This chapter presents evidence on the detailed components of the benefit formula for the Supplemental Nutrition Assistance Program (SNAP) and examines their impact on the purchasing power of SNAP allotments and the implications for the definition of the allotments’ adequacy (see Box 2-2 in Chapter 2 for a detailed description of the calculation of SNAP allotments). Specific components of the benefit formula examined by the committee include the maximum benefit guarantee, the benefit reduction rate, and various deductions to net income. Additionally, the committee reviewed evidence on such factors as the geographic adjustment of benefits and the timing of benefit updating and receipt that can have either a direct or indirect impact on the benefit formula and thus on allotment adequacy, as well as factors that influence the types of foods purchased with SNAP benefits, including dietary knowledge, preferences, and cultural influences. Factors such as nutrition education and incentives and restrictions on benefit usage, as well as information on retail food outlets, also are considered because they provide a more complete picture of how SNAP benefits are used and the possible implications for the adequacy of SNAP allotments. The chapter ends with a summary of findings and conclusions.
Following certification for participation in SNAP, a monthly allotment is computed based on (1) the maximum SNAP benefit for the household size, (2) the benefit reduction rate, and (3) the household’s or individual’s net
income. The following discussion reviews evidence identified by the committee on components of the SNAP benefit formula and the factors that influence them, and assesses their relationship to the purchasing power of SNAP allotments.
Maximum Benefit Guarantee
Entitlement to SNAP benefits is derived from the cost of the Thrifty Food Plan (TFP) for a family of four. The TFP is based on the cost of purchasing foods consumed by individuals in four age-gender groups. The U.S. Department of Agriculture (USDA) developed four food plans (described in Chapter 2) based on market baskets of food that can provide a diet meeting dietary recommendations for individuals.1 The foods in each market basket are based on current consumption patterns, dietary recommendations, and food composition data and prices. In determining SNAP benefits, the following age-gender groups are used: a male and a female aged 19-50, a child aged 6-8, and a child aged 9-11. In 2006, the market baskets were revised to reflect the Dietary Reference Intakes (IOM, 1997, 1998, 2000, 2001, 2005a,b), the 2005 Dietary Guidelines for Americans (DGA) (USDA and HHS, 2005), the 2005 MyPyramid Food Guidance System (USDA, 2005), and changes in food prices and consumption patterns.
Household Size and the Benefit Level
As noted in Chapter 2, the TFP is designed for a reference family of two adults and two children, and the cost is then adjusted for families of different sizes to reflect economies of scale in food purchases. As described in Box 5-1, relative to the per-person benefit for a family of four, the perperson benefit is increased by 5 percent for a family of three, by 10 percent for a family of two, and by 20 percent for a family of one. Per-person benefits are reduced by 5 percent for families with five or six members and by 10 percent for families with seven or more members. These adjustment factors do not appear to be in line with differential spending patterns for food across families of different sizes, however. According to calculations from the 2010 Consumer Expenditure Survey, per-person expenditures on food are 11 percent higher for families of three than for families of four; families of two and one spend 36 and 57 percent, respectively, more per person than families of four. When restricted to purchases of food consumed at home, the numbers are slightly different and suggest that a more realistic economies-of-scale multiplier would be 44 percent for a one-person family,
1Detailed information on the plans is available at www.Cnpp.usda.gov/Publications/FoodPlans/MiscPubs/TFP2006Report.pdf (accessed March 11, 2013).
Food costs for the Thrifty Food Plan are based on individuals in the context of a reference four-person family. For households that are larger or smaller than the reference, per-person food costs are adjusted for economies of scale using a suggested adjustment such as the following:
• One person—add 20 percent
• Two persons—add 10 percent
• Three persons—add 5 percent
• Five or six persons—subtract 5 percent
• Seven or more persons—subtract 10 percent
SOURCE: CNPP, 2011.
33 percent for a two-person family, and 13 percent for a three-person family. The published tables do not allow for separately calculating multipliers for households of five or larger. These calculations are based on averages for all consumer units and are not restricted to low-income households, whose purchasing patterns may differ from those of other households. In addition, they are based on actual consumption patterns and do not account for differences in nutritional intake or adequacy that may exist across different household sizes. Nonetheless, the evidence reviewed by the committee suggests that the current economies-of-scale multipliers may be substantially underestimated for small households.
Household Composition and the Benefit Level
Recommended nutrient intake varies by individual characteristics such as sex, age, and level of activity. Therefore, the cost of food under the TFP also varies by these characteristics, with lower levels for the elderly and young children. Instead of being adjusted to meet each household’s individual characteristics, the SNAP benefit amount is set for a representative “reference family,” allocating all households of a certain size the same benefit even if their individual characteristics (age, sex, activity level) vary. As discussed in Chapter 2, a 1975 U.S. Circuit Court decision took issue with this assumption and directed USDA to either individualize benefits or set them at a high enough level “so that virtually all recipients are swept
within it.”2 USDA opted for the latter approach, which it operationalized by rewriting the food plans to better account for nutritional guidance and to fit a four-person reference family that included two school-aged children and an adult male and female. The reference family benefit amount was adjusted for different family sizes using economies of scale (Box 5-1).
Based on 2010 household composition data, 30.7 percent of all SNAP participants were school-aged children, 15.9 percent were preschool children, 45.6 percent were nonelderly adults, and 7.9 percent were elderly adults. The SNAP reference family comprises a male and a female aged 19-50, one child aged 9-11, and one child aged 6-8. Using June 2011 data, females aged 19-50 require $156.70 per month in food expenditures, a female aged 12-13 requires $129.00, and a female aged 14-18 requires $157.20 (CNPP, 2011). In contrast, a male aged 19-50 requires $176.00 per month, a male aged 12-13 requires $158.60, and a male aged 14-18 requires $164.50. The current monthly individual food expenses for the reference family are shown in Table 5-1.
The reference family’s food expenditures come to $612.00 per month, and this amount is used to set the maximum benefit, which is then adjusted by the economies-of-scale multipliers to account for different family sizes. “Unusual” household composition will obviously cause variation from this formula. For example, a household of four nonelderly adult males would fall short of meeting the reference family criteria for a maximum-benefit four-person household (by $92 a month), whereas a household with one adult, two preschool-aged children, and one school-aged child would be eligible for the maximum benefit for the household size even though the benefit would exceed the household’s requirement by as much as $129.00 a month.
While it would be possible to issue benefits based on the age and sex of household members at a point in time, any change to the current law would require great care. Households likely to lose the most benefits would be those with a disproportionate number of small children and those with more elderly adults, because they require less food expenditure per month. Those most likely to benefit would be households with disproportionately more adolescents or nonelderly adults, particularly males. The committee was unable to estimate the cost fraction that would increase or decrease the allotment if the estimate were based on individual household composition rather than the reference family, because the data needed to do so were unavailable, and the time and resources required to produce such an estimate were beyond the scope of this study.
2Rodway v. United States Department of Agriculture, 514 F.2d 809, 168 (U.S. App. D.C. 387, 1975).
Monthly Food Expenditure Under the TFP ($)
Male, aged 19-50
Female, aged 19-50
Child, aged 9-11
Child, aged 6-8
NOTES: TFP = Thrifty Food Plan.
SOURCE: CNPP, 2011.
The maximum SNAP benefit varies only by family size in the contiguous United States, but is adjusted upward in both Alaska and Hawaii, presumably because of higher food costs. The presumption, then, is that the variation in prices from the average used in constructing the TFP in the lower 48 states and the District of Columbia is not sufficient to warrant the additional complication of program administration entailed in making similar adjustments. These complications include identifying the appropriate data source and then determining how to apply it meaningfully to households that live on the border of one or another geographic area. The maximum benefit is adjusted each October based on the Consumer Price Indexes (CPIs) for the 29 food categories in the TFP that have a corresponding CPI or set of CPIs for each age-sex group (Carlson et al., 2007). There has been a long-standing assumption that the variation in prices for these 29 categories is not significant across the contiguous states and the District of Columbia. The challenge in questioning this assumption is that the Bureau of Labor Statistics (BLS) does not produce an official CPI for different areas of the country, or one for the TFP. The CPI for All Urban Consumers (CPI-U) spans 87 percent of the population (BLS, 1998), and from this set BLS releases a monthly CPI for the 3 largest metro areas, a bimonthly index for 11 more metropolitan statistical areas (MSAs), and a semiannual index for 12 additional metro areas. However, these subnational price indices do not cover all MSAs or any nonmetro/rural areas. This historic lack of data on regional food prices led the National Academy of Sciences Panel on Poverty and Family Assistance to recommend that cost-of-living differences in the poverty threshold be adjusted only for differences in housing as captured by the U.S. Department of Housing and Urban Development’s Fair Market Rents Index (NRC, 1995). Presently, the approach followed by the Census Bureau in its Supplemental Poverty Measure is to follow the recommendation of the Committee on Poverty Measurement of adjusting the poverty threshold only for differences in housing costs, but using differences
As described in Chapter 4, a series of recent papers from the Economic Research Service has documented substantive regional differences in food prices (Gregory and Coleman-Jensen, 2012; Leibtag, 2007; Todd et al., 2011). Leibtag (2007) shows that, based on Nielsen Homescan data, food prices in the West and Northeast are above average, while those in the South and Midwest are below average, meaning that the SNAP dollar can go further in the South and Midwest than in the West and Northeast. Although low-income consumers adopt coping mechanisms to stretch the SNAP dollar, Leibtag (2007) finds that differences in prices across regions exceed differences in prices paid across demographic (income) groups. Todd and colleagues (2011) provide corroborative evidence that geographic price variation in healthy compared with unhealthy foods may help explain geographic differences in health outcomes. Indeed, Gregory and Coleman-Jensen (2012), using local prices from the Quarterly at Home Food Survey merged with Current Population Survey (CPS) data on food insecurity, find that this regional price variation affects food insecurity—a one standard deviation increase in the cost of a TFP-type basket of goods results in an 8.4 percent increase in adult food insecurity and a 15.9 percent increase in child food insecurity (see Chapter 4).
The committee considered evidence from Children’s HealthWatch because these studies assessed the influence of regional price variations on the purchasing power of SNAP benefits. These studies included a series conducted in Boston and Philadelphia in 2008 and 2011 that examined local costs of purchasing foods consistent with the assumptions of the TFP based on the maximum SNAP benefit (Breen et al., 2011; Thayer et al., 2008). For these studies, the authors assembled grocery lists comprising 107 items from the TFP to feed a two-adult, two-child family (the SNAP reference family). In the 2008 study, four neighborhoods in each city were selected, and within each neighborhood, four stores were selected (two small, one medium, one large). The authors found that families receiving the maximum SNAP benefit needed to spend an additional $2,520 in Boston and $3,165 in Philadelphia per year to purchase foods that meet the TFP guidelines, or roughly 40 to 50 percent more than the maximum annual benefit amount of $6,504 for a four-person family in fiscal year (FY) 2008. This deficit, while varying in magnitude, was present across all four store sizes. The authors also found that 16 and 38 percent of the 107 items were unavailable in the Boston and Philadelphia stores, respectively. In a 2011 follow-up study in Philadelphia, the deficit was lower, but a still substantial $2,352 per year. Although this evidence is limited, the committee did not find additional evidence to support a converse perspective.
In 2009, as part of the national stimulus package, SNAP benefits were increased by 13.6 percent, effective April 2009. Four-person families received a maximum benefit increase of $80 per month (presumably explaining in part the reduced TFP deficit found in 2011 in Philadelphia by Children’s HealthWatch). For a household of three, the maximum benefit increased from $463 to $526 per month. Future increases would be based on 2009, and therefore their impact would be reduced each year once inflation was taken into account (CBO, 2012). The American Recovery and Reinvestment Act of 20093 (ARRA) also allowed states to suspend time limits for unemployed able-bodied adults through FY 2010, increased the minimum benefit from $14 to $16 per month, and increased administrative funding to states. Subsequent legislation set an expiration date of November 2013 for the 13.6 percent benefit adjustment.
USDA found that “the food security of low-income households (those with incomes in the eligible range for SNAP) improved from 2008 to 2009, and a substantial share of that improvement may be due to the increase in SNAP benefits implemented under ARRA” (Nord and Prell, 2011, p. iii). During that period, the SNAP benefit received by the typical low-income household increased by about 5.4 percent (Nord and Prell, 2011). Food security did not increase, however, for households only a little above the SNAP eligibility level. In 2012, the benefit level for a four-person household remains at $668 per month, while the TFP for this category is set at $611.70, resulting in a $56 difference per month.
Regional differences in food prices discussed above, coupled with a number of food access challenges and reduced food insecurity attributed to the ARRA expansion, have led some stakeholders to call for permanent increases in the TFP or for the maximum benefit to be linked to another USDA food plan, such as the Low-Cost Food Plan (Children’s HealthWatch, 2012; FRAC, 2012). The counterargument for permanently adjusting the maximum benefit or linking it to the Low-Cost Food Plan is that to make such a revision cost-neutral, participation would have to be restricted and/or some other aspect of the net income formula (discussed below) would have to be altered to reduce the benefits of those not at the maximum so as to hold total spending in check. Cost neutrality, however, is a requirement linked to the TFP. Moving from the TFP to the Low-Cost Food Plan would necessitate a higher cost that is not supported by the current statute. In the absence of cost neutrality, neither restriction of participation nor reduction of benefits would be necessary, but given that the Low-Cost Food Plan is
3American Recovery and Reinvestment Act of 2009, Public Law 111-5, 111th Congress (February 17, 2009).
about one-quarter more expensive than the TFP, the cost considerations cannot be ignored. This evidence informed the view that determining the adequacy of the TFP as the benchmark for the maximum benefit appears more pressing today given that 40 percent of the SNAP caseload is receiving the maximum benefit (Eslami et al., 2011), suggesting that SNAP is the primary source of food support for a large fraction of the caseload.
Benefit Reduction Rate
As described in Chapter 2, SNAP benefits are calculated as the difference between the maximum benefit guarantee for a given unit size and 30 percent of the unit’s net income (see Box 5-2). In other words, benefits are reduced by 30 cents for each additional dollar of a household’s net income. This benefit reduction rate (BRR) has remained unchanged since the 1977 Food Stamp Act (see Box 5-2).4 The rationale is that benefits are a supplement to households’ food purchases and that participants with incomes should be able to contribute 30 percent of their own cash resources toward food purchases. The 30 percent figure was based in part on an analysis of 1955 USDA consumption data showing that the median family spent one-third of its income on food (Orshansky, 1957). Since not all of a household’s income is counted to determine the SNAP allotment, in practice the formula assumes that recipients can spend 20-25 percent of their total monthly cash income on food (Committee on Ways and Means, 2004; Ziliak, 2008).
Evidence reviewed by the committee suggests that the BRR of 30 percent does not reflect current spending patterns for most U.S. households. In contrast to the findings of Orshansky (1957), the median family in the United States today typically spends a lower share of its income on food than the BRR assumes. According to the Consumer Expenditure Survey (CES), in 2010 the average “consumer unit”5 spent just under 13 percent of its pretax income on food consumed both at home and away (BLS, 2011a). Lower-income consumers typically spend a higher share of their income on food, but even among low-income families, the fraction spent on food is substantially lower today than in 1955. For example, data from the 2010 CES show that consumers with pretax incomes of $5,000 to $9,999 spent 16.8 percent of their income on food, those earning $20,000 to $29,999 spent 13.7 percent, and those earning over $70,000 spent 11.7 percent (BLS, 2011a).
4Reimbursement of Census Enumerators for Telephone Tolls and Charges, Public Law 88-535 (August 31, 1964).
5“Consumer units include families, single persons living alone or sharing a household with others but who are financially independent, or two or more persons living together who share expenses” (BLS, 2011b).
The SNAP benefit reduction rate is the rate at which the maximum benefit is reduced per dollar of income. The current benefit reduction rate is 30 percent and is based on the assumption that an average household will spend 30 percent of its net income on food. Thus, for each additional dollar of net income, the maximum SNAP benefit is reduced by 30 cents. The minimum benefit after all income-related reductions for one- and two-person households in the contiguous United States in 2009 was $16 per month.
SOURCES: FNS, 2012e,f.
The committee identified important design trade-offs involved in setting the BRR that can influence the amount of the SNAP benefit a participant receives. A high BRR keeps program costs lower and directs more of the benefits toward recipients with the lowest incomes. Holding other factors constant, a high BRR keeps program costs lower because it reduces the benefit at a faster rate as labor and other taxable income increases. A higher BRR also keeps program costs lower because fewer people are eligible. That is, using the notation of Box 2-2 in Chapter 2, the “break-even” income level (Y(b)) for eligibility can be defined as Y(b) = G/BRR + D, where G is the maximum benefit, BRR is the benefit reduction rate, and D is deductions and exemptions used in constructing net income Y(n). Holding G and D fixed, a higher BRR results in a lower Y(b) and thus fewer people eligible.
On the other hand, a higher BRR also poses a disincentive for recipients to work because their benefits will be reduced at a relatively high rate for each additional dollar earned. Although evidence on the work disincentive effect of the BRR generally suggests the effect is small or modest (Fraker and Moffitt, 1988; Hagstrom, 1996; Hoynes and Schanzenbach, 2012), the BRRs accumulate across programs, leading to potentially large aggregate work disincentive effects when SNAP benefits are received in conjunction with other transfers, such as Temporary Assistance for Needy Families (TANF), housing, and the earned income tax credit (Keane and Moffitt, 1998). Of importance, a lower BRR preserves the incentive to work for participants who are not near the eligibility threshold, but for participants who are close to the eligibility threshold, earning more may make them ineligible for the program. In the extreme case in which a recipient is exactly on the margin of eligibility and receives the minimum SNAP benefit of $16 per month, earning $1 would have the net impact of losing $15. In some
The effective tax rate on earnings is somewhat complicated because two of the deductions used to compute net income are themselves functions of income. The earned income deduction of 20 percent reduces the effective tax rate on benefits. The excess shelter cost deduction (see page 157) is calculated as the amount of shelter costs over 50 percent of net income after other deductions are taken, so an increase in income can reduce this deduction. As a result, an increase in income can result in a benefit reduction that is greater than the base (Ohls and Beebout, 1993).
Net Income Determination
Earned Income Deduction
An important change within the SNAP population is that an increasing proportion of the SNAP caseload is employed (Eslami et al., 2011). This shift toward a greater number of employed participants can have an impact on the purchasing power of the SNAP allotment because of expenses related to employment, such as transportation to work and child care expenses, which reduce the disposable resources available to purchase food. To account for the cost of being employed, the SNAP formula allows certain deductions in the calculation of a household’s net income on which the benefit level is based. Twenty percent of earned income is deducted, and recipients can deduct their spending on dependent care (prior to 2008 the dependent care deduction was capped at $175 per month per dependent) (CBPP, 2010).
There has, however been less recognition that being employed reduces the time available to prepare meals (Davis and You, 2010; Rose, 2007). As discussed in previous chapters, the cost of the TFP does not take into account time costs for food procurement and meal preparation, and therefore does not explicitly account for the trade-off between the costs of more expensive, intermediate-prepared foods and the labor costs of preparation. For example, a household may prefer to purchase prepared foods (e.g., precut carrots or shredded lettuce) instead of spending the time to prepare meals from raw ingredients. Given this trade-off, the earned income deduction at its current level may reduce the overall purchasing power of the SNAP allotment, especially for those facing time constraints such as households headed by a working single mother. Employment among single mothers accelerated with the reforms of the 1990s toward a more work-based safety net, notably the expansions of the earned income tax credit that increased the reward for working and the 1996 Welfare Reform Act,6
6Personal Responsibility and Work Opportunity Reconciliation Act of 1996, Public Law 104-193 (August 22, 1996).
Changes in participation rates for some subgroups in the SNAP population may be attributable to a combination of effects. For example, changes in the economy, in program rules, in the availability of other public assistance programs, and in the participation decisions of eligible individuals all contribute to fluctuations in SNAP participation. Participation by children, individuals in households with earnings, small households, nondisabled childless adults subject to work requirements, and noncitizens all increased in FY 2008 and 2009 (Leftin et al., 2011). At the same time, participation by the elderly and by individuals in households earning at about the poverty threshold remained relatively unchanged.
Recent evidence shows that the shelter deduction, which consists of expenses such as rent, mortgage payments, and utilities,7 is claimed by more than 70 percent of all households, and more than 28 percent of these households have housing expenses that exceed the SNAP shelter cap (Eslami et al., 2011). The actual amount deducted from income is that portion of a household’s shelter costs that exceeds 50 percent of its income after all other deductions. However, the shelter deduction may not exceed $459 in 2012. As mentioned in Chapter 2, the shelter deduction cap is adjusted every fiscal year to reflect changes using the CPI-U for the previous 12 months ending November 30.8 Households with elderly or disabled members are not subject to the cap.
In a study carried out in 2002, the Center on Budget and Policy Priorities found that in the Northeast, Midwest, South, and West, 57, 53, 47, and 57 percent, respectively, of households had shelter costs exceeding 50 percent of their income (Rosenbaum et al., 2002). However, the study also found that the substantial differences in the amount households pay for their housing “is not a geographical phenomenon” and that variation in housing costs paid by SNAP-eligible households exists within all regions of the country. This finding was based on quality control data from USDA’s Food and Nutrition Service (FNS) as well as from the 1999 American Housing Survey. The American Housing Survey was updated in 2007; however, the committee is not aware of updates to this study. Because geographic variation is so great within rather than among regions and states, the shelter deduction and the other individualized deductions are one way to account in part for geographic price differences. Thus, the question arises
7Households can claim actual utility costs or use a standard allowance, which varies by state.
8Agriculture, Rural Development, Food and Drug Administration, and Related Agencies Appropriation Act of 2001, Public Law 106-387, Sec. 846 (October 28, 2000).
All households receive a standard deduction from gross income that is intended to account for unusual or unexpected household expenses that could limit food purchasing power. The deduction varies by household size and is adjusted annually. The deduction is set at 8.31 percent of the income eligibility standard, not to exceed 8.31 percent for a family of six. The 2012 standard deductions are $147 for households of one to three members, $155 for households of four, $181 for households of five, and $208 for households of six or more. Wilde (2002) estimated that a $1.00 increase in the standard deduction raises SNAP benefits by $0.30 to $0.45 for households with positive net cash income. This variation occurs because of an interaction between the standard deduction and the excess shelter deduction. That is, a $1.00 increase in the standard deduction raises benefits by $0.30 for those without an excess shelter deduction, but raises them by $0.45 for those who also have the shelter deduction but are below the shelter cap.
Geographic Adjustment of SNAP Benefits
In addition to the adjustment to the maximum benefit for residents of Alaska and Hawaii, several aspects of the current SNAP benefit formula directly or indirectly accommodate differences in cost of living across regions of the country. This has the effect of either lifting some to the maximum benefit (because the deductions lower net income to zero) or raising the monthly benefit payment. This geographic adjustment is accomplished directly by the excess shelter deduction, which, as described previously, permits the deduction of housing and housing-related costs above 50 percent of net income after other deductions. Because housing costs vary widely across the nation, this deduction accommodates to some extent the geographic variation in cost of living. In FY 2012, however, this deduction was capped at $459 per month, and nearly 30 percent of recipients have housing costs in excess of this cap, suggesting that the cap is a binding constraint for many SNAP households.
There are two major deductions available to working SNAP recipients that implicitly introduce geographic differences in SNAP benefits. The first is the 20 percent deduction of earnings from gross income. Wages vary greatly across the country because of differences in local labor markets (Moretti, 2011); moreover, wages for the same job in the same company but in different locations vary greatly both within and across countries (Ashenfelter, 2012). Thus two SNAP recipients working full-time as cashiers
at the same major fast-food chain—for example, one in New York City and the other in Kansas City—would have different levels of deductions for the purpose of net income: net income would be lower and the SNAP benefit higher for the recipient in New York City than for the one in Kansas City. The second deduction for working SNAP recipients (and for those seeking work or students/trainees) is the dependent care deduction. Prior to the Farm Bill of 2008,9 the dependent care deduction was capped at $175 per month, but since 2008 it has not been capped. This means that actual costs of dependent care, such as the direct cost of care, transportation to and from care, copayments for subsidized care, unreimbursed payments for care, and fees for unused care, may be deducted in the net income calculation (CBPP, 2010). Because the cost of child care varies across states (NACCRRA, 2011), the amount of the dependent care deduction will vary accordingly.
The committee identified three additional implicit geographic cost-of-living adjustments in the benefit formula that can have an effect on the SNAP allotment. First, the elderly and disabled can deduct their monthly out-of-pocket medical expenses in excess of $35 from net income. Given that regional differences in medical spending are substantial (CBO, 2008; Fisher et al., 2009), this introduces geographic cost-of-living differences into the benefit formula. A second regionally focused deduction comes from the child support payment allowance. Pirog and colleagues (1998) document cross-state differences in child support awards, and this, too, may introduce geographic variation in SNAP benefits. The third adjustment works in the opposite direction from the others by reducing the size of the SNAP allotment. Income from other transfer programs, such as TANF, reduces the size of the SNAP benefit, and since the TANF benefit is set at the state level and tends to be higher in high-cost states, this has the effect of “taxing” the SNAP allotment in high-cost areas since it is set nationally at a fixed level for the lower 48 states.
Because most of the geographic differences in cost of living in the SNAP benefit formula are implicit rather than explicit, the question arises of whether making the adjustment more direct would facilitate definition of the benefit’s adequacy. For example, the evidence of regional differences in prices across the lower 48 states (recall that Alaska and Hawaii already have upward adjustments to benefits) suggests that in lieu of moving from the TFP to the Low-Cost Food Plan as the baseline for the maximum benefit, one could instead index the benefit for differences in the cost of living. That is, the indexed benefit for location i in time period t (Bit) would be the product of the price index (Pit) and the federal maximum benefit guarantee (G), or Bit = Pit × G. Normalizing the average price index to 1, Pit would be
9Food, Conservation, and Energy Act of 2008, Public Law 110-234 (May 22, 2008).
<1 for low-cost locations and >1 for high-cost locations. This implies that the maximum benefit could actually fall for many areas, which is not allowed under current law and arguably may not be efficient in terms of meeting the program goals of improving food security and access to a healthy diet, given the evidence that higher benefits improve these outcomes. This suggests an alternative of restricting Pit to ≥1, which means that for average or low-cost areas, benefits would be linked to the TFP as is current practice, but those living in high-cost areas would receive an upward adjustment. Presumably this approach would address some of the food benefit gap as identified by Children’s HealthWatch in Boston and Philadelphia (Breen et al., 2011), and likewise in similar high-cost locations. Conversely, this asymmetric adjustment would lead to increased program costs.
The challenge of implementing geographic cost-of-living adjustments is that at present, BLS does not produce a regional price index. As stated by BLS, the CPI for the four major census regions (Northeast, South, Midwest, West), along with that for the 27 major MSAs, “measures how much prices have changed over a specific period in that particular area; it does not show whether prices or living costs are higher or lower in that area relative to another. In general, the composition of the market basket and the relative prices of goods and services in the market basket during the expenditure base period vary substantially across areas” (BLS, 2011c, FAQ 14). Researchers at the Bureau of Economic Analysis are conducting ongoing research into the production of a regional price index (Aten et al., 2012), while those at the Census Bureau involved in poverty measurement are adjusting poverty thresholds only for differences in spending on housing (Renwick, 2011; Short, 2012). In the short term, adjusting the maximum benefit geographically for differences in cost of living (or even food) is likely to be infeasible until further progress is made on regional price indices.
Timing of Benefits
SNAP benefits are deposited onto an Electronic Benefit Transfer (EBT) card near the beginning of each month. USDA research shows that about 80 percent of benefits typically are used up within the first 14 days, and by the 21st day of the month, more than 91 percent has been spent (FNS, 2012a). Because families that run out of benefits usually do so at the end of the month, it has been suggested that benefits be issued semimonthly to level out spending. Benefits were in fact issued semimonthly at one time, but that was when there was a purchase requirement (see Chapter 2). Those who use food pantries and other private food assistance to supplement their SNAP benefits might be expected to change the timing of their usage to reflect a semimonthly cycle. Whether this would be advantageous to food providers
The committee considered a number of factors not directly related to the SNAP allocation that influence the type of foods purchased with SNAP benefits. These factors—incentives and restrictions on benefit usage, eligibility rules for retail outlets, and nutrition education—are examined here only in the context of how they might influence the feasibility of defining the adequacy of SNAP allotments consistent with the goals of increasing food security and access to a healthy diet. However, the committee recognizes that these factors are not directly linked to defining the adequacy of SNAP allotments. Further, in carrying out its charge, the committee was asked not to consider revisions to the TFP. As a consequence, the committee derived no conclusions or recommendations from the following discussion that would directly alter the TFP.
SNAP places few limits on the use of benefits. As discussed in Chapter 2, eligible foods include any food or food product for home consumption, as well as seeds and plants (FNS, 2012b), but SNAP benefits may not be used for the purchase of hot foods or any food sold for on-premises consumption. Nonfood items, such as tobacco products, pet foods, soaps, paper products, medicines and vitamins, household supplies, grooming items, and cosmetics, also are ineligible for purchase with SNAP benefits.
Several times in the history of SNAP, Congress has considered placing limits on the types of food that can be purchased (FNS, 2012b). However, it was concluded that designating foods as luxury or non-nutritious would be administratively costly and burdensome. In addition to Congress, cities and states have expressed interest in limiting the use of SNAP benefits to purchase certain foods and beverages (Barnhill, 2011). Because the criteria for SNAP purchases are federally regulated policies, however, any state that wishes to impose its own restrictions must apply to USDA for a waiver. To date, USDA has not approved any applications for waivers.
The discussion below illustrates the complexity of the issue of potentially restricting purchases made with program benefits. Potential impacts on participants’ dietary intake and nutritional status must be weighed carefully against concerns about program administrative complexity and program access, as well as participants’ freedom to make their own purchasing
Studying the impact of SNAP on dietary quality presents a challenge because of selection bias; SNAP participants often are worse off than eligible nonparticipants with respect to financial and nutritional needs (Martin et al., 2012; Nord and Golla, 2009). Because SNAP participants and nonparticipants may not be sufficiently comparable, it is difficult to determine whether observed differences are due to the program or to unobserved differences between the groups, such as a family’s economic situation, nutritional needs, health status, food security status, or motivation to enroll in the program. Evidence reviewed by the committee related to the association between SNAP and diet quality was inconclusive. Some studies suggest that SNAP improves the quality of the diet (Basiotis et al., 1998; Lee et al., 2006; Mabli et al., 2010; Rose et al., 1998), while others suggest that it does not (Chen et al., 2005; Cole and Fox, 2008; Fox et al., 2004; Wilde et al., 1999; You et al., 2009). Because of the confounding effect of a self-selection bias among the SNAP population, using analysis of diet quality to inform a definition of the adequacy of SNAP allotments may not be feasible.
Most recently, debates over restrictions of SNAP purchases have focused largely on whether the purchase of sugar-sweetened beverages (SSBs) should be permitted (Brownell and Ludwig, 2011). Evidence is limited on patterns of beverage purchases among SNAP recipients. Andreyeva and colleagues (2012) collected grocery store scanner data from January to June 2011 and concluded that SSBs account for 58 percent of beverage purchases made by SNAP households. Their data and analysis are limited, however, by a regional focus on a single grocery chain in New England and the inclusion of only SNAP households with a history of recent participation in the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC). Further, purchases of SSBs accounted for 5 and 7 percent of total grocery spending by SNAP and WIC households, respectively, and 9 and 7 percent, respectively, of total spending on all beverage refreshment categories (Andreyeva et al., 2012). Many arguments have been made for and against the policy of restricting SNAP purchases of SSBs (Brownell and Ludwig, 2011; FNS, 2007). The following discussion summarizes only those related to access to a healthy diet (e.g., whether allowing for the purchase of SSBs with SNAP benefits contributes directly to an unhealthy diet).
Economic theory suggests that restricting the purchase of some items, such as SSBs, might not impact the overall purchasing behavior of SNAP recipients. SNAP was designed to supplement participants’ food purchases,
and most participants purchase food for home consumption with resources both from SNAP benefits and from other cash income (e.g., from employment and Social Security) (Fraker et al., 1995). In the hypothetical case that SNAP program rules were changed to prohibit the purchase of SSBs with SNAP benefits, theory predicts that participants would continue to purchase the same amount of SSBs as long as spending on these products was less than the amount of cash they typically spent on overall food purchases. Given that restricting the purchase of SSBs would be unlikely to change household purchases, proponents of the change argue that taxpayer dollars should not be used to purchase SSBs, while opponents argue that the administrative burden of restricting the purchase of SSBs would be too costly, and that such restrictions are paternalistic and could further stigmatize SNAP recipients (Brownell et al., 2009; Shenkin and Jacobson, 2010).
There are some important exceptions to this prediction about purchasing behavior. If a participant typically spent more on SSBs than the amount of cash typically used to supplement the household’s food purchases, theory predicts that the household would reduce purchases of SSBs in response to such a restriction. Furthermore, items purchased with SNAP benefits are not subject to sales tax. SSBs are subject to sales tax in 33 states (mean tax rate = 5.2 percent) (Brownell et al., 2009). As a result, restricting SSB purchases with SNAP benefits would in some states increase the price for SSBs faced by SNAP households by making them liable for sales and excise taxes (McGranahan and Schanzenbach, 2011). Although there is little evidence on what effect this price increase would have on the consumption behavior of SNAP participants, Fletcher and colleagues (2010) found that current state soda taxes reduce adolescents’ consumption modestly. Several other studies have demonstrated that a 10 percent increase in the price of SSBs could reduce consumption by 8 to 11 percent on average (Andreyeva et al., 2011; Bahl et al., 2003; Bergtold et al., 2004; Yen et al., 2004).
An extensive discussion of the literature related to SSB consumption and obesity risk is beyond the scope of this report. Of particular relevance to this report, however, are studies that specifically examine whether the purchase of SSBs is associated with poor diet quality; nonetheless, the committee was unable to identify research on this topic (IOM, 2012).
Hot Prepared Foods
Historically, SNAP has restricted the purchase of hot prepared foods in eligible food outlets because the program was designed to supplement purchases of food for home preparation and consumption. Challenges entailed in preparing food at home, however, have led some areas to start their own restaurant programs with FNS approval, such as Rhode Island’s Food Access Prepared Meals Pilot Program (RIDHS, 2012) and California’s
Restaurant Meals Program (Hodges and Emerson, 2012). The aim of these programs is to offer elderly, homeless, and disabled SNAP participants the opportunity to use their benefits for hot prepared meals at approved restaurants. California is one of only eight states taking advantage of the long-standing option to authorize certain restaurants to accept SNAP benefits for the elderly and disabled. It has certified 1,081 establishments, compared with 106 in Arizona and 47 in Michigan. No other state has more than 9 certified (FNS, 2012c). Other allowable meal services include drug and alcohol treatment centers, communal dining facilities for the elderly, and homeless meal providers (see Chapter 2 for further background on the restrictions on hot foods).
Use of Incentives to Promote a Healthy Diet for SNAP Participants
In addition to restriction or expansion of SNAP benefits, incentives offer another mechanism for encouraging the purchase of healthy foods. Evidence suggests that the use of financial incentives to promote health behavior change is effective (Kane et al., 2004; Volpp et al., 2009a,b). Incentives can be framed as rewards or as penalties. The behavioral economics literature suggests that financial incentives framed as rewards may have smaller effects than penalties of equivalent size (Arrow, 2004) because of “loss aversion” (Conrad and Perry, 2009, p. 359). For healthy behavior changes, however, such as dieting or smoking cessation, rewards have been shown to motivate behavior change effectively (Volpp et al., 2008, 2009a,b). While penalty incentives are widely used for behavior change, moreover, there is a lack of evidence directly comparing positive and negative incentives (Volpp et al., 2009a). The Farm Bill of 200810 authorized $20 million for pilot projects (e.g., Healthy Incentives Pilot [HIP]) to evaluate health and nutrition promotion in the SNAP program and to determine whether financial incentives provided to SNAP recipients at the point of sale increase the purchase of fruits and vegetables or other healthful foods (FNS, 2012d). The evaluation data for HIP were not available as this report was being written.
Retail Food Outlets
More than 231,000 retail outlets accepted SNAP benefits by the end of FY 2011, including a small number of restaurants that served the elderly, disabled, and homeless. One of the most dramatic changes over the years has been the participation by farmers’ markets. In FY 2010, 6,132 farmers
10Food, Conservation, and Energy Act of 2008, Public Law 110-234, Sec. 4141 (May 22, 2008).
markets were operating, and 1,611 of these markets and individual farmers were authorized to accept SNAP benefits totaling $7,547,028. The number of markets and farmers increased by 263 percent over FY 2009, and redemptions increased by 49 percent over the previous 5 fiscal years (FNS, 2011a). Overall, 83 percent of all benefits in FY 2010 were redeemed by supermarkets or super stores, 6 percent by grocery stores, and 4 percent by convenience stores. Among food outlets where SNAP benefits are redeemed, however, only 17 percent are supermarkets or super stores, about 15 percent are grocery stores, 36 percent are convenience stores, 23 percent are combination stores, 2 percent are meal services, and 7 percent represent all other stores (FNS, 2012a).
As discussed in Chapter 1, to be authorized to accept SNAP benefits, a store must sell food for home preparation and offer for sale on a continuous basis a variety of food items that include meat, fish or poultry, breads or cereals, vegetables or fruits, and dairy products, with perishables (including frozen foods) in at least two of these groups. If a store does not meet this definition, it may be authorized if at least 50 percent of its total sales volume is in staple food sales.
USDA has been working to increase the number of farmers’ markets that accept SNAP benefits and recently announced grants to expand wireless technology. Currently, markets receive free EBT point-of-sale devices only when redemptions are $100 or more per month. The $4 million in grants is the result of funding provided through the 2012 Consolidated and Further Continuing Appropriations Act.11 These grants will help markets that lack access to phone lines or electricity. It should be noted that the committee acknowledges the concerns of feeding programs for the elderly about their problems with accepting SNAP donations in the EBT environment.12 The difficulty of determining which outlets should be eligible to redeem benefits lies in the need to consider issues of access, pricing, quality, variety, and business integrity. This issue continues to attract attention by the program’s administrators, client advocates, the retail food associations, and Congress.
Providing nutrition education to SNAP participants through SNAP-Education (SNAP-Ed) is not a program requirement. Nutrition education funding is available to states that opt to provide nutrition education to their SNAP participants. This component of the SNAP program has grown considerably
11Consolidated and Further Continuing Appropriations Act, 2012, Public Law 112-55 (November 18, 2011).
12Personal communication, Enid Borden, Meals On Wheels, March 28, 2012.
in the last two decades. In 1992 only seven states had approved nutrition education plans, and the federal share of funding was $661,000. By 2011, all states and the District of Columbia had approved plans, and the federal share of funding was $372 million (FNS, 2011b). However, the Healthy, Hunger-Free Kids Act of 2010 placed a cap on federal funding for SNAP-Ed of $375 million in FY 2011 and then indexed funding to inflation in future years.
As part of its examination of the evidence, the committee discussed the role of SNAP in providing nutrition education. Three alternative scenarios were highlighted in this discussion:
• SNAP should offer nutrition education because it serves one in seven Americans and therefore has an opportunity to impact national nutrition and health.
• Because SNAP participants have many of the same dietary problems experienced by the population as a whole, nutrition education should be undertaken equally for all Americans and funded accordingly (i.e., SNAP funds should not be diverted to nutrition education).
• The low-income population, as represented by SNAP participants, has special challenges and burdens that should be addressed through unique nutrition education approaches funded by the SNAP program.
SNAP nutrition education programs need more and better evaluation, including studies investigating optimal approaches to delivering educational messages. The committee did consider the role of nutrition education in the food purchasing decisions made by SNAP participants to better inform its assessment of the feasibility of defining the adequacy of SNAP allotments.
The evidence presented in this chapter highlights a number of challenges related to the calculation of SNAP benefits that have an impact on defining their adequacy. The committee’s findings and conclusions based on this evidence focus on the maximum benefit guarantee, the BRR, and the net income calculation.
Maximum Benefit Guarantee
The TFP does not account for the time costs of food acquisition and preparation or for geographic variation in the cost of food. Limited evidence from community-level studies indicates that some SNAP households
with zero net income residing in high-cost locales with limited food access are unable to purchase foods within the cost and food choice assumptions of the TFP. The costs of foods that are value-added and have some builtin preparation time are not accounted for in the maximum benefit. The committee found compelling evidence on geographic price differences and time costs of food. Less compelling, however, is the evidence on how to incorporate these factors into the SNAP benefit formula, particularly for the maximum benefit. Moreover, because 80 percent of SNAP benefits are redeemed in supermarkets, the national prevalence of challenges similar to those identified in the community studies is unclear.
The committee concludes that specific areas of research could fill the evidence gap. These research areas include ways to incorporate time costs into the TFP; geographic price adjustments to the maximum benefit; and the effectiveness of alternative food plans, such as the Low-Cost Food Plan, in helping to achieve the program goals in areas where pricing variation negatively impacts the adequacy of SNAP allotments.
Benefit Reduction Rate
The committee’s review of the evidence led to the finding that the five-decades-old assumption that the average household spends 30 percent of its income on food purchases is inconsistent with current spending patterns of American families, regardless of income. Today the average family spends about 13 percent of its income on food, and the current SNAP benefit formula is not aligned with this change.
From the evidence reviewed, the committee concluded that a BRR more in line with current spending patterns would result in increased incentive for households to combine work with SNAP participation because a lower BRR would reduce the penalty due to working. Holding other factors constant, moreover, a lower BRR would be expected to increase the SNAP allotment for those with positive net income, thereby enhancing the opportunity of these households to achieve improved food security and access to a healthy diet.
Calculation of Net Income
Evidence reviewed by the committee suggests that a substantial proportion of SNAP households face very high housing costs and that the cap on the excess shelter deduction is binding for nearly 30 percent of these households. Evidence is limited, however, on the extent to which the earned income deduction has an impact on the adequacy of SNAP allotments. As noted, the TFP does not incorporate the time costs of food preparation, and this is a concern in particular for households headed by a working single
parent, who faces significant time pressures as a result of his or her employment status. This pressure could be relieved somewhat by an earned income deduction that gave employed recipients a larger benefit that could be used to purchase more partially prepared foods, which in turn could shorten meal preparation time. At the same time, out-of-pocket expenses on transportation and clothing for work typically are higher for the employed. It is unclear whether the 20 percent earned income deduction is adequate to address all of these additional expenses.
Likewise, the medical deduction is allowed only for limited populations of the elderly and disabled, for out-of-pocket medical expenses. In light of the rising cost of health care and the increasing percentage of the nonelderly population with chronic diseases, coupled with reductions in employer-provided insurance and uncertainties associated with implementation of the Patient Protection and Affordable Care Act of 2010,13 the impact of the burden of out-of-pocket medical costs on the purchasing power of SNAP allotments for the nonelderly and nondisabled is unknown.
The committee drew two conclusions from these findings. First, raising the shelter deduction cap to reflect geographic differences in housing more accurately would likely decrease the net income of SNAP households and thereby increase the amount of the allotment available for food purchases. Second, further evidence is needed on the effectiveness of the current earned income deduction in addressing the time costs of food preparation for working SNAP participants, as well as on whether the deduction for out-of-pocket medical expenses should be extended to all SNAP units regardless of age and disability status.
The currently available secondary and administrative data infrastructure is likely inadequate to address many of the research needs identified above. Some will require multisite, multiyear demonstration projects, coupled with rigorous evaluation, to obtain the necessary data, while others will require new survey data, especially on the development of a regional price index to provide a better understanding of geographic differences in the cost of foods.
Andreyeva, T., F. J. Chaloupka, and K. D. Brownell. 2011. Estimating the potential of taxes on sugar-sweetened beverages to reduce consumption and generate revenue. Preventive Medicine 52(6):413-416.
Andreyeva, T. J. Leudicke, K. E. Henderson, and A. S. Tripp. 2012. Refreshment beverage choices of SNAP and WIC participants: Analysis of grocery store scanner data. American Journal of Preventive Medicine 43(4):411-418.
13The Patient Protection and Affordable Care Act, Public Law 111-148 (March 23, 2010).
Ashenfelter, O. 2012. Comparing real wage rates. American Economic Review 102(2):617-642.
Aten, B. H., E. B. Figueroa, and T. M. Martin. 2011. Research spotlight: Regional price parities by expenditure class, 2005-2009. Survey of Current Business 91(5):73-87.
Bahl, R., R. Bird, and M. B. Walker. 2003. The uneasy case against discriminatory excise taxation: Soft drink taxes in Ireland. Public Finance Review 31(5):510-533.
Barnhill, A. 2011. Impact and ethics of excluding sweetened beverages from the SNAP program. American Journal of Public Health 101(11):2037-2043.
Basiotis, P. P., C. S. Kramer-LeBlanc, and E. T. Kennedy. 1998. Maintaining nutrition security and diet quality: The role of the Food Stamp Program and WIC. Family Economics and Nutrition Review 11(1-2):4-16.
Bergtold, J., E. Akobundu, and E. B. Peterson. 2004. The FAST method: Estimating unconditional demand elasticities for processed foods in the presence of fixed effects. Journal of Agricultural and Resource Economics 29(2):276-295.
Bhargava, A., and A. Amialchuk. 2007. Added sugars displaced the use of vital nutrients in the National Food Stamp Program Survey. Journal of Nutrition 137(2):453-460.
Blank, R. M. 2002. Evaluating welfare reform in the United States. Journal of Economic Literature 40(4):1105-1166.
BLS (Bureau of Labor Statistics). 1998. U.S. Department of Labor program highlights: Guide to available CPI data. BLS Fact Sheet 94-1 (revised). http://www.bls.gov/cpi/cpifact8.pdf (accessed August 13, 2012).
BLS. 2011a. Table 46. Income before taxes: Shares of average annual expenditures and sources of income, Consumer Expenditure Survey, 2010. ftp://ftp.bls.gov/pub/special.requests/ce/share/2010/income.txt (accessed August 14, 2012).
BLS. 2011b. Consumer spending in 2010. Focus on Prices and Spending: Consumer Expenditure Survey, 2010 2(12), http://www.bls.gov/opub/focus/volume2_number12/cex_2_12.htm (accessed August 14, 2012).
BLS. 2011c. Consumer Price Index frequently asked questions (FAQs). http://www.bls.gov/cpi/cpifaq.htm#Question_14 (accessed August 13, 2012).
BLS. 2012. U.S. Department of Labor program highlights: Guide to available CPI data, BLS Fact Sheet 94-1 (revised). http://www.bls.gov/cpi/cpifact8.pdf (accessed August 13, 2012).
Breen, A. B., R. Cahill, S. E. De Cuba, J. Cook, and M. Chilton. 2011. The real cost of a healthy diet: 2011. Philadelphia, PA: Center for Hunger-Free Communities, Children’s HealthWatch, Drexel University School of Public Health. http://www.centerforhungerfreecommunities.org/sites/default/files/pdfs/RCOHD_Report2011-FINAL.pdf (accessed August 13, 2012).
Brownell, K. D., and D. S. Ludwig. 2011. The Supplemental Nutrition Assistance Program, soda, and USDA policy: Who benefits? Journal of the American Medical Association 306(12):1370-1371.
Brownell, K. D., T. Farley, W. C. Willett, B. M. Popkin, F. J. Chaloupka, J. W. Thompson, and D. S. Ludwig. 2009. The public health and economic benefits of taxing sugar-sweetened beverages. New England Journal of Medicine 361(16):1599-1605.
Carlson, A., M. Lino, W. Y. Juan, K. Hanson, and P. P. Basiotis. 2007. Thrifty Food Plan, 2006, CNPP-19. Washington, DC: USDA, CNPP. http://www.cnpp.usda.gov/Publications/FoodPlans/MiscPubs/TFP2006Report.pdf (accessed May 31, 2012).
CBO (Congressional Budget Office). 2008. Geographic variation in health care spending. Washington, DC: CBO. http://www.cbo.gov/sites/default/files/cbofiles/ftpdocs/89xx/doc8972/02-15-geoghealth.pdf (accessed August 13, 2012).
CBO. 2012. The Supplemental Nutrition Assistance Program. http://www.cbo.gov/publication/43173 (accessed August 13, 2012).
CBPP (Center on Budget and Policy Priorities). 2010. The food stamp dependent care deduction: Help for families with child care costs. http://www.cbpp.org/cms/index.cfm?fa=view&id=3130 (accessed August 13, 2012).
Chen, Z., S. T. Yen, and D. B. Eastwood. 2005. Effects of food stamp participation on body weight and obesity. American Journal of Agricultural Economics 87(5):1167-1173.
Children’s HealthWatch. 2012. Policy recommendations. http://www.childrenshealthwatch.org/page/PolicyRecs (accessed August 13, 2012).
CNPP (Center for Nutrition Policy and Promotion). 2011. Official USDA food plans: Cost of food at home at four levels, U.S. average, June 2011. http://www.cnpp.usda.gov/Publications/FoodPlans/2011/CostofFoodJun2011.pdf (accessed August 13, 2012).
Cole, N., and M. K. Fox. 2008. Diet quality of Americans by food stamp participation status: Data from the National Health and Nutrition Examination Survey, 1999-2004. Submitted by Abt Associates, Inc. to U.S. Department of Agriculture, Food and Nutrition Service, Alexandria, VA. http://www.fns.usda.gov/ora/menu/Published/snap/FILES/Participation/NHANES-FSP.pdf (accessed February 9, 2012).
Committee on Ways and Means. 2004. Background material and data on programs within the jurisdiction of the Committee on Ways and Means (Green Book). Washington, DC: Government Printing Office. http://www.gpo.gov/fdsys/search/pagedetails.action?granuleId=&packageId=GPO-CPRT-108WPRT108-6 (accessed June 4, 2012).
Conrad, D. A., and L. Perry. 2009. Quality-based financial incentives in health care: Can we improve quality by paying for it? Annual Review of Public Health 30:357-371.
Dachner, N., L. Ricciuto, S. I. Kirkpatrick, and V. Tarasuk. 2010. Food purchasing and food insecurity among low-income families in Toronto. Canadian Journal of Dietetic Practice and Research 71(3):e50-e56.
Davis, G. C., and W. You. 2010. The time cost of food at home: General and food stamp participant profiles. Applied Economics 42(20):2537-2552.
Eslami, E., K. Filion, and M. Strayer. 2011. Characteristics of Supplemental Nutrition Asssistance Program households: Fiscal year 2010. Submitted by Mathematica Policy Research, Inc. to U.S. Department of Agriculture, Food and Nutrition Service, Alexandria, VA. http://www.fns.usda.gov/ora/menu/Published/snap/FILES/Participation/2010Characteristics.pdf (accessed May 24, 2012).
Fisher, E., D. Goodman, J. Skinner, and K. Bronner. 2009. Health care spending, quality, and outcomes: More isn’t always better, Dartmouth Atlas Project Topic Brief. http://www.dartmouthatlas.org/downloads/reports/Spending_Brief_022709.pdf (accessed August 13, 2012).
Fletcher, J. M., D. Frisvold, and N. Tefft. 2010. Taxing soft drinks and restricting access to vending machines to curb childhood obesity. Health Affairs 29(5):1059-1066.
FNS (Food and Nutrition Service). 2007. Implications of restricting the use of food stamp benefits—Summary. http://www.fns.usda.gov/ora/menu/published/snap/files/programoperations/fspfoodrestrictions.pdf (accessed May 18, 2012).
FNS. 2011a. We welcome SNAP putting healthy food within reach: 2010 annual report. Alexandria, VA: USDA, FNS. http://www.fns.usda.gov/snap/retailers/pdfs/2010-annualreport.pdf (accessed June 14, 2012).
FNS. 2011b. Nutrition program facts: Supplemental Nutrition Assistance Program Education (SNAP-Ed). http://www.nal.usda.gov/snap/SNAP-EdFactsheet2011.pdf (accessed August 13, 2012).
FNS. 2012a. Building a healthy America: A profile of the Supplemental Nutrition Assistance Program. Alexandria, VA: USDA, FNS. http://www.fns.usda.gov/ora/MENU/published/snap/FILES/Other/BuildingHealthyAmerica.pdf (accessed May 22, 2012).
FNS. 2012b. Supplemental Nutrition Assistance Program: Eligible food items. http://www.fns.usda.gov/snap/retailers/eligible.htm (accessed June 7, 2012).
FNS. 2012c. Food and Nutrition Service 2012 Budget Explanatory Notes. http://www.obpa.usda.gov/30fns2012notes.pdf (accessed August 14, 2012).
FNS. 2012d. Healthy Incentives Pilot (HIP). http://www.fns.usda.gov/snap/hip/ (accessed June 7, 2012).
FNS. 2012e. Supplemental Nutrition Assistance Program: Eligibility. http://www.fns.usda.gov/snap/applicant_recipients/eligibility.htm (accessed August 13, 2012).
FNS. 2012f. FY 2012 minimum SNAP allotment. http://www.fns.usda.gov/snap/government/FY12_Minimum_Allotments.htm (accessed August 13, 2012).
Fox, M. K., W. Hamilton, and B. H. Lin. 2004. Effects of food assistance and nutrition programs on nutrition and health: Volume 3, literature review, FANRR 19-3. Submitted by Abt Associates, Inc. to U.S. Department of Agriculture, Economic Research Service, Washington, DC. http://webarchives.cdlib.org/sw1s17tt5t/http://ers.usda.gov/Publications/FANRR19-3/ (accessed July 3, 2012).
FRAC (Food Research and Action Center). 2010. A review of strategies to bolster SNAP’s role in improving nutrition as well as food security. Washington, DC: FRAC. http://frac.org/wp-content/uploads/2011/06/SNAPstrategies.pdf (accessed June 7, 2012).
Fraker, T., and R. Moffitt. 1988. The effect of food stamps on labor supply. A bivariate selection model. Journal of Public Economics 35(1):25-56.
Fraker, T. M., A. P. Martini, and J. C. Ohls. 1995. The effect of food stamp cashout on food expenditures. Journal of Human Resources 30(4):633-649.
Frazao, E., M. Andrews, D. Smallwood, and M. Prell. 2007. Food spending patterns of low-income households: Will increasing purchasing power result in healthier food choices? Economic Information Bulletin Number 29-4. USDA, ERS.
Gregory, C. A., and A. Coleman-Jensen. 2012. Do food prices affect food security? Evidence from the CPS 2002-2006. In Agricultural and Applied Economics Association 2011 Annual Meeting. Pittsburgh, PA.
Hagstrom, P. A. 1996. The food stamp participation and labor supply of married couples: An empirical analysis of joint decisions. Journal of Human Resources 31(2):383-403.
HHS and USDA (U.S. Department of Health and Human Services and U.S. Department of Agriculture). 2005. Dietary Guidelines for Americans. Washington, DC: U.S. Government Printing Office. http://www.health.gov/dietaryguidelines/dga2005/document (accessed June 4, 2012).
Hodges, S., and B. Emerson. 2012. A primer on the Restaurant Meals Program in California. http://www.snaprmp.org/docs/RestaurantMealsProgram.pdf (accessed March 14, 2013).
Hoynes, H. W., and D. W. Schanzenbach. 2012. Work incentives and the Food Stamp Program. Journal of Public Economics 96(1-2):151-162.
IOM (Institute of Medicine). 1997. Dietary reference intakes for calcium, phosphorus, magnesium, vitamin D, and fluoride. Washington, DC: National Academy Press.
IOM. 1998. Dietary reference intakes for thiamin, riboflavin, niacin, vitamin B6, folate, vitamin B12, pantothenic acid, biotin, and choline. Washington, DC: National Academy Press.
IOM. 2000. Dietary reference intakes for vitamin C, vitamin E, selenium, and carotenoids. Washington, DC: National Academy Press.
IOM. 2001. Dietary reference intakes for vitamin A, vitamin K, arsenic, boron, chromium, copper, iodine, iron, manganese, nickel, silicon, vanadium, and zinc. Washington, DC: National Academy Press.
IOM. 2005a. Dietary reference intakes for energy, carbohydrate, fiber, fat, fatty acids, cholesterol, protein, and amino acids. Washington, DC: The National Academies Press.
IOM. 2005b. Dietary reference intakes for water, potassium, sodium, chloride, and sulfate. Washington, DC: The National Academies Press.
Kane, R. L., P. E. Johnson, R. J. Town, and M. Butler. 2004. A structured review of the effect of economic incentives on consumers’ preventive behavior. American Journal of Preventive Medicine 27(4):327-352.
Keane, M., and R. Moffitt. 1998. A structural model of multiple welfare program participation and labor supply. International Economic Review 39(3):553-589.
Lee, B. J., L. Mackey-Bilaver, M. Chin, and T. A. Majchrowicz. 2006. Effects of WIC and Food Stamp Program participation on child outcomes, CCR-27. Washington, DC: USDA, ERS. http://naldc.nal.usda.gov/catalog/33688 (accessed July 5, 2012).
Leftin, J., E. Eslami, and M. Strayer. 2011. Trends in Supplemental Nutrition Assistance Program participation rates: Fiscal year 2002 to fiscal year 2009: Final report. Submitted by Mathematica Policy Research, Inc. to U.S. Department of Agriculture, Food and Nutrition Service, Alexandria, VA. http://www.mathematica-mpr.com/publications/PDFs/nutrition/trends2002-09.pdf (accessed October 10, 2012).
Leibtag, E. 2007. Can food stamps do more to improve food choices? An economic perspective—stretching the food stamp dollar: Regional price difference affect affordability of food, EIB-29-2. Washington, DC: USDA, ERS. http://www.ers.usda.gov/publications/eib-economic-information-bulletin/eib29-2.aspx (accessed August 13, 2012).
Lin, B. H., and A. Carlson. 2010. SNAP benefits and eating out: Wise choices required. Amber Waves. http://webarchives.cdlib.org/sw1vh5dg3r/http://www.ers.usda.gov/AmberWaves/March10/Findings/SnapBenefits.htm (accessed December 31, 2012).
Mabli, J., L. Castner, J. Ohls, M.K. Fox, M.K. Crepinsek, and E. Condon. 2010. Food expenditures and diet quality among low-income households and individuals: Final report. Submitted by Mathematica Policy Research, Inc. to U.S. Department of Agriculture, Food and Nutrition Service, Alexandria, VA. http://www.mathematica-mpr.com/publications/PDFs/nutrition/FoodExpendDietQuality.pdf (accessed August 13, 2012).
Martin, K. S., E. Havens, K. E. Boyle, G. Matthews, E. A. Schilling, O. Harel, and A. M. Ferris. 2012. If you stock it, will they buy it? Healthy food availability and customer purchasing behaviour within corner stores in Hartford, CT, USA. Public Health Nutrition 15(10):1973-1978.
McGranahan, L., and D. W. Schanzenbach. 2011. Who would be affected by soda taxes? Chicago Federal Letter, Number 284, http://www.chicagofed.org/digital_assets/publications/chicago_fed_letter/2011/cflmarch2011_284.pdf (accessed December 31, 2012).
Meyer, B. D., and D. T. Rosenbaum. 2001. Welfare, the earned income tax credit, and the labor supply of single mothers. Quarterly Journal of Economics 116(3):1063-1114.
Moretti, E. 2011. Local labor markets. In Handbook of Labor Economics, Vol. 4, Part B, edited by O. Ashenfelter and D. Card. Amsterdam: North Holland. Pp. 1237-1313.
NACCRRA (National Association of Child Care Resources and Referral Agencies). 2011. Child care in America: 2011 state fact sheet. http://www.naccrra.org/sites/default/files/default_site_pages/2011/childcareinamericafacts_2011_final.pdf (accessed August 14, 2012).
Nord, M., and M. Golla. 2009. Does SNAP decrease food insecurity? Untangling the self-selection effect, ERR-85. Washington, DC: USDA, ERS. http://www.ers.usda.gov/publications/err-economic-research-report/err85.aspx (accessed August 13, 2012).
Nord, M., and M. Prell. 2011. Food security improved following the 2009 ARRA increase in SNAP benefits, ERR-116. Washington, DC: USDA, ERS. http://www.ers.usda.gov/publications/err116 (accessed June 7, 2012).
NRC (National Research Council). 1989. Diet and health: Implications for reducing chronic disease. Washington, DC: National Academy Press.
NRC. 1995. Measuring poverty: A new approach. Washington, DC: National Academy Press.
Orshansky, M. 1957. Food consumption and dietary levels of households in the United States: Some highlights from the Household Food Consumption Survey, Spring 1955, ARS 62-6. Washington, DC: USDA, ARS.
Pirog, M. A., M. E. Klotz, and K. V. Byers. 1998. Interstate comparisons of child support orders using state guidelines. Family Relations 47(3):289-295.
Renwick, T. 2011. Geographic adjustments of Supplemental Poverty Measure thresholds: Using the American Community Survey five-year data on housing costs. Presented at the January 2011 Annual Convention of the Allied Social Sciences Association in Denver, CO. http://www.census.gov/hhes/povmeas/methodology/supplemental/research/Renwick_SGE2011.pdf (accessed August 13, 2012).
RIDHS (Rhode Island Department of Human Services). 2012. The Rhode Island Supplemental Nutrition Assistance Program (SNAP) prepared meals “Food Access Pilot Project.” http://www.dhs.ri.gov/Portals/0/Uploads/Documents/SNAP/Q_and_A.pdf (accessed June 7, 2012).
Rose, D. 2007. Food stamps, the Thrifty Food Plan, and meal preparation: The importance of the time dimension for US nutrition policy. Journal of Nutrition Education and Behavior 39(4):226-232.
Rose, D., J. P. Habicht, and B. Devaney. 1998. Household participation in the Food Stamp and WIC Programs increases the nutrient intakes of preschool children. Journal of Nutrition 128(3):548-555.
Rosenbaum, D., D. Tenny, and S. Elkin. 2002. The Food Stamp Shelter Deduction: Helping households with high housing burdens meet their food needs. Washington, DC: CBPP. http://www.cbpp.org/files/7-1-02fs.pdf (accessed August 13, 2012).
Shenkin, J. D., and M. F. Jacobson. 2010. Using the food stamp program and other methods to promote healthy diets for low-income consumers. American Journal of Public Health 100(9):1562-1564.
Short, K. S. 2012. The Supplemental Poverty Measure: Examining the incidence and depth of poverty in the U.S. taking account of taxes and transfers in 2010, SEHSD working paper # 2012-05. http://www.census.gov/hhes/povmeas/methodology/supplemental/research/sea2011.pdf (accessed August 13, 2012).
Thayer, J., C. Murphy, J. Cook, de Cuba, S. E., R. DaCosta, and M. Chilton. 2008. Coming up short: High food costs oustrip food stamp benefits. Boston, MA: Children’s Sentinel Nutrition Assessment Program. http://www.childrenshealthwatch.org/upload/resource/RCOHD_Report_Final.pdf (accessed August 13, 2012).
Todd, J., E. Leibtag, and C. Penberthy. 2011. Geographic differences in the relative price of healthy foods, EIB-78. Washington, DC: USDA, ERS. http://www.ers.usda.gov/publications/eib-economic-information-bulletin/eib78.aspx (accessed July 24, 2012).
U.S. Census Bureau. 2012. Current Population Survey Annual Social and Economic Supplement (CPS ASEC). http://www.census.gov/hhes/www/poverty/publications/pubs-cps.html (accessed August 6, 2012).
USDA and HHS (U.S. Department of Agriculture and U.S. Department of Health and Human Services). 2005. Dietary guidelines for Americans. 6th ed. Washington, DC: U.S. Government Printing Office.
Volpp, K. G., L. K. John, A. B. Troxel, L. Norton, J. Fassbender, and G. Loewenstein. 2008. Financial incentive-based approaches for weight loss: A randomized trial. Journal of the American Medical Association 300(22):2631-2637.
Volpp, K. G., M. V. Pauly, G. Loewenstein, and D. Bangsberg. 2009a. Market watch. P4P4P: An agenda for research on pay-for-performance for patients. Health Affairs 28(1):206-214.
Volpp, K. G., A. B. Troxel, M. V. Pauly, H. A. Glick, A. Puig, D. A. Asch, R. Galvin, J. Zhu, F. Wan, J. Deguzman, E. Corbett, J. Weiner, and J. Audrain-McGovern. 2009b. A randomized, controlled trial of financial incentives for smoking cessation. New England Journal of Medicine 360(7):699-709.
Wilde, P. 2001. Understanding the Food Stamp Benefit Formula: A tool for measuring the component effects. Washington, DC: USDA, ERS. http://www.ers.usda.gov/media/330003/fanrr14fm_1_.pdf (accessed August 13, 2012).
Wilde, P. 2002. Issues in Food Assistance: The Standard Deduction in the Food Stamp Benefit Formula, FANNR 26-3. USDA, ERS. http://naldc.nal.usda.gov/download/36552/PDF (accessed August 13, 2012).
Wilde, P. E., P. E. McNamara, and C. K. Ranney. 1999. The effect of income and food programs on dietary quality: A seemingly unrelated regression analysis with error components. American Journal of Agricultural Economics 81(4):959-971.
Yen, S. T., B. H. Lin, D. M. Smallwood, and M. Andrews. 2004. Demand for non-alcoholic beverages: The case of low-income households. Agribusiness 20(3):309-321.
You, W., G. Zhang, B. M. Davy, A. Carlson, and B. H. Lin. 2009. Food consumed away from home can be a part of a healthy and affordable diet. Journal of Nutrition 139(10):1994-1999.
Ziliak, J. P. 2008. Effective tax rates and guarantees in the Food Stamp Program. University of Kentucky. http://gatton.uky.edu/Faculty/ziliak/ERS_FSP_Rates&Guarantees_042308.pdf (accessed June 5, 2012).