Embed Size (px)
Citation preview
1
Federal Food Policy and Childhood Obesity:
Part of the Problem or Solution?
Rachel Tolbert Kimbro
Elizabeth Rigby*
*Authors are listed alphabetically. Rachel Tolbert Kimbro
([email protected]) is an assistant professor of sociology
at Rice University and Elizabeth Rigby ([email protected]) is
an assistant professor of political science at the
University of Houston. The authors gratefully acknowledge
a RIDGE grant from Mississippi State University’s Southern
Rural Development Center, a grant from the University of
Kentucky Center for Poverty Research through the U.S.
Department of Health and Human Services, Office of the
Assistant Secretary for Planning and Evaluation, grant
number 2 U01 PE000002-06, and the Faculty Initiatives
Program and Health Economics Program at the James A. Baker
Institute for Public Policy at Rice University. The
Fragile Families and Child Wellbeing Study was funded by
NICHD grant number NIH 5R01HD36916 05 and a consortium of
private foundations. In addition, the authors acknowledge
the generous support of the Robert Wood Johnson Foundation
Health & Society Scholars Program and the excellent
research assistance of Jenny Kingsley and Abby Banks-Golub.
2
Abstract
Amid growing concern about childhood obesity, the U.S.
spends billions of dollars on food assistance—providing
meals and subsidizing food purchases. We examine the
relationship between food assistance and Body Mass Index
(BMI) for low-income, young children: a primary target
population for federal food programs and for efforts to
prevent childhood obesity. Our findings indicate that food
assistance may contribute to the childhood obesity problem
in cities with high food prices. We also find that
subsidized meals at school or daycare are beneficial for
children’s weight status, and argue that expanding access
to subsidized meals may be the most effective tool to
combat childhood obesity for poor children.
3
Introduction
With approximately one-third of U.S. children
overweight and 16% obese (1), childhood overweight is a
significant public health issue, and overweight children
are more likely to develop serious health problems (2-7).
Notably, children are struggling with weight problems at
earlier ages, particularly low-income children (8, 9).
Since overweight children are more likely to become
overweight adolescents and adults (10-12), preventing
obesity early in childhood is the most cost-effective
solution for the epidemic (13, 14).
To combat the “obesity epidemic” among America’s young
children, policymakers are experimenting with a range of
novel policy proposals such as regulating food served in
schools, taxing snack foods, and limiting advertising to
children (15-17). Although these proposals garner much of
the media attention on anti-obesity policy, in reality the
federal government is already deeply engaged in food policy
through its $53 billion annual expenditure in food
assistance programs.
During early childhood, low-income children may
participate in all of the primary federal food assistance
programs. This receipt often begins with the Special
Supplemental Nutrition Program for Women, Infants, and
4
Children (WIC), which primarily provides vouchers for
nutritious foods to pregnant women and children up to age
five who are at nutritional risk. If a child’s family
income is below 130 percent of the federal poverty line
(FPL), the family may receive assistance in the form of an
electronic benefits card which may be used to purchase
virtually any food item as part of the Supplemental
Nutrition Assistance Program (SNAP), formerly known as the
Food Stamp Program. If a child attends a child care,
healthy snacks and meals may be provided through the Child
and Adult Care Food Program. Once they reach school-age,
low-income children may receive meals as part of the School
Breakfast Program or National School Lunch Program.
Since federal food assistance programs, by design,
intervene early in children’s lives and serve the low-
income children who are most likely to be overweight (18),
they hold real potential to combat the obesity epidemic by
reversing early risk factors and setting in place healthy
weight trajectories. The reach of these programs is wide:
one-half of all U.S. children will participate in the
supplemental nutrition program (formerly known as Food
Stamps) alone, including 90% of African-American children
(19). There is also reason, however, to question the
ability of the programs to combat the obesity epidemic
5
(20). These programs were originally designed to encourage
food expenditures at a time when alleviating hunger —
rather than the prevention of obesity — was the primary
policy objective (21). Indeed, mounting evidence suggests
that the Supplemental Nutrition Assistance Program may be
associated with increased Body Mass Index (BMI) for adult
women (22-25).
To help inform this debate, we examine new data on
young children, ages 3 and 5, who comprise a primary target
population for the various federal food assistance
programs. Since children may receive food assistance
through multiple federal programs during early childhood
and since each program may have a unique effect on
children’s weight, we examine the effect of each type of
food assistance, controlling for simultaneous participation
in the other programs. The young children in our sample
are drawn from twenty of America’s larger cities with
varied costs-of-living, giving us contextual variation
within which we can compare effects of program
participation.
Federal Food Assistance Programs and Children’s Weight:
What Do We Know?
Despite the potential for federal food assistance
programs to assist in efforts to combat childhood obesity,
6
research documenting the effects of these programs on
children’s weight is surprisingly limited and has produced
mixed results (26). Hofferth and Curtin (27) found no
relationship between receipt of three food assistance
programs and older children’s BMI. However, other research
points to a relationship between receipt of federal food
assistance and children’s weight status. Among very young
children, Bitler et al. found that — accounting for
characteristics associated with enrolling in the program —
infants and toddlers who participated in the Special
Supplemental Nutrition Program for Women, Infants, and
Children had healthier weight-for-age (28).
Federally-subsidized school meals have the potential
to enhance children’s nutrition, as the meals must adhere
to nutritional requirements. For example, one study that
estimated effects of both the Supplemental Nutrition
Assistance Program and the National School Lunch Program
found that for girls living in food insecure households
(households without access to adequate food), participation
in the programs reduced weight gain (29). One careful
analysis found the School Breakfast Program to have a
positive effect on children’s nutritional quality (30).
Another study found, however, that children who regularly
ate a school lunch had much higher rates of obesity (31),
7
although this may have been due to other characteristics
relating to food preferences (32).
These mixed findings stem from research that treats
receipt of federal food assistance as the same experience
across different contexts. However, even though
eligibility is set at the federal level, the meaning and
experience of these benefits may differ in places where
food, as well as other expenses, is more or less expensive.
This type of contextual effect was identified by Zhang and
Chen (25), who illustrated how local food prices
contributed to body mass index (BMI) for adult women
receiving Supplemental Nutrition Assistance. These findings
provided a powerful incentive for our current analysis.
How Cost-of-Living May Shape the Effects of Federal Food
Assistance
Underlying any relationship between food assistance
and childhood obesity is the possibility that the use,
value, and effects of food assistance vary in places with
high versus low costs-of-living. When cost-of-living is
high, eligible families likely experience greater need than
eligible families in other communities. As a result, a
poverty-line income of $18,310 (for a family of three) can
be stretched much further in a low-cost-of-living city such
as Nashville than in a high-cost-of-living setting such as
8
Boston. This increased difficulty making ends meet may
lead more eligible families to participate in food
assistance programs (33), and may make receipt of food
assistance more critical for promoting children’s well-
being.
One important aspect of cost-of-living is captured in
food prices. When the local cost of food is higher, the
same food assistance benefit simply buys less food (34,
35). This may lead to substitution of less expensive,
unhealthy foods for more expensive, healthier foods (36).
Recent evidence suggests that the diets of low-income
Americans are sensitive to price changes such that lower
prices for fruits and vegetables result in more consumption
by adults (37) and smaller increases in BMI for children
(38, 39).
Expectations
Given the different intentions and delivery mechanisms
of these programs, we distinguish among our expectations
for how each program may shape children’s weight
trajectories. Most important is the degree to which the
program provides food assistance in general, or limits this
assistance to particular types of foods. For example, the
Special Supplemental Nutrition Program for Women, Infants,
and Children specifies the types of food that may be
9
purchased, and includes nutrition education. And the
school lunch and breakfast programs have nutritional
guidelines that must be followed. Therefore, we expect a
more protective effect (a negative relationship with BMI)
for these programs due to their more prescriptive
guidelines regarding the foods that can be purchased for or
provided to young children. In contrast, Supplemental
Nutrition Assistance recipients have more flexibility over
consumption choices; thus, we expect this program to do
less to prevent gains in BMI. For the same reason, we
expect local food prices to play a larger role in shaping
the relationship between this program and children’s BMI
than for the other forms of food assistance.
Data
We examine direct and context-dependent effects of
changes in federal food program participation on young
children’s weight status (BMI) using data from the
longitudinal Fragile Families and Child Wellbeing Survey
(FFCWS), waves III and IV. Participating parents and
infants were recruited in 1998-1999 in twenty large U.S.
cities (with populations of 200,000 or more) and sampled
via a stratified random sampling process. Families were
surveyed for the first time in the hospital within 48 hours
of the child’s birth and followed-up when the child was 1
10
year, 3 years, and 5 years old. At the 3- and 5-year
follow-ups (Waves III and IV), a portion of the sample also
participated in the In-Home Longitudinal Study of Pre-
School Aged Children (for more information about the study,
please visit http://www.fragilefamilies.princeton.edu/).
For this paper, we limited the sample to low-income
(below 185% of FPL) children born to U.S.-born parents who
participated in the In-Home portion of the study at both
age 3 (2001-2002) and age 5 (2003-2004) and for whom we had
valid child and mother BMI measurements at both time points
(N=681). Children’s height and weight data were gathered
by trained researchers using standard techniques. BMI for
the children was calculated and converted to the sex-
specific BMI-for-age percentiles issued by the National
Center for Health Statistics of the Centers for Disease
Control and Prevention (40). Other child-, mother-, and
family-demographic characteristics were gathered via self-
report from mothers at each time point. Descriptive
statistics for this sample are presented in Exhibit 1.
[Exhibit 1 about here]
Since children and families may use multiple food
assistance programs concurrently, we generated both a
general measure indicating whether a child received any
form of food assistance as well as three indicator
11
variables, each identifying receipt of one type of food
assistance: (1) the Special Supplemental Nutrition Program
for Women, Infants, and Children, (2) Supplemental
Nutrition Assistance Program, and (3) Child Care/School
Meals (provided in a child care, preschool, kindergarten,
or summer care setting, funded by federal dollars). In
this sample of low-income children (Exhibit 1), 90%
participated in at least one food assistance program at age
3, rising to 93% at age 5. Considering each type of food
assistance: participation in the Special Supplemental
Nutrition Program for Women, Infants, and Children declined
from age 3 (50%) to age 5 (44%); Supplemental Nutrition
Assistance Program participation rates remained fairly
stable — 71% at age 3 and 72% at age 5; and child
care/school meals, provided to 41% of children at age 3,
nearly doubled at age 5 with 80% receiving federally-
subsidized meals at daycare, preschool, or kindergarten.
To explore differences in cost-of-living, we used 2003
American Chamber of Commerce Researchers Association
(ACCRA) (41) data to classify cities into those with, on
average, high and low grocery prices, defined as those with
ACCRA-generated grocery cost-of-living indices above or
below 103. In our sample of 20 cities, we coded the
following as having relatively-low food prices: Austin,
12
Baltimore, Corpus Christi, Indianapolis, Milwaukee,
Nashville, Norfolk, Pittsburgh, San Antonio, and Toledo.
Those classified as high cost-of-living are: Boston,
Chicago, Detroit, Jacksonville, New York City, Newark,
Oakland, Philadelphia, Richmond, and San Jose.
Analytic Strategy
Because we know that many of the same characteristics
affecting children’s weight also influence whether or not
families are eligible for and use federal food assistance
programs, we had to use a modeling strategy that could
account for these “selection effects.” To do this, we
capitalize on the fact that our data include information on
children, and their families, when they are three-years-
old, as well as five-years-old. We pooled together these
two waves of data collection (681 children observed twice
for a total sample size of 1,362) and included individual-
level fixed-effects to absorb the influence of stable
characteristics of children and families (i.e. race,
parenting capacity) in order to isolate changes in
individual children’s participation in food assistance.
Exhibit 2 presents the dynamics in program
participation across the two time points at ages three and
five. Only 3% of this low-income sample received no form of
food assistance, with most (87%) receiving food assistance
13
at both time points. But, the cases most central to our
analysis are those that switched status during the two time
points (taking up or moving off of food assistance). For
the overall measure of food assistance, only 10% of the
sample changed status. Yet, when we look at the results by
program, larger proportions of the sample changed status
between the two time points: 25% for Special Supplemental
Nutrition Program for Women, Infants, and Children, 24% for
Supplemental Nutrition Assistance Program, and 49% for
childcare/school meals.
[Exhibit 2 about here]
To capture other factors that may have also changed
between age three and age five, we controlled for a range
of dynamic characteristics that could influence either use
of food assistance or children’s BMI. These factors
included the child’s age; whether the child attended
daycare, preschool, or kindergarten; family structure
(married vs. cohabiting or single); maternal education (did
not graduate high school vs. high school degree and/or some
college); maternal employment (does not work outside the
home vs. works part-time or works full-time); the mother’s
BMI; the mother’s BMI squared in case the effect is not
linear; the household’s income-to-poverty ratio; Temporary
Aid to Needy Families receipt; whether the family lives in
14
public housing; and food insecurity (based on the USDA 18-
item scale which measures a household’s access to food)
(42). Our results were not sensitive to alterative
specifications for our control variables.
Results
Exhibit 3 presents the coefficients from individual-
level fixed effects models examining, first, the basic
relationship between transitions into and out of different
food assistance programs and young children’s BMI (top
panel), and then re-estimating these relationships for
those living in cities with low food prices (middle panel)
versus high food prices (bottom panel).
[Exhibit 3 about here]
The first pair of models examines the effect of movement
onto or off of food assistance, while the second pair of
models disaggregates this measure of participation to
capture whether the child (or her family) receives aid from
the Special Supplemental Program for Women, Infants, and
Children and/or Supplemental Nutrition Assistance Program
and/or child care/school meals. We present results, first
for the full sample of low-income children — those living
in families with incomes less than 185% of FPL at both time
points (N=681 children). These children were eligible for
both reduced price lunch and the Special Supplemental
15
program for Women, Infants, and Children. Then we estimate
parallel models for the more disadvantaged sub-set: those
with family incomes at or below 130% of the Federal Poverty
Line who were eligible for free lunch and for Supplemental
Nutrition Assistance Program benefits.
Examining first the relationship between receiving any
food assistance and children’s BMI, we find no significant
relationship across the full sample of 20 cities. Yet, when
we divide these city samples into those with low versus
high food prices, we see a significant relationship for
children with family incomes at or below 130% of Federal
Poverty Line. For these children, we identify a negative
relationship between receiving food assistance and BMI in
cities with low food prices (B = -16.17, se = 7.09) but a
positive relationship in cities with high food prices (B =
15.01, se = 7.14). These findings indicate that food
assistance may contribute to the childhood obesity problem
in cities with high food prices but may serve to protect
children from excessive weight gain in cities with low food
prices.
The second pair of models considers the potential for
different influences for each form of food assistance. For
the models estimated on the full sample across all 20
cities, we see only one significant relationship: a
16
negative impact of receiving child care/school meals on
children’s BMI (B = -4.00, se = 2.00 for those <=185% FPL;
B = -5.10, se = 2.27 for those <=130% of FPL), meaning that
the receipt of meals at school or childcare helps children,
particularly low-income children, maintain a healthy
weight. When we test these relationships in the cities with
low food prices, we see the same pattern for the poorer
sub-set of children (B = -6.31, se = 3.06). Yet, in cities
with higher food prices, we find a positive relationship
between participation in the Supplemental Nutrition
Assistance Program and children’s BMI (B = 7.66, se =
4.44)—suggesting that in areas where food is more
expensive, participation may be contributing to the early
childhood obesity problem.
[Exhibit 4 about here]
These findings are presented visually in Exhibit 4.
The bars labeled “any” capture the expected change in BMI
when children transition onto or off of food assistance
(defined broadly to include any program). The subsequent
sets of bars reflect this relationship for each form of
food assistance. Note that these individual relationships
are smaller because they capture the effect of each
program, controlling for simultaneous receipt of the other
food assistance programs.
17
Conclusion
Our results demonstrate that participation in federal
food assistance programs can affect the BMI of young
children. However, we must distinguish among the multiple
food assistance programs that low-income children may
receive during their early childhood years, as well as the
food environment in which these programs are experienced.
In particular, we find a protective effect from food
assistance programs that subsidize meals as opposed to
those programs that provide financial assistance to
purchase food. This effect is strongest among the low-
income children most at-risk for later weight problems.
This finding suggests that efforts to combat childhood
obesity might be enhanced by increasing access to
subsidized meals through a range of strategies, including:
increasing outreach to child care providers not
participating in the Child and Adult Food Program,
providing school-wide presumptive eligibility for Title I
schools, and instituting summer food programs for school
children and their families. Another anti-obesity strategy
is to increase the nutritional value of food provided by
all the federal food assistance programs, which would
require strengthening nutritional guidelines for subsidized
meals, as well as limiting access to foods of minimal
18
nutritional value as part of the supplemental food
assistance programs.
Yet, our findings also warn of the potential for food
assistance programs to exacerbate the childhood obesity
problem. As mentioned above, this was most likely for the
Supplemental Food Assistance Program—the program that
provides the most broad food choice. In cities with high
food prices — those in which Supplemental Nutrition
Assistance Program benefits provide the least purchasing
power — participation in the Supplemental Nutrition
Assistance Program is related to increases in BMI among the
poorest children. Efforts to prevent childhood obesity need
to take seriously this important role of local context, in
which the same federal program, with the same federal
guidelines and benefit plans, can have a different effect
in some parts of the country compared with others. This
study highlights one such factor: food prices, although we
acknowledge many others that may influence the relationship
between federal food assistance and childhood obesity, such
as: food availability, housing density, and local farmers
markets.
This analysis only begins to unpack the complex
relationships between food assistance and children’s weight
trajectories. We plan to follow these children into the
19
next wave of Fragile Families and Child Wellbeing Study
data (collected when the children are nine years old) in
order to capture the full early childhood period. But, this
initial study highlights the potential for federal food
assistance to serve as part of the early childhood obesity
solution, as well as warns of ways in which federal policy
could undermine other childhood obesity prevention efforts.
Literature Cited
1. Ogden CL, Carroll MD, Flegal KM. High body mass index
for age among US children and adolescents, 2003-2006. JAMA.
2008;299(20):2401.
2. Fagot-Campagna A, Saadinem JB, Flegal KM, Beckles GL.
Emergence Of Type 2 Diabetes Mellitus in children:
epidemiologic evidence. J Pediatr Endocrinol Metab.
2000;13(6):1395-405.
3. To T, Vydykhan TN, Dell S, Tassoudji M, Harris JK. Is
obesity associated with asthma in young children? J
Pediatr. 2004;144(2):162-8.
4. Dietz WH. Health consequences of obesity in youth:
childhood predictors of adult disease. Pediatrics.
1998;101:518-25.
5. Freedman DS, Mei Z, Srinivasan SR, Berenson GS, Dietz
WH. Cardiovascular risk factors and excess adiposity among
20
overweight children and adolescents: the bogalusa heart
study. J Pediatr. 2007;150(1):12-7.
6. Reilly JJ, Methven E, McDowell ZC, Hacking B,
Alexander D, Stewart L, et al. Health consequences of
obesity. Arch Dis Child. 2003;88:748-52.
7. Hardy LR, Harrell JS, Bell RA. Overweight in children:
definitions, measurements, confounding factors, and health
consequences. J Pediatr Nurs. 2004;19(6):376-84.
8. Kimbro RT, Brooks-Gunn J, McLanahan S. Racial and
ethnic differentials in overweight and obesity among 3-
year-old children. Am J Public Health. 2007;97:298-305.
9. Mei Z, Scanlon KS, Grummer-Strawn LM, Freedman DS, Yip
R, Trowbridge FL. Increasing prevalence of overweight among
US low-income preschool children: the centers for disease
control and prevention pediatric nutrition surveillance,
1983 to 1995. Pediatrics. 1998;101(1):103-5.
10. Whitaker RC, Wright JA, Pepe MS, Seidel KD, Dietz WH.
Predicting obesity in young adulthood from childhood and
parental obesity. N Engl J Med. 1997;337:869-73.
11. Guo SS, Chumlea WC. Tracking of body mass index in
children in relation to overweight in adulthood. Am J Clin
Nutr. 1999;70:145S-8S.
21
12. Nader PR, O'Brien M, Houts R, Bradley R, Belsky J,
Crosnoe R, et al. Identifying risk for obesity in early
childhood. Pediatrics. 2006;118(3):e594-601.
13. Daviglus ML. Health care costs in old age are related
to overweight and obesity earlier in life. Health Aff.
2005;24:5.
14. Trasande L, Liu Y, Fryer GE. Trends in u.s.
hospitalizations, charges, and costs associated with
childhood obesity, 1999-2005. Health Aff. 2009;28(4):W751.
15. Kuchler F, Golan E, Variyam JN, Crutchfield SR.
Obesity policy and the law of unintended consequences.
Amber Waves. 2005;3(3):26-33.
16. Nestle M. Food politics: how the food industry
influences nutrition and health. University of California
Press; 2003.
17. Engelhard CL, Arthur Garson J, Dorn S. Reducing
obesity: policy strategies from the tobacco wars. The Urban
Institute. July 2009.
18. Gleason P, Rangarajan A, Olson C. Dietary intake and
dietary attitudes among food stamp participants and other
low-income individuals. USDA, Food and Nutrition Service,
FSP-00-DI; 2000.
22
19. Rank M R, Hirschl T A. Estimating the risk of food
stamp use and impoverishment during childhood. Arch Pediatr
Adolesc Med. 2009; 163(11):994-999.
20. Kennedy E. Public policy in nutrition: the US
nutrition safety net—past, present and future. Food Policy.
1999;24(2-3):325-33.
21. Ver Ploeg M, Ralston K. Food stamps and obesity: what
do we know? In: ERS Report Summary, USDA. 2008.
22. Townsend MS, Peerson J, Love B, Achterberg C, Murphy
SP. Food insecurity is positively related to overweight in
women. J Nutr. 2001;131(6):1738-45.
23. Gibson D. Food stamp program participation is
positively related to obesity in low income women. J Nutr.
2003;133(7):2225-31.
24. Zagorsky J, Smith P. Does the US food stamp program
contribute to adult weight gain? Econ Hum Biol.
2009;7(2):246-58.
25. Zhang Q, Chen Z. Impact of local food prices on the
relationship between food stamp program participation and
bodyweight status of adults in the south. RIDGE Project
Summaries, 2007: Food Assistance and Nutrition Research.
26. Dinour LM, Bergen D, et.al. The food insecurity-
obesity paradox: a review of the literature and the role
23
food stamps may play. J Am Diet Assoc. 2007; 107(11):1952-
1961.
27. Hofferth SL, Curtin S. Poverty, food programs, and
childhood obesity. J Policy Anal Manage. 2005;24(4):703-
72624.
28. Bitler MP, Currie J, Scholz JK. WIC eligibility and
participation. J Hum Resour. 2003:1139-79.
29. Jones S, Jahns L, Laraia B, Haughton B. Lower risk of
overweight in school-aged food insecure girls who
participate in food assistance results from the panel study
of income dynamics child development supplement. Arch
Pediatr Adolesc Med. 2003;157(8):780-4.
30. Bhattacharya J, Currie J, et.al. Breakfast of
champions?: the school breakfast program and the nutrition
of children and families. J Hum Resour. 2006; 41(3):445.
31. Schanzenbach D. do school lunches contribute to
childhood obesity? Harris School Working Paper. 2005.
32. Gordon A, Devaney B, et.al. Dietary effects of the
national school lunch program and the school breakfast
program. Am J Clin Nutr. 1995. 61(1):221S-231.
33. Datar A, Nicosia N. Food prices and transitions in
school meal participation during elementary school. RAND
Working Paper. 2009.
24
34. Golan E, Stewart H, Kuchler F, Dong D. Can low-income
americans afford a healthy diet? Amber Waves. 2008;6(5):26-
33.
35. Kupillas L, Nies M. Obesity and poverty: are food
stamps to blame? Home Health Care Manag Pract.
2007;20(1):41.
36. Drewnowski A, Specter S. Poverty and obesity: the role
of energy density and energy costs. Am J Clin Nutr.
2004;79(1):6-16.
37. Dong D, Lin B-H. Fruit and vegetable consumption by
low-income americans: would a price reduction make a
difference? In: Service ER, editor. USDA; 2009.
38. Sturm R, Datar A. Body mass index in elementary school
children, metropolitan area food prices and food outlet
density. Public Health. 2005;119(12):1059-68.
39. Powell LM, Bao Y. Food prices, access to food outlets
and child weight. Econ Hum Biol. 2009;7(1):64-72.
40. Centers for Disease Control and Prevention. CDC Growth
Charts: United States, 2000.
41. ACCRA Cost of Living Index. The Council for Community
and Economic Research; 2003; Available from:
http://www.coli.org/.
42. Guide to measuring food insecurity. Washington, DC.
USDA. 2000.
25
EXHIBIT LIST
EXHIBIT 1 (table)
Caption/Headline: Descriptive Statistics
SOURCE: Authors’ analysis of FFCWS data.
EXHIBIT 2 (table)
Caption/Headline: Percent of FFCWS Children Receiving
Federal Food Assistance
SOURCE: Authors’ analysis of FFCWS data.
EXHIBIT 3 (table)
Caption/Headline: Relationship Between Food Assistance and
Child BMI
SOURCE: Authors’ analysis of FFCWS data.
EXHIBIT 4 (figure)
Caption/Headline: Relationship between BMI and Receipt of
Food Assistance Program
SOURCE: Authors’ analysis of FFCWS data.
26
Exhibit 1: Descriptive Statistics
Min. Max. 3 yrs 5 yrs
BMI (percentile) 0 99.99 62.08 (30.40) 65.45 (28.57)
Any Food Assistance 0 1 0.90 0.93
WIC 0 1 0.50 0.44
SNAP 0 1 0.71 0.72
CC/School Meals 0 1 0.41 0.80
Child Age 32 76 37.55 (2.99) 63.42 (2.94)
In CC/School 0 1 0.53 0.75
Mom Married 0 1 0.13 0.13
Mom Cohab. 0 1 0.21 0.13
Mom HS Degree 0 1 0.38 0.39
Mom College 0 1 0.20 0.21
Mom Work 2 0 1 0.20 0.20
Mom Work 3 0 1 0.25 0.30
Mom BMI 16 66 30.81 (7.73) 31.60 (8.04)
Poverty Ratio 0 8.8 .73 (.50) 0.75 (.50)
Receive TANF 0 1 0.45 0.40
Live in Public Housing 0 1 0.23 0.10
Food Insecure 0 1 0.24 0.23
Range Mean (sd)
Source: Authors' analysis of FFCWS data. Sample limited to those with
family incomes at or below 185% of poverty line at both time points
(N=681 children)
27
Exhibit 2: Percent of FFCWS Children Receiving Federal Food
Assistance
No
Receipt
Constant
Receipt
Changed
Receipt
No
Receipt
Constant
Receipt
Changed
Receipt
Food Assistance
Any program 3% 87% 10% 2% 91% 7%
WIC 40% 34% 25% 38% 36% 27%
SNAP 16% 59% 24% 12% 67% 20%
Meals 15% 36% 49% 13% 37% 51%
Note: Values may not equal 100% due to rounding.
Source: Authors' analysis using FFCWS data.
Income < = 185% FPL Income < = 130% FPL
28
Exhibit 3: Relationship between Food Assistance and Child BMI
Forms of Assistance
Any Assistance
<=185% <=130% <=185% <=130%
All Cities
Any Program 0.28 0.53
(3.70) (5.02)
WIC -1.06 0.35
(2.31) (2.47)
SNAP 0.35 1.83
(2.42) (2.92)
Meals -4.00 * -5.10 *
(2.00) (2.27)
N (observations) 1362 1104 1362 1104
N (children) 681 552 681 552
Cities w/Low Food Prices
Any Program -0.67 -16.17 *
(5.00) (7.09)
WIC -0.27 -0.50
(3.48) (3.79)
SNAP -1.91 -0.59
(3.22) (3.73)
Meals -4.16 -6.31 *
(2.76) (3.06)
N (observations) 712 556 712 556
N (children) 356 278 356 278
Cities w/High Food Prices
Any Program 0.94 15.01 *
(5.20) (7.14)
WIC -1.63 0.13
(3.21) (3.39)
SNAP 3.13 7.66 +
(3.88) (4.44)
Meals -4.00 -3.24
(3.00) (3.29)
N (observations) 650 548 650 548
N (children) 325 274 325 274
Note: Coefficients from OLS regression models with individual-level fixed effects.
All models also control for childrens's age and child care/school attendance;
Mother's marital status, education, employment status, TANF receipt, and BMI; and
families' income-to-poverty ratio, public housing, and food insecurity
Source: Authors’ analysis using FFCWS data.
29
Exhibit 4: Relationship between BMI and Receipt
of Food Assistance Program
Note: Bars represent coefficients from Exhibit 2 estimated for
children with family incomes <= 130% of FPL, * = p<.05, + = p<.10.
Source: Authors' analysis using FFCWS data.
-20
-10
0
10
20
Any WIC SNAP Meals
All Cities Low Food Prices High Food Prices
