6, 0.7)
3.9
(3.9, 3.9)
3.8
(3.7, 3.8)
−0.2
(−0.2, −0.1)
−0.2
(−0.3, −0.2)
0.8
(0.8, 0.9)
0.9
(0.8, 1.0)
Starchy vegetable varieties 1.0
(1.0, 1.0)
1.1
(1.1, 1.1)
0.1
(0.1, 0.1)
0.1
(0.1, 0.1)
0.7
(0.7, 0.7)
0.7
(0.6, 0.7)
−0.1
(−0.1, 0.0)
−0.1
(−0.1, −0.1)
0.1
(0.1, 0.1)
0.2
(0.1, 0.2)
a

Significance threshold for primary outcomes (FV dollars and ounces) is 0.00625 so 99.4% confidence intervals are presented, significance thresholds for secondary outcomes (FV variety, processed food and SSB ounces) is 0.00833 so 99.2% confidence intervals are presented

Note: boldface indicates statistical significance after Bonferroni adjustment for multiple comparisons

CVB: cash value benefit; FV: fruits and vegetables; SSB: sugar sweetened beverages; oz: ounces

Weighted estimates are marginal effects of the pre/post variable holding the group variable at WIC and non-WIC, and their difference, from overlap weighted and multivariable two-part (dollars and ounces) and Poisson (varieties) models

Figure 2.

Figure 2.

Figure 2.

Figure 2.

Changes in (A) dollar expenditures (B) volume (oz) and (C) unique varieties purchased of CVB-eligible fruits, nutrient-dense vegetables, starchy vegetables, processed foods, and SSBs per month per shopper after the CVB increase among WIC shoppers, non-WIC shoppers, and the difference-in-differences

*Statistically significant after Bonferroni adjustment for multiple comparisons

Estimates are marginal effects of the pre/post variable holding the group variable at WIC and non-WIC, and their difference, from overlap weighted and multivariable two-part (dollars and ounces) and Poisson (varieties) models

Results when excluding October 2020 and 2021 did not differ substantively (Appendix Table 4).

Discussion

After the CVB increase, there were estimated increases of $9.3 and 67.1 ounces per month (about 8.5 cups/month) in CVB-eligible FV purchases among WIC shoppers compared to non-WIC shoppers. These are meaningful increases considering recommended FV intake for a young child ranges from 1⅓ to 3 cups per day and very few children currently meet FV recommendations, although we cannot be sure that purchases translate directly to consumption among children.3 The findings are consistent with existing studies in which WIC participants reported they consumed more FV after the CVB increase.13,14,16

The observed increase in FV expenditures is less than the per person amount of the CVB increase. This may be attributable to shoppers replacing previous out-of-pocket spending on FV with CVB spending on FV. Shoppers may have also redeemed part of their CVB at other retailers since this dataset only includes purchases from one grocery store chain. Another possible explanation is low redemption; however, this may be less likely as the CVB has a high redemption rate compared to other food package components and one study found increases in CVB redemption after the 2021 CVB increase.9,34 Future studies using more complete food purchasing data from all retailers and WIC administrative data are needed to fully understand the effects of this policy change.

Increases in varieties of FV purchased were observed among WIC shoppers. Though we cannot be sure that increased purchases translated to increased consumption, other studies have reported increases in FV consumption after the CVB increase.13–16,35 This finding is consistent with qualitative studies in which WIC participants reported the CVB increase allowed them to purchase previously cost-prohibitive FV varieties.13,14,16 Given the importance of consuming a variety of FV in pregnancy, postpartum, and early childhood, estimating change in varieties purchased is a unique contribution.

Small increases in processed food and SSB purchases were observed among WIC shoppers. Some WIC participants have reported using the CVB increase to substitute WIC funds for FV they ordinarily would have purchased with non-WIC funds and using the saved non-WIC funds to purchase meat, seafood, grains, or other household items.13–16 The relationship between the CVB increase and changes in the total diet should be explored in future research using datasets with more food groups available.

This work has implications for current and future WIC policy such as revisions to the WIC food packages and efforts to increase redemption of the CVB. Additionally, given the small increases observed in processed food and SSB purchases, it may be wise to pair FV incentives with disincentives for ultraprocessed foods such as restrictions on marketing to children, excise taxes or warning labels.36

Limitations

As with all loyalty card data, demographic data are not available, so racial, ethnic, or geographic differences in purchases cannot be measured despite documented differences in experiences with the pandemic and policy changes by these characteristics.37–39 Additionally, this dataset only includes food purchases made at this retail chain with loyalty cards, but the majority of grocery transactions involve loyalty cards (90–95% of transactions according to some estimates25,26). Moreover, this retailer is one of the top two food retailers in North Carolina and the preferred retailer for WIC shoppers, so this dataset likely captures a majority of WIC shoppers’ food purchases.16 The dollar amount of each household’s CVB increase is unknown because it was dependent on household composition. Also, substitution of payment methods for purchases of specific food groups and purchases of a wide variety of food groups cannot be measured.

Finally, while it is the goal of the non-WIC shoppers and the overlap weights to control for unmeasured, time invariant characteristics and the effects of secular trends and other policy changes, shoppers are weighted based on pre-CVB increase shopping characteristics, not demographic characteristics. Therefore, it is possible that some of the changes observed in purchases are due, in part, to differences in unmeasured demographic characteristics or to secular events or policy changes that may have differentially affected WIC and non-WIC shoppers, such as the Expanded Child Tax Credits. However, the weights account for elements like SNAP use and total food expenditures that are correlated with income and household composition, so it is unlikely this introduces a major source of bias. Despite these limitations, the use of a quasi-experimental design and overlap weights is a robust approach to evaluating a national-level policy change.18

Conclusions

This study adds to the existing research on the 2021 CVB increase that suggests this policy was associated with an increase in the amount and variety of FV purchased and consumed by WIC participants.13–16,34 While this policy appears to have been effective, a comprehensive suite of nutrition policies that promote FV and disincentivize ultraprocessed foods and beverages will likely be needed to reduce disparities in FV consumption in early childhood and bring young children’s total dietary intake closer in line with recommendations.

Supplementary Material

Appendix

Acknowledgements

We are very grateful to the participating grocery chain for sharing their transaction data that allowed for this study. We also wish to thank Neal Curran at Reinvestment Partners, Amy Lo, Caitlin Lowery, and Alex Ross for research assistance and Sierra Mullen for data management assistance.

Funding

Funding that made this study possible comes from the Carolina Population Center NICHD-NRSA Population Research Training (T32 HD007168), Carolina Population Center (NIH P2C HD050924), a grant from the T32 Cancer Health Disparities Training Grant from the National Cancer Institute of the National Institutes of Health (T32CA128582), and earlier funding from NC TraCS (National Center for Advancing Translational Sciences (NIH UL1TR001111), the Center for Health Equity Research at the University of North Carolina, and Arnold Ventures. K01HL147713 from the National Heart, Lung, and Blood Institute of the NIH supported Marissa Hall’s time writing the paper. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH. Article contents have not been previously published elsewhere.

Footnotes

Declarations of Interest

No financial disclosures were reported by the authors of this paper.

References

  • 1.Aune D, Giovannucci E, Boffetta P, et al. Fruit and vegetable intake and the risk of cardiovascular disease, total cancer and all-cause mortality-a systematic review and dose-response meta-analysis of prospective studies. Int J Epidemiol 2017;46(3):1029–1056. 10.1093/ije/dyw319. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Wang X, Ouyang Y, Liu J, et al. Fruit and vegetable consumption and mortality from all causes, cardiovascular disease, and cancer: systematic review and dose-response meta-analysis of prospective cohort studies. Bmj 2014;349:g4490. 10.1136/bmj.g4490. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Dietary Guidelines for Americans 2020–2025. United States Department of Health and Human Services, United States Department of Agriculture;2020. [Google Scholar]
  • 4.Hamner HC. Fruit, Vegetable, and Sugar-Sweetened Beverage Intake Among Young Children, by State—United States, 2021. MMWR Morbidity and Mortality Weekly Report. 2023;72. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Wambogo EA, Ansai N, Ahluwalia N, Ogden CL. Fruit and vegetable consumption among children and adolescents in the United States, 2015–2018. 2020. [PubMed] [Google Scholar]
  • 6.Drewnowski A, Rehm CD. Socioeconomic gradient in consumption of whole fruit and 100% fruit juice among US children and adults. Nutr J 2015;14(1):1–9. 10.1186/1475-2891-14-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Fadeyev K, Nagao-Sato S, Reicks M. Nutrient and Food Group Intakes among U.S. Children (2–5 Years) Differ by Family Income to Poverty Ratio, NHANES 2011–2018. Int J Environ Res Public Health. 2021;18(22). 10.3390/ijerph182211938. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.United States Department of Agriculture. WIC Program:Total Participation. 2020.
  • 9.National Academies of Sciences, Engineering, and Medicine. Review of WIC Food Packages: Improving Balance and Choice: Final Report. Washington (DC): National Academies Press (US). 2017. [PubMed] [Google Scholar]
  • 10.Mennella JA, Jagnow CP, Beauchamp GK. Prenatal and postnatal flavor learning by human infants. Pediatrics. 2001;107(6):E88. 10.1542/peds.107.6.e88. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Mennella JA, Reiter AR, Daniels LM. Vegetable and Fruit Acceptance during Infancy: Impact of Ontogeny, Genetics, and Early Experiences. Adv Nutr 2016;7(1):211s–219s. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.USDA to Incentivize Purchase of Fruits and Vegetables under WIC for 4 Months with American Rescue Plan Funding [press release]. 2020.
  • 13.Halverson MM, Karpyn A. WIC Participants’ Perceptions of the Cash-Value Benefit Increase during the COVID-19 Pandemic. Nutrients. 2022;14(17):3509. 10.3390/nu14173509. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Martinez CE, Ritchie LD, Lee DL, Tsai MM, Anderson CE, Whaley SE. California WIC Participants Report Favorable Impacts of the COVID-Related Increase to the WIC Cash Value Benefit. Int J Environ Res Public Health. 2022;19(17):10604. 10.3390/ijerph191710604. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Gago C, Colchamiro R, May K, Rimm EB, Kenney EL. Caregivers’ Perceived Impact of WIC’s Temporary Cash-Value Benefit (CVB) Increases on Fruit and Vegetable Purchasing, Consumption, and Access in Massachusetts. Nutrients. 2022;14(23):4947. 10.3390/nu14234947. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Duffy EW, Vest DA, Davis CR, et al. “I think that’s the most beneficial change that WIC has made in a really long time”: Perceptions and Awareness of an Increase in the WIC Cash Value Benefit. Int J Environ Res Public Health. 2022;19(14):8671. 10.3390/ijerph19148671. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Ritchie L, Lee D, Sallack L, et al. Multi-State WIC Participant Satisfaction Survey: Cash Value Benefit Increase During COVID. 2021. http://aws.upl.s3.amazonaws.com/nwica.org/nwamulti-state-wic-participant-satisfaction-surveynationalreportfinal.pdf. Accessed June 12,2022.
  • 18.Fox MK, Hamilton W, Lin B-H. Effects of food assistance and nutrition programs on nutrition and health. Food Assistance and Nutrition Research Report No. FANRR19–4. US Department of Agriculture. 2004. [Google Scholar]
  • 19.Berkowitz SA, Curran N, Hoeffler S, Henderson R, Price A, Ng SW. Association of a Fruit and Vegetable Subsidy Program With Food Purchases by Individuals With Low Income in the US. JAMA Netw Open. 2021;4(8):e2120377. 10.1001/jamanetworkopen.2021.20377. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Lowery CM, Henderson R, Curran N, Hoeffler S, De Marco M, Ng SW. Grocery Purchase Changes Were Associated With A North Carolina COVID-19 Food Assistance Incentive Program: Study examines a COVID-19 food assistance incentive program in North Carolina. Health Affairs. 2022;41(11):1616–1625. 10.1377/hlthaff.2022.00902. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Vercammen K, Grummon AH, Zatz LY, et al. A Descriptive Analysis of Redemption Patterns by Vendor Type Among: WIC Participants in Massachusetts. Journal of the Academy of Nutrition and Dietetics. 2023;123(4):626–636. e2. 10.1016/j.jand.2022.09.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Mintel Global New Products Database. London, UK. Continuous Edition. https://www.mintel.com/global-new-products-database. Accessed August 27, 2022. [Google Scholar]
  • 23.USDA Food and Nutrient Database for Dietary Studies, 1.0. 2004. Beltsville, MD: Agricultural Research Service, Food Surveys Research Group. [Google Scholar]
  • 24.Label Insight. Chicago, Illinois. Continuous Edition. https://www.labelinsight.com/. Accessed August 27, 2022. [Google Scholar]
  • 25.Andreyeva T, Tripp AS. The healthfulness of food and beverage purchases after the federal food package revisions: The case of two New England states. Prev Med Oct 2016;91:204–210. doi: 10.1016/j.ypmed.2016.08.018. [DOI] [PubMed] [Google Scholar]
  • 26.Andreyeva T, Luedicke J. Incentivizing fruit and vegetable purchases among participants in the Special Supplemental Nutrition Program for Women, Infants, and Children. Public Health Nutr Jan 2015;18(1):33–41. doi: 10.1017/s1368980014000512 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.WIC North Carolina. Healthy Foods Shopping Guide. 2022.
  • 28.WIC North Carolina. North Carolina WIC Approved Product List. In:2020–2021.
  • 29.Thomas LE, Li F, Pencina MJ. Overlap Weighting: A Propensity Score Method That Mimics Attributes of a Randomized Clinical Trial. JAMA 2020;323(23):2417–2418. 10.1001/jama.2020.7819. [DOI] [PubMed] [Google Scholar]
  • 30.Linden A, Adams JL. Applying a propensity score-based weighting model to interrupted time series data: improving causal inference in programme evaluation. J Eval Clin Pract 2011;17(6):1231–1238. 10.1111/j.1365-2753.2010.01504.x. [DOI] [PubMed] [Google Scholar]
  • 31.Belotti F, Deb P, Manning WG, Norton EC. twopm: Two-part models. The Stata Journal. 2015;15(1):3–20. [Google Scholar]
  • 32.Cameron AC, Trivedi PK. Microeconometrics using stata. Vol 2: Stata press; College Station, TX; 2010. [Google Scholar]
  • 33.Bhaskaran K, Gasparrini A, Hajat S, Smeeth L, Armstrong B. Time series regression studies in environmental epidemiology. Int J Epidemiol 2013;42(4):1187–1195. 10.1093/ije/dyt092. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Anderson CE, Au LE, Yepez CE, Ritchie LD, Tsai MM, Whaley SE. Increased WIC Cash Value Benefit is Associated with Greater Amount and Diversity of Redeemed Fruits and Vegetables among Participating Households. Cur Dev Nutr 2023;7(9):101986. 10.1016/j.cdnut.2023.101986. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Whaley SE, Anderson CE, Tsai MM, Yepez CE, Ritchie LD, Au LE. Increased WIC benefits for fruits and vegetables increases food security and satisfaction among California households with young children. J Acad Nutr Diet 2023. 10.1016/j.jand.2023.05.018. [DOI] [PubMed] [Google Scholar]
  • 36.Valizadeh P, Popkin BM, Ng SW. Linking a sugar-sweetened beverage tax with fruit and vegetable subsidies: A simulation analysis of the impact on the poor. Am J Clin Nutr 2022;115(1):244–255. 10.1093/ajcn/nqab330. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Elliott S, Satterfield SJ, Solorzano G, Bowen S, Hardison-Moody A, Williams L. Disenfranchised: How Lower Income Mothers Navigated the Social Safety Net during the COVID-19 Pandemic. Socius 2021;7:23780231211031690. [Google Scholar]
  • 38.Barofsky J, Vargas A, Rodriguez D, Barrows A. Spreading Fear: The Announcement Of The Public Charge Rule Reduced Enrollment In Child Safety-Net Programs: Study examines whether the announced change to the federal public charge rule affected the share of children enrolled in Medicaid, SNAP, and WIC. Health Affairs. 2020;39(10):1752–1761. [DOI] [PubMed] [Google Scholar]
  • 39.Cleaveland C, Lee M, Gewa C. “I thought I was going to die there:” Social-political contexts and the plight of undocumented Latinx in the COVID-19 pandemic. SSM-Qualitative Research in Health. 2023:100242. 10.1016/j.ssmqr.2023.100242. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Appendix

RESOURCES