This is the accessible text file for GAO report number GAO-08-415 
entitled 'Food Stamp Program: Options for Delivering Financial 
Incentives to Participants for Purchasing Targeted Foods' which was 
released on July 30, 2008.

This text file was formatted by the U.S. Government Accountability 
Office (GAO) to be accessible to users with visual impairments, as part 
of a longer term project to improve GAO products' accessibility. Every 
attempt has been made to maintain the structural and data integrity of 
the original printed product. Accessibility features, such as text 
descriptions of tables, consecutively numbered footnotes placed at the 
end of the file, and the text of agency comment letters, are provided 
but may not exactly duplicate the presentation or format of the printed 
version. The portable document format (PDF) file is an exact electronic 
replica of the printed version. We welcome your feedback. Please E-mail 
your comments regarding the contents or accessibility features of this 
document to Webmaster@gao.gov. 

This is a work of the U.S. government and is not subject to copyright 
protection in the United States. It may be reproduced and distributed 
in its entirety without further permission from GAO. Because this work 
may contain copyrighted images or other material, permission from the 
copyright holder may be necessary if you wish to reproduce this 
material separately. 

Report to the Chairman, Committee on Agriculture, Nutrition, and 
Forestry, U.S. Senate: 

United States Government Accountability Office: 
GAO: 

July 2008: 

Food Stamp Program: 

Options for Delivering Financial Incentives to Participants for 
Purchasing Targeted Foods: 

GAO-08-415: 

GAO Highlights: 

Highlights of GAO-08-415, a report to the Chairman, Committee on 
Agriculture, Nutrition, and Forestry, U.S. Senate. 

Why GAO Did This Study: 

In fiscal year 2007, the Food Stamp Program provided about $30.4 
billion in nutrition assistance benefits to 26.5 million individuals. 
Benefits are issued through Electronic Benefit Transfer (EBT) cards, 
similar to debit cards, to purchase eligible foods at authorized retail 
stores. The diets of many low-income individuals, like the U.S. 
population overall, do not meet federal dietary guidelines. One 
potential strategy for increasing the purchases of targeted foods that 
contribute to a healthy diet is to incorporate into the program 
financial incentives for purchasing these foods. GAO was asked to 
identify (1) what is known about the effectiveness of financial 
incentives and other approaches intended to increase the purchase of 
targeted foods, (2) the key factors to consider in designing a 
financial incentive program, and (3) options available to the U.S. 
Department of Agriculture’s Food and Nutrition Service (FNS) for 
implementing financial incentives. GAO interviewed agency and state 
officials, retailers and associations, private EBT contractors, and 
other stakeholders; convened a panel of 17 experts; and conducted a 
literature review. 

In commenting on this report, FNS generally agreed with GAO’s findings 
and concluding observations. 

What GAO Found: 

A variety of approaches, including financial incentives and nutrition 
education, can increase the consumption of targeted foods, but little 
is known about the effectiveness of efforts to increase access to 
targeted foods. A few studies examining the effectiveness of financial 
incentives have demonstrated short-term positive effects on purchases, 
consumption, or weight loss. A study that reviewed 92 nutrition 
education studies found that most studies reported significant positive 
effects on consumption. Because of a lack of reported research, little 
is known about the effectiveness of approaches intended to improve 
access to targeted foods. 

Factors to consider in designing a program that delivers financial 
incentives through an additional food stamp allotment tied to the 
purchase of targeted foods include the following: 

* Selection of foods: Selecting which foods to promote could be a 
controversial and challenging part of designing an incentive program. 

* Incentive amount: The amount of the incentive will affect participant 
response and program costs. 

* Informing participants: Participants must be informed of the 
availability of incentives to take full advantage of a new incentive 
program. 

* Program monitoring and evaluation: Monitoring and evaluating the 
incentive program is critical to maintaining program integrity and 
determining the effects of the program. 

An incentive program could be implemented through either electronic or 
paper methods, and the different options would have implications for 
ease of implementation, program integrity, and cost. Electronic options 
include adding food benefits to the EBT cards currently used or to a 
separate card on the basis of the amount that participants spend on 
targeted foods. Providing incentives using existing EBT cards would 
build on the current checkout technology and process, and could require 
less time to complete transactions compared with using a separate card. 
However, delivering financial incentives to participants’ EBT accounts 
on the basis of their purchases of certain foods would require several 
changes to the EBT system, such as modifications to retailer and EBT 
contractor software to separately track the amount spent on the 
targeted food items. Administering incentives using a separate card, 
such as an additional EBT card, may be more costly and complicated to 
implement. Alternatively, providing participants with paper vouchers 
for the purchase of targeted foods would not require changes to the EBT 
system, but could be more burdensome to use, increase fraud risk, and 
increase state administrative costs. With the new authority provided as 
part of the Food, Conservation, and Energy Act of 2008, FNS will have 
the opportunity to develop and administer a pilot incentive program and 
to determine both its effects on participant purchasing and consumption 
patterns and the costs associated with such a program. 

To view the full product, including the scope and methodology, click on 
[hyperlink, http://www.gao.gov/cgi-bin/getrpt?GAO-08-415]. For more 
information, contact Kay Brown at (202) 512-7215 or brownke@gao.gov. 

[End of section] 

Contents: 

Letter: 

Results in Brief: 

Background: 

Variety of Approaches Can Increase the Consumption of Targeted Foods: 

Factors to Consider Include the Selection of Targeted Foods, Incentive 
Amount, Participant Education, and Monitoring and Evaluation: 

Financial Incentives Could Be Implemented Electronically or Using Paper 
Methods, but Each Option Has Potential Implications for Ease of 
Implementation, Program Integrity, and Cost: 

Concluding Observations: 

Agency Comments and Our Evaluation: 

Appendix I: Objectives, Scope, and Methodology: 

Appendix II: GAO Contact and Staff Acknowledgments: 

Related GAO Products: 

Tables: 

Table 1: Percentage of Authorized Firms and FSP Redemptions, by 
Category, for Fiscal Year 2007: 

Table 2: Required Changes to Implement the Option Preferred by Expert 
Panelists: 

Figures: 

Figure 1: Example of a FSP EBT Transaction: 

Figure 2: Percentage of U.S. Population Aged 2 Years and Older, Meeting 
Specific Dietary Guidelines for the Consumption of Food Groups (Data 
Collected between 1988 and 1996): 

Figure 3: Theoretical Relationships between Nutrition Promotion 
Approaches and Outcomes and Impacts: 

Figure 4: Number of Studies Reporting Significant Nutrition Education 
Effects: 

Figure 5: Incentive Amount Added to Existing EBT Card on the Basis of 
the Amount Spent on Targeted Foods: 

Figure 6: Incentive Amount Added to a Separate Targeted Foods Card on 
the Basis of the Amount Spent on Targeted Foods: 

Figure 7: Incentive Provided to Participants Using Paper Vouchers as a 
Fixed Amount to Be Spent on Targeted Foods: 

Abbreviations: 

EBT: Electronic Benefit Transfer: 

EFNEP: Expanded Food and Nutrition Education Program: 

ERS: Economic Research Service: 

FNS: Food and Nutrition Service: 

FSNE: Food Stamp Nutrition Education: 

FSP: Food Stamp Program: 

PIN: personal identification number: 

PLU: Price Look-Up code: 

POS: point of sale: 

TANF: Temporary Assistance for Needy Families: 

UPC: Universal Product Code: 

USDA: U.S. Department of Agriculture: 

WIC: Special Supplemental Nutrition Program for Women, Infants, and 
Children: 

[End of section] 

United States Government Accountability Office:
Washington, DC 20548: 

July 30, 2008: 

The Honorable Tom Harkin: 
Chairman: 
Committee on Agriculture, Nutrition, and Forestry: 
United States Senate: 

Dear Mr. Chairman: 

The diets of many low-income individuals, like the U.S. population 
overall, do not meet federal dietary guidelines. The typical American 
diet includes higher-than-recommended intakes of fat, added sugars, and 
sodium and lower-than-recommended servings of milk, whole grains, 
fruits, and vegetables. Poor eating habits can increase the risk of 
serious health problems, such as heart disease, cancer, stroke, 
diabetes, and hypertension. Changing food purchasing and consumption 
patterns, therefore, might produce improvements in the nation's health. 
The Food Stamp Program (FSP), whose mission is to improve the nutrition 
of low-income families, provides one possible vehicle for influencing 
dietary patterns for large numbers of individuals. This program 
provided about $30.4 billion in food benefits to about 26.5 million 
individuals each month during fiscal year 2007. 

The FSP is jointly administered by the U.S. Department of Agriculture's 
(USDA) Food and Nutrition Service (FNS) and states. States generally 
administer the program from local offices where caseworkers determine 
whether households meet the program's eligibility requirements, 
calculate monthly benefits for qualified households, and issue benefits 
through Electronic Benefit Transfer (EBT) cards. These cards, similar 
to debit cards, allow participants to purchase eligible foods at retail 
stores authorized by FNS to accept FSP benefits. At the end of fiscal 
year 2007, about 165,000 retailers--ranging from large supermarkets to 
convenience stores and farmers' markets--were authorized to accept FSP 
benefits. In addition to providing food assistance, states have the 
option to offer federally funded nutrition education to FSP 
participants and eligible nonparticipants. FNS also administers the 
Special Supplemental Nutrition Program for Women, Infants, and Children 
(WIC) through grants to the states. Although benefits in this program 
are primarily delivered through a paper-based voucher system, FNS has 
recently provided some states with funds to pilot an EBT system. The 
WIC program provides food, nutrition education, and health care 
referrals to low-income pregnant and postpartum women, infants, and 
children under aged 5 years who are determined to be at nutritional 
risk. In fiscal year 2007, the WIC program served an average of about 
8.3 million women, infants, and children per month, and federal program 
costs amounted to over $5.5 billion. 

The Food, Conservation, and Energy Act of 2008 (the 2008 Farm Bill) 
includes provisions authorizing USDA to fund pilot projects in the FSP 
to assess whether healthier food purchases result from projects that 
provide financial incentives, nutrition education, and improved access 
to healthy foods, among other options.[Footnote 1] Given the variety of 
approaches available to influence food purchases, you asked us to 
assess what is known about their effectiveness in increasing purchases 
of targeted foods that could contribute to a healthy diet. In addition, 
because the technical and operational challenges of implementing a 
system of financial incentives at the point of sale (POS) are largely 
unknown, you asked us to determine the key factors to consider in 
designing such a program, options available to implement financial 
incentives, and advantages and challenges involved in their 
implementation. To address these issues, we examined the following: 

1. What is known about the effectiveness of financial incentives and 
other approaches intended to increase the purchase of targeted foods 
that could contribute to a healthy diet? 

2. What are the key factors to consider in designing a program that 
provides FSP participants with financial incentives to purchase certain 
foods? 

3. What options are available to FNS for implementing financial 
incentives, and what are the advantages and challenges involved in 
implementing such options? 

To identify recent studies that examine the effectiveness of nutrition 
promotion approaches, we conducted a literature review and consulted 
with USDA staff. We identified related studies published from January 
2000 through February 2008 and conducted detailed reviews of 16 
studies, including reviews or summaries of original research from peer- 
reviewed journals, published research from USDA's Economic Research 
Service (ERS), and studies of WIC farmers' market programs. We reviewed 
each study's research methodology and determined that the studies were 
reliable for the purposes of our work. 

To understand the issues involved in designing a financial incentive 
program, we interviewed representatives of stakeholder groups that 
would be affected by implementation of an incentive program, including 
USDA officials; officials from four states with WIC EBT experience 
(Ohio, Texas, Washington, and Wyoming); officials from one state 
(California) that passed legislation authorizing the provision of 
financial incentives to FSP participants for purchasing targeted foods; 
three EBT contractors serving almost all of the states; retailer 
associations and retailers representing large chain and smaller 
independent grocery stores; food and nutrition advocacy groups; and 
researchers. We also facilitated group discussions about the options 
for delivering financial incentives and implementation issues with 
state officials, retailers, and industry representatives at three 
national conferences. To further explore financial incentive options 
and implementation issues, we convened a panel of 17 experts 
representing stakeholder groups that would be involved in implementing 
an incentive program, including FNS, states, state EBT contractors, 
retailers, and manufacturers of retailer checkout systems. Appendix I 
explains our scope and methodology in more detail. We conducted this 
performance audit from May 2007 through July 2008 in accordance with 
generally accepted government auditing standards. Those standards 
require that we plan and perform the audit to obtain sufficient, 
appropriate evidence to provide a reasonable basis for our findings and 
conclusions based on our audit objectives. We believe that the evidence 
obtained provides a reasonable basis for our findings and conclusions 
based on our audit objectives. 

Results in Brief: 

A variety of approaches, including financial incentives and nutrition 
education, can increase the consumption of targeted foods that could 
contribute to a healthy diet, but little is known about the 
effectiveness of efforts to increase access to such foods. A few 
studies examining the effectiveness of financial incentives have 
demonstrated short-term positive effects on purchases of targeted 
foods, consumption, or weight loss. For example, an evaluation of a 
Farmers' Market Nutrition Program implemented in a county in Michigan 
found that after giving $20 in farmers' market coupons to 564 low- 
income women, they self-reported a significant increase in fruit and 
vegetable consumption. Nutrition education approaches, such as 
individual counseling and group classes, can positively affect fruit 
and vegetable consumption, particularly among populations at risk of or 
diagnosed with disease. Providing nutrition information at the point of 
purchase--such as posters, brochures, and shelf and food labels--is 
another nutrition education approach that can influence food purchasing 
behavior and consumption, although the results from reported research 
on these approaches are mixed. Little is known about the effectiveness 
of approaches aimed at improving access to targeted foods because of a 
lack of reported research. Study limitations, including a lack of 
information on costs and sustained effects as well as methodological 
differences among studies, make it difficult to compare approaches or 
determine which approaches are most effective. 

Factors to consider in designing a financial incentive program include 
the foods to target for promotion, the incentive amount, how 
participants would be informed of the new benefit, and how the program 
would be monitored and evaluated. 

* Selection of foods: Expert panelists and many stakeholders we 
interviewed agreed that an incentive program would be easier to 
implement if incentives were provided for a general food category, such 
as fruits and vegetables, rather than for individual food items that 
either meet certain nutritional criteria or contain a particular 
ingredient, such as whole grains. Even using general food categories 
could be challenging, however, since some foods with less nutritive 
value might be included in a category. Another issue to consider is the 
availability of the targeted foods, which could vary by retailer. 

* Incentive amount: Although the optimal value of an incentive to 
increase targeted food purchases has not been determined, ERS estimated 
in a study that a 20 percent reduction in the price of fruits and 
vegetables would result in a modest increase in these purchases by 
average FSP participants. The incentive amount would affect overall 
program costs. 

* Informing participants: Participants must be informed of the 
availability of incentives to take full advantage of a new incentive 
program, but reaching them could be a challenge and would likely 
increase the workload of local assistance offices and nutrition 
education providers. 

* Program monitoring and evaluation: FNS indicated it would need to 
incorporate any new incentive program into its existing oversight 
activities by monitoring the total amount spent with the incentive 
benefits on the targeted foods and developing criteria to flag 
suspicious transactions since opportunities for fraud may be increased. 
Expert panelists agreed that evaluating the incentive program is 
critical to ensure that program integrity is not eroded and to 
determine the effects of the incentives on participants' food choices. 

A financial incentive program--defined as an additional FSP allotment 
tied to the purchase of certain foods targeted for promotion--could be 
implemented through electronic or paper methods, and the different 
options would have implications for ease of implementation, program 
integrity, and cost. Electronic options include providing additional 
FSP benefits to the cards currently used or to a separate card on the 
basis of the amount that participants spend on targeted foods. 
Providing incentives using existing EBT cards would build on the 
current checkout technology and process, and could require less time to 
complete transactions compared with using a separate card. Implementing 
incentives electronically on EBT cards could also minimize fraud risk 
by requiring entry of a personal identification number (PIN) to 
complete transactions. However, delivering financial incentives to 
participants' EBT accounts on the basis of their purchases of certain 
foods would require several changes to the EBT system. For example, it 
would require modifications to retailer and EBT contractor software to 
separately track the amount spent on the targeted food items. 
Administering incentives using a separate card, such as an additional 
EBT card, may be more costly and complicated to implement. 
Alternatively, providing FSP participants with paper vouchers for the 
purchase of targeted foods would not require changes to the EBT system, 
but could be more burdensome to use, increase fraud risk, and increase 
state administrative costs. With the new authority provided as part of 
the 2008 Farm Bill, FNS will have the opportunity to develop and 
administer a pilot incentive program and to determine both the 
program's effects on participant purchasing and consumption patterns 
and the costs associated with such a program. 

In commenting on this report, FNS officials agreed with our findings 
and concluding observations. However, they raised concerns about the 
opportunity for fraud in a financial incentive program as well as the 
potential cost of such a program. 

Background: 

The FSP is intended to help low-income individuals and families obtain 
a more nutritious diet by supplementing their income with benefits to 
purchase food. Most food items sold in grocery stores can be purchased 
with FSP benefits, except for hot foods, tobacco products, and 
alcoholic beverages.[Footnote 2] FNS pays the full cost of FSP benefits 
and shares the states' administrative costs--with FNS usually paying 
approximately 50 percent. FNS is responsible for promulgating program 
regulations and ensuring that state officials administer the program in 
compliance with program rules. States usually administer the program 
from local assistance offices where staff determine whether households 
meet the program's income and asset requirements, calculate monthly 
benefits for qualified households, and issue benefits to participants 
on an EBT card. The local assistance offices often administer other 
benefit programs as well, such as Temporary Assistance for Needy 
Families (TANF) and Medicaid.[Footnote 3] 

Determination of Eligibility and Benefits: 

FSP eligibility is largely based on a household's income and assets. To 
determine a household's eligibility, a caseworker first determines the 
household's gross income and net income. Gross income cannot exceed 130 
percent of the federal poverty level for that year as determined by the 
Department of Health and Human Services, and net income cannot exceed 
100 percent of the federal poverty level (or about $20,652 annually for 
a family of four living in the continental United States in fiscal year 
2008). Net income is determined by deducting a portion of specified 
expenses from gross income, such as dependent care, medical needs, 
utilities, and housing. 

Use of Electronic Benefit Transfer Cards to Redeem FSP Benefits: 

Prior to EBT, participants used paper coupons to pay for allowable 
foods. The Personal Responsibility and Work Opportunity Reconciliation 
Act of 1996 required each state agency to implement an EBT system to 
electronically distribute FSP benefits. The last state completed its 
implementation in fiscal year 2004. States are responsible for 
coordinating and managing the EBT system in their state, but they 
generally award contracts to private sector companies to develop and 
operate the system. Under the EBT system, FSP participants receive an 
EBT card imprinted with their name and account number. FSP benefits are 
automatically credited to the participants' accounts once a month. 
According to FNS, more than half of the states use EBT cards to 
distribute benefits from other federal programs, such as TANF cash 
assistance.[Footnote 4] 

The EBT system uses magnetic strip EBT cards that work like commercial 
debit cards and an online system where transactions are processed as 
they occur. As shown in figure 1, to complete a FSP EBT purchase, the 
store clerk totals the amount of eligible items on the cash register, 
and the amount of the EBT purchase is entered into the retailer's POS 
device. The EBT card is then swiped through the POS device at the 
grocery checkout counter, and the participant enters a PIN to access 
the account. POS devices read the recipient identification information 
from the magnetic strip on the back of the EBT card. That information, 
along with the total amount of the EBT purchase, is sent to the state's 
EBT contractor for approval. Information about the items purchased is 
not transmitted through the FSP EBT system. Electronically, the EBT 
contractor verifies certain information and sends an authorization or 
denial back to the retailer. The purchase amount is then subtracted 
from the participant's account balance and added to the retailer's bank 
account through a settlement process at the end of each business day. 
Retailers can also be reimbursed for EBT purchases by using manual 
paper procedures if the EBT system or an EBT card is inoperable or if 
the retailer does not have access to telephone lines and electricity. 

Figure 1: Example of a FSP EBT Transaction: 

[See PDF for image] 

This figure is an illustration of an FSP EBT transaction, as follows: 

Customer-retailer interaction: 
* Retailer rings up $50 worth of groceries; 
* Customer swipes EBT card and enters PIN to access account; 
* Customer received $50 worth of groceries. 

Request for payment is sent to State EBT contractor: 
* State EBT contractor verifies retailer’s license number and 
customer’s account number, PIN, and account balance, then authorizes 
transaction; 
* Approval is returned to retailer. 

Federal Reserve Bank: 
* State EBT contractor orders $50 reimbursement for retailer from FNS’s 
Federal Reserve Bank account. 

Retailer's bank: 
* Retailer is reimbursed $50 as part of the daily settlement process at 
end of business day. 

Source: GAO, Nova Development Corporation. 

[End of figure] 

Retailers use different checkout systems to process EBT transactions. 
Many large volume retailers have integrated checkout systems that 
automatically identify and subtotal the FSP-eligible items when items 
are scanned and display the EBT purchase amount on the POS device. 
Retailers with scanning systems develop and maintain a computer file 
that contains the prices associated with the Universal Product Codes 
(UPC) and Price Look-Up (PLU) codes of the items sold in the store. 
[Footnote 5] The computer file can be programmed to contain a FSP 
eligibility "flag" associated with each item, which can help the 
retailer ensure that only FSP-eligible items are purchased with EBT 
cards. Some retailers use commercial POS equipment that can accept 
multiple payment types, such as EBT, debit, and credit cards. Retailers 
that do not accept debit or credit cards but average more than $100 in 
monthly FSP sales often have state-provided, stand-alone POS equipment 
that can only process EBT transactions. These stand-alone devices are 
not integrated with the retailer's cash register system. Without a 
scanner and an integrated checkout system, store clerks have to 
manually separate and subtotal the eligible and ineligible items and 
manually enter the EBT purchase amount in the POS device. 

Characteristics of Authorized Retailers: 

FNS has the primary responsibility for authorizing and overseeing 
retailers that accept FSP benefits. To become an authorized retailer, a 
store must offer on a continuing basis a variety of foods in each of 
the four staple food categories--meats, poultry, or fish; breads or 
cereals; vegetables or fruits; and dairy products--or 50 percent of its 
sales must be in a staple group, such as meat or bakery items. At the 
end of fiscal year 2007, about 165,000 retailers were authorized to 
accept FSP benefits. Large stores redeem most of the FSP benefits, as 
shown in table 1. In fiscal year 2007, supermarkets and superstores 
accounted for about 21 percent of the authorized stores but redeemed 
about 85 percent of FSP benefits.[Footnote 6] 

Table 1: Percentage of Authorized Firms and FSP Redemptions, by 
Category, for Fiscal Year 2007: 

Type of firm: Supermarkets; 
Percentage of total authorized firms: 13.16%; 
Percentage of total FSP dollars redeemed: 49.98%. 

Type of firm: Superstores; 
Percentage of total authorized firms: 7.83%; 
Percentage of total FSP dollars redeemed: 35.18%. 

Type of firm: Grocery stores[A]; 
Percentage of total authorized firms: 17.65%; 
Percentage of total FSP dollars redeemed: 6.24%. 

Type of firm: Convenience stores; 
Percentage of total authorized firms: 35.45%; 
Percentage of total FSP dollars redeemed: 4.32%. 

Type of firm: Combination stores[B]; 
Percentage of total authorized firms: 15.48%; 
Percentage of total FSP dollars redeemed: 1.62%. 

Type of firm: All other stores[C]; 
Percentage of total authorized firms: 8.72%; 
Percentage of total FSP dollars redeemed: 2.43%. 

Type of firm: Meal services[D]; 
Percentage of total authorized firms: 1.71%; 
Percentage of total FSP dollars redeemed: 0.23%. 

Type of firm: Total; 
Percentage of total authorized firms: 100%; 
Percentage of total FSP dollars redeemed: 100%. 

Source: Food and Nutrition Service data. 

[A] This category includes small, medium, and large grocery stores. 

[B] This category includes stores such as independent drug stores, 
dollar stores, and general stores. 

[C] This category includes farmers' markets, wholesalers, specialty 
food stores, health/natural food stores, nonprofit food-buying co-ops, 
military commissaries, and delivery routes. 

[D] This category includes drug and alcohol treatment centers, group 
homes, and communal dining facilities or meals-on-wheels for seniors. 

[End of table] 

In addition to approving retailers to participate in the program, FNS 
headquarters officials collect and monitor EBT transaction data to 
detect suspicious patterns of transactions by retailers. As part of the 
monitoring process, FNS collects each day's FSP transaction data from 
the states' EBT contractors and adds these data to its EBT transaction 
database for analysis. Information on the amount of the transaction is 
reported. Information on the items being purchased is not available 
through EBT. The system scans these data to identify transactions or 
sets of transactions that fit a certain set of criteria defined by 
established patterns of fraudulent activity. 

Adherence to Federal Dietary Guidelines: 

According to ERS, research has shown that the diets of most Americans 
do not follow the pattern recommended by the 2005 Dietary Guidelines 
for Americans and USDA's MyPyramid. The federal dietary guidelines 
describe a healthy diet as one that emphasizes fruits, vegetables, 
whole grains, and fat-free or low-fat milk and milk products; that 
includes lean meats, poultry, fish, beans, eggs, and nuts; and that is 
low in saturated fats, trans fats, cholesterol, salt, and added sugars. 
Two cups of fruit (4 servings) and 2½ cups of vegetables (5 servings) 
per day are recommended for a reference 2,000-calorie intake, with 
higher or lower amounts depending on the calorie level. The federal 
guidelines recommend that total fat intake be between 20 percent and 35 
percent of daily calories, with most fats coming from sources of 
polyunsaturated and monounsaturated fatty acids, such as fish, nuts, 
and vegetable oils. 

As shown in figure 2, less than half of the U.S. population met 
specific dietary guidelines for the consumption of many of the 
recommended food groups and nutrients.[Footnote 7] A greater percentage 
of people in higher-income households met recommendations for the 
consumption of certain food groups and nutrients than did people living 
in lower-income households. For example, 50 percent of those living in 
higher-income households met the vegetable recommendations, compared 
with 42 percent in lower-income households. 

Figure 2: Percentage of U.S. Population Aged 2 Years and Older, Meeting 
Specific Dietary Guidelines for the Consumption of Food Groups (Data 
Collected between 1988 and 1996): 

[See PDF for image] 

This figure is a multiple vertical bar graph depicting the following 
data: 

Food groups and nutrients: Fruits; 
Lower-income households meeting dietary guidelines: 23%; 
Higher-income households meeting dietary guidelines: 29%. 

Food groups and nutrients: Vegetables; 
Lower-income households meeting dietary guidelines: 42%; 
Higher-income households meeting dietary guidelines: 50%. 

Food groups and nutrients: Grains; 
Lower-income households meeting dietary guidelines: 44%; 
Higher-income households meeting dietary guidelines: 53%. 

Food groups and nutrients: Whole grains; 
Lower-income households meeting dietary guidelines: 4%; 
Higher-income households meeting dietary guidelines: 7%. 

Food groups and nutrients: Total fat; 
Lower-income households meeting dietary guidelines: 30%; 
Higher-income households meeting dietary guidelines: 34%. 

Food groups and nutrients: Saturated fat; 
Lower-income households meeting dietary guidelines: 33%; 
Higher-income households meeting dietary guidelines: 36%. 

Food groups and nutrients: Sodium; 
Lower-income households meeting dietary guidelines: 25%; 
Higher-income households meeting dietary guidelines: 20%. 

Food groups and nutrients: Calcium; 
Lower-income households meeting dietary guidelines: 39%; 
Higher-income households meeting dietary guidelines: 47%. 

Source: Ronette R. Briefel, “The Changing Consumption Patterrns and 
Health and Nutritional Status in the United States: Evidence from 
National Surveys, ”The Nation’s Nutrition, International Life Sciences 
Institute (Washington, D.C.: 2007). 

Notes: 

Data are from the 1994 to 1996 Continuing Survey of Food Intakes by 
Individuals for fruits, vegetables, grains, whole grains, total fat, 
and saturated fat and from the 1988 to 1994 National Health and 
Nutrition Examination Survey for sodium and calcium. 

Lower-income households have income at or below 130 percent of the 
federal poverty threshold. 

[End of figure] 

Other research has examined how the diets of FSP participants compare 
with the diets of low-income and higher-income individuals not 
participating in the FSP. This research suggests that the diets of FSP 
participants and income-eligible nonparticipants do not differ 
significantly. An ERS-funded study found that FSP participants were 
more likely than higher-income nonparticipants to have poor diets, on 
the basis of an assessment of the intake of grains, vegetables, fruits, 
dairy, meats, fat, cholesterol, and sodium as well as on the variety of 
foods consumed.[Footnote 8] 

Nutrition Education: 

Nutrition education can be defined as any set of learning experiences 
designed to facilitate the voluntary adoption of eating and other 
nutrition-related behaviors that are conducive to health and well- 
being.[Footnote 9] USDA, the lead agency for the nation's nutrition 
education efforts, funds and administers a variety of nutrition 
education efforts. One program, the Expanded Food and Nutrition 
Education Program (EFNEP), is specifically designed to educate low- 
income families and youths about nutrition and nutrition-related 
subjects. EFNEP is administered at the state level by Cooperative 
Extension Service offices, which oversee the allocation of federal 
EFNEP funds. In 2006, USDA reported that EFNEP delivered nutrition 
education to 150,270 adults and 409,389 youths, and that 84 percent of 
the families served were at or below the federal poverty level. 

While there is no legislative mandate for nutrition education, states 
can provide nutrition education to FSP participants and eligible 
nonparticipants, known as Food Stamp Nutrition Education (FSNE). USDA 
reimburses states 50 percent of allowable nutrition education costs. To 
provide nutrition education, the state agency administering the FSP 
submits a state nutrition education plan. About one-half of the 93 
agencies implementing FSNE are affiliated with the Cooperative 
Extension Service, which is the same entity that administers EFNEP. In 
2007, 52 state agencies had approved nutrition education plans for 
which more than $270 million in federal funds were approved. 

Variety of Approaches Can Increase the Consumption of Targeted Foods: 

Recent research describes a variety of approaches to encourage healthy 
eating habits, such as providing financial incentives and nutrition 
education as well as improving access to targeted foods. Studies show 
that financial incentives and nutrition education can increase the 
consumption of targeted foods, although there is less reported research 
on financial incentives. Little is known about the effectiveness of 
access approaches due to a lack of reported research. Study 
limitations, including a lack of data regarding cost and information on 
sustained effects, as well as methodological differences among studies, 
make it difficult to compare approaches or determine which approaches 
are the most effective. 

Recent Research Describes a Variety of Approaches: 

A variety of approaches have been undertaken to increase the purchase 
and consumption of targeted foods, and many combinations of approaches 
have been studied. Financial incentives are used to make the purchase 
and consumption of the targeted foods more affordable and appealing 
than alternative food choices. Examples of financial incentives include 
providing additional unrestricted food or cash benefits; providing 
targeted financial incentives (e.g., price discounts at the POS, 
coupons or vouchers for targeted foods, or financial rewards based on 
the adoption of the desired behavior); or taxing foods on the basis of 
nutrient content. Nutrition education approaches--including nutrition 
classes, counseling, point-of-purchase information, and social 
marketing--attempt to increase individual knowledge and skills (e.g., 
reading food labels and shopping) as well as influence underlying food 
preferences.[Footnote 10] Access approaches focus on increasing the 
availability of targeted foods or limiting the availability of foods 
with low nutritional value. These approaches can vary by setting (e.g., 
grocery stores, schools, workplaces, or communities); scope (e.g., 
individual or larger scale); target population (e.g., healthy or at 
risk of or diagnosed with diet-related disease); and targeted foods. 

The goal of these approaches is to influence one or more of the many 
factors that affect adult food purchasing decisions, such as food 
prices, household income, knowledge about food choices that contribute 
to a healthy diet, and food availability. Taste and convenience also 
influence food purchasing decisions. Figure 3 illustrates the 
theoretical effects of these approaches on household food expenditures, 
eating habits, and health. 

Figure 3: Theoretical Relationships between Nutrition Promotion 
Approaches and Outcomes and Impacts: 

[See PDF for image] 

This figure is an illustration of the theoretical relationships between 
nutrition promotion approaches and outcomes and impacts, as follows: 

Type of Nutrition Promotion Approach: 
* Financial incentives: 
- Increased cash or food benefits; 
- Targeted price reductions/subsidies; 
- Targeted taxes; 
Expected Outcomes and Impacts: 
* Increase food purchasing power; 
* Change relative prices favoring targeted healthy foods; 
* Increased household expenditures on targeted healthy foods; 
* Improved individual dietary intake; 
* Improved weight status and other health outcomes. 

Type of Nutrition Promotion Approach: 
* Nutrition education: 
- Point-of-purchase information (e.g., shelf labels and signs); 
- Nutrition classes and counseling; 
- Social marketing; 
Expected Outcomes and Impacts: 
* Improve nutrition knowledge, attitudes, and skills; 
* Increased household expenditures on targeted healthy foods; 
* Improved individual dietary intake; 
* Improved weight status and other health outcomes. 

Type of Nutrition Promotion Approach: 
* Access: 
- Use EBT at farmers’ markets; 
- Develop new stores; 
- Increase capacity of existing stores; 
- Prohibit use of FSP benefits for certain types of food; 
Expected Outcomes and Impacts: 
* Increase availability of targeted healthy foods; 
* Reduce availability of less healthy foods; 
* Increased household expenditures on targeted healthy foods; 
* Improved individual dietary intake; 
* Improved weight status and other health outcomes. 

Source: GAO. 

[End of figure] 

Studies Show Financial Incentives and Nutrition Education Can 
Positively Affect the Consumption of Targeted Foods, but Little Is 
Known about Efforts to Improve Access: 

Several nutrition promotion studies have demonstrated positive effects 
on the purchase or consumption of targeted foods. However, results vary 
and changes in consumption patterns as a result of approaches may not 
be substantial enough to meet federal dietary guidelines. 

Financial Incentives: 

A review of research published as of April 2005 identified four studies 
of financial incentive programs that were able to compare a group of 
participants that received incentives with a similar group that did not 
in order to identify program impact. All four studies demonstrated 
short-term positive effects on the purchases of targeted foods, 
consumption, or weight loss.[Footnote 11] The studies evaluated the 
provision of WIC farmers' market coupons, price discounts on low-fat 
items offered in school and work-site vending machines, and payments 
contingent on weight loss. For example, an evaluation of a Farmers' 
Market Nutrition Program implemented in a county in Michigan studied 
the following approaches with WIC and Commodity Supplemental Food 
Program participants: education about the use, storage, and nutritional 
value of fruits and vegetables; farmers' market coupons; both education 
and coupons; and no education or coupons.[Footnote 12] The study found 
that after $20 in farmers' market coupons were given to 564 low-income 
women, they self-reported a statistically significant increase in fruit 
and vegetable consumption. This study also found that the maximum 
impact of the intervention on attitudes about fruit and vegetable 
consumption and intake was achieved through a combination of education 
and coupons. Additional evidence from the WIC program suggests that 
providing cash value vouchers for purchasing fruits and vegetables can 
increase the consumption of these items. New York State found that 81 
percent of the $5 checks that were provided to households with children 
in the WIC program for the purchase of fruits and vegetables were 
redeemed.[Footnote 13] Another study evaluated the effect of providing 
$10 vouchers once a week for 6 months for fresh fruit and vegetable 
purchases at a farmers' market or supermarket.[Footnote 14] 
Participants in this study included 602 low-income women from the WIC 
program in Los Angeles. This study found that participants increased 
their consumption of fruits and vegetables and sustained the increase 6 
months after the program ended. The results of this study show that 90 
percent of the coupons were redeemed; farmers' market participants 
showed an increase in the consumption of 1.4 servings per 1,000 
calories of consumed food, and supermarket participants showed an 
increase of 0.8 servings per 1,000 calories of consumed food.[Footnote 
15] 

Economic research suggests that providing low-income households with 
targeted financial incentives may be more effective at increasing fruit 
and vegetable consumption than providing additional unrestricted cash 
or food benefits. In a 2007 study, ERS estimated how spending on fruits 
and vegetables by low-income households would change in response to 
changes in the prices of fruits and vegetables, using statistical 
models and data from the 1987-88 Nationwide Food Consumption Survey. 
[Footnote 16] Specifically, ERS estimated that a 10 percent reduction 
in the price of fruits and vegetables would result in increases in the 
amount purchased, ranging from 6 percent to 7 percent, and would 
increase consumption by the average FSP participant from 1.95 cups per 
day to an estimated 2.08 cups per day. A 20 percent reduction in price 
would raise consumption by about 1/4 cup, bringing total daily 
consumption up to 2.20 cups.[Footnote 17] 

In contrast, another ERS study suggested that a modest increase in 
income would not result in increased spending on fruits and vegetables 
in low-income households. ERS examined the relationship between 
increased income, which could be spent on any food or nonfood item, and 
fruit and vegetable expenditures in low-income households. This study 
found no relationship between a 10 percent increase in income and 
expenditures on fruits, vegetables, eggs, dairy products, or baked 
foods among low-income households.[Footnote 18] However, ERS found a 
statistically significant relationship between the additional income 
and expenditures for beef and frozen entrees and other frozen prepared 
foods among low-income households, possibly because of taste 
preferences and convenience. 

California passed legislation in September 2006 authorizing the Healthy 
Food Purchase Pilot Program, which includes the delivery of financial 
incentives to FSP participants for purchases of fresh produce; however, 
the pilot program has not been implemented. The pilot program aims to 
increase access to fresh produce in selected low-income communities, in 
part by adding a percentage of what FSP participants' spend on fresh 
produce back to their EBT accounts. The law requires the state agency 
to consult with USDA's ERS to design the pilot evaluation. The state 
convened a workgroup made up of state and local health and FSP 
administrators, information systems staff, retailers, and advocacy 
organizations to begin discussions about the pilot. However, the 
California state legislature did not provide funding as part of the 
legislation, and implementation is on hold until funding is made 
available. 

Nutrition Education: 

A review of 92 studies analyzed the results of nutrition education 
approaches, including individual counseling; support groups; and 
classes provided in school-based, work-site, community, and health care 
settings.[Footnote 19] The review compared the results of studies that 
used similar outcome measures and found that most studies reported that 
the approaches had statistically significant effects on consumption, as 
shown in figure 4. For example, more than three quarters of the studies 
reporting changes in fruit and vegetable consumption (17 of 22) found 
significant increases in the consumption of these foods, with an 
average increase of 0.6 servings per day. Similarly, 86 percent of the 
studies reporting results for total fat and 87 percent of the studies 
reporting results for saturated fat found that the approaches had a 
significant effect in reducing fat consumption. Of the studies 
reporting results for total fat, groups receiving nutrition education 
decreased fat intake more than groups that did not receive nutrition 
education, representing an average decrease of 7.3 percent of daily 
calories from fat. According to this review, approaches may be more 
effective if educators help participants set goals and involve small 
groups of participants. 

Figure 4: Number of Studies Reporting Significant Nutrition Education 
Effects: 

[See PDF for image] 

This figure is a stacked vertical bar graph depicting the following 
data: 

Number of Studies Reporting Significant Nutrition Education Effects: 

Outcome: Total fat; 
Number of studies that did not report significant effect: 7; 
Number of studies that reported significant effect: 42. 

Outcome: Saturated fat; 
Number of studies that did not report significant effect: 4; 
Number of studies that reported significant effect: 26. 

Outcome: Fruit and vegetable intake; 
Number of studies that did not report significant effect: 5; 
Number of studies that reported significant effect: 17. 

Source: Alice Ammerman, et al., “The Efficacy of Behavioral 
Interventions to Modify Dietary Fat and Fruit and Vegetable Intake: A 
Review of the Evidence,” Preventive Medicine 35, 25-41 (2002). 

Note: Total fat and saturated fat are measured as either a percentage 
of daily energy intake or in grams per day. Fruit and vegetable intake 
is measured as servings per day or in other units, such as fruit and 
vegetable intake scores. 

[End of figure] 

Another review of 15 community-based nutrition education studies found 
that increased fruit and vegetable consumption was reported in many 
studies.[Footnote 20] According to the authors, the most effective 
approaches gave clear messages, incorporated multiple strategies that 
reinforced the messages, involved the family, were more intensive 
(e.g., multiple contacts and multiple components) and provided over a 
longer period, and were based on a theoretical framework. Another 
review of 44 nutrition education studies suggested that small increases 
in the consumption of fruits and vegetables can be achieved in a 
variety of ways. This review found that fruit and vegetable consumption 
increased when participants were provided with face-to-face education 
or counseling. Less intense approaches, such as telephone contacts and 
computer-generated information tailored to the participant, can also be 
effective.[Footnote 21] Increases in fruit and vegetable consumption 
ranged from 0.1 to 1.4 servings per day in approaches used for healthy 
adults.[Footnote 22] Nutrition education approaches appeared to be more 
successful at positively changing dietary behavior among populations at 
risk of or diagnosed with disease than among general, healthy 
populations, according to two study reviews.[Footnote 23] According to 
the study reviews, these approaches may be more successful because (1) 
higher-risk individuals are more motivated to change and (2) approaches 
with higher-risk populations, such as those offered in health care 
settings, may be more intense than approaches used with lower-risk 
populations in community, workplace, or school settings. 

Providing nutrition information at the point of purchase--such as 
posters, brochures, and shelf and food labels--is another nutrition 
education approach that can positively affect food purchasing behavior, 
although the results from reported research on this approach are mixed. 
In another review of eight studies that used only information 
strategies (e.g., labels and signs) to promote targeted items in a 
grocery store setting, five studies reported increased sales of some of 
the targeted items.[Footnote 24] The grocery store intervention studies 
that showed the greatest changes in behavior lasted for 2 years, 
involved a large supermarket chain, and involved multiple components. 
One of the studies reviewed found significant increases in the sales of 
low-calorie or reduced-calorie foods after a 2-year effort involving 
shelf labeling, pamphlets, and advertisements on television and radio 
and in newspapers. The authors of this review concluded that nutrition 
information approaches and financial incentives appear to be more 
successful in settings where food choices are limited, such as 
workplace cafeterias or school settings, rather than in grocery stores. 
Food labels may also have an effect on dietary intake. In a study 
estimating the effect of the Nutrition Labeling and Education Act of 
1990, which took effect in 1994 and made nutrition labeling mandatory 
for most processed foods, researchers found that the new labels were 
associated with a decrease in body weight and the probability of 
obesity among non-Hispanic white women. The new labels were not 
associated with a statistically significant decrease in body weight or 
the probability of obesity among other groups of women or among men. 
[Footnote 25] 

Access: 

We did not identify any recent studies evaluating the effects of 
approaches designed to improve access to targeted foods, but there have 
been efforts to address the availability of foods that contribute to a 
healthy diet. California's Healthy Food Purchase Pilot Program, which 
was not funded, included proposed assistance for retailers to obtain 
refrigerated produce display cases and technical assistance on the 
purchase, storage, marketing, and display of fresh produce. The FSP 
authorizes farmers' markets to accept FSP benefits, an approach that 
may increase the availability of fresh fruits and vegetables. In 2007, 
532 authorized farmers' markets redeemed $1.6 million in FSP benefits, 
according to FNS. 

Another access strategy that has been proposed but not implemented or 
evaluated is to prohibit FSP participants from using their benefits to 
buy foods with limited nutritional value. This strategy would require 
an FNS-approved waiver of FSP regulations. For example, Minnesota 
proposed prohibiting the purchase of candy and soft drinks with FSP 
benefits. While this strategy has not been implemented, FNS and ERS 
reports have described several challenges to implementing this idea. 
[Footnote 26] For example, FNS reported that food restrictions would 
increase program complexity and cost, since either FNS or food 
manufacturers and producers would need to assess and certify a large 
number of food products that meet new federal standards. ERS reported 
that restricting purchases of "unhealthful" foods and beverages may not 
be a promising strategy for dietary improvement, in part because FSP 
participants may use their own money to buy prohibited items or 
substitute the prohibited foods for other foods similar to the 
prohibited items. 

Lack of Information on Costs and Sustained Effects and Methodological 
Differences among Studies Make It Difficult to Identify the Most 
Effective Approaches: 

The studies reviewed provided little information about the cost or 
sustainability of the nutrition promotion approaches. Five reviews of 
studies found that cost information, such as total costs, cost per 
participant, and the cost-effectiveness of the approach, was not 
reported. In addition, few approaches were shown to be effective over a 
sustained period of time, such as a year or more. For example, one 
review of studies found that many approaches last only a few weeks, and 
studies have short durations, which prevents researchers from 
identifying long-term effects. 

The studies reviewed were designed differently, and they often used 
different outcome measures to evaluate effects, thereby making it 
difficult to compare the results of the approaches. For example, not 
all studies used a control group to compare individuals receiving the 
intervention with similar individuals who did not. Although measures of 
changes during and after an intervention provide useful insights, 
without a control group that does not participate in the intervention, 
it is difficult to directly attribute any changes the researchers 
observed to the intervention. In addition, three reviews of studies 
found that various outcome measures were used, which prevented study 
comparisons and conclusions about the most effective approaches. For 
example, one review found that studies used various self-reported 
dietary assessment tools or sales data to measure changes in 
consumption. A lack of standardized outcome data has also been 
identified as a barrier to evaluating the effectiveness of nutrition 
education activities, according to ERS. One review of studies that was 
able to compare study results found that the dissimilarity across 
studies was a major barrier to drawing comparisons and reaching broad 
conclusions about the effectiveness of approaches. 

The 2008 Farm Bill authorizes funding for pilot projects to evaluate 
health and nutrition promotion in the FSP. The pilot projects may 
include increased benefit allotments to participating households, 
increased access to farmers markets, incentives for authorized 
retailers to increase the availability of healthy foods, stricter 
retailer requirements to stock healthy foods, incentives at the point 
of purchase for targeted foods, and select school-based nutrition 
education strategies. All pilot projects require an independent 
evaluation using rigorous methodologies, particularly random assignment 
or other methods that are capable of producing scientifically valid 
information regarding which activities are effective. 

Factors to Consider Include the Selection of Targeted Foods, Incentive 
Amount, Participant Education, and Monitoring and Evaluation: 

There are several factors that must be taken into consideration in 
designing a program that provides financial incentives for the purchase 
of targeted foods in the FSP. These factors include the selection of 
foods to target for promotion, the incentive amount, how participants 
would be informed of the new benefit, and how the program would be 
monitored and evaluated. 

Selection of Foods to Promote Could Be Controversial and Challenging: 

Selecting the foods an incentive program is designed to promote could 
be controversial. Various interests--such as food manufacturers, 
nutritionists, participant advocacy groups, and states--may have 
differing opinions about which foods should be targeted for promotion 
and which should be excluded. Food manufacturers may want the targeted 
foods to include a wide variety of products because they could see 
financial gain, nutritionists may not agree on which foods are the 
healthiest, and advocacy groups may not want FSP participants to be 
restricted in their food choices. States may also advocate for locally 
grown or produced foods to be targeted for promotion. 

Developing a list of individual food items that either meet certain 
nutritional criteria or contain a particular ingredient, such as whole 
grains, would be challenging because the nutritive content of thousands 
of food items would need to be evaluated to identify the eligible 
items. For example, if low-fat items were targeted, retailers would 
need to identify each of the items sold in their stores that meet that 
criteria. Similarly, targeting all foods that contain whole grains may 
be difficult, given the variety of foods that have whole grains as 
ingredients. 

Expert panelists and many stakeholders we interviewed agreed that 
providing incentives for a general food category that is relatively 
easy to identify--such as fruits and vegetables--would be easier for 
participants and retailers, rather than identifying and selecting 
specific items on a list. However, using a general food category could 
allow some foods with less nutritive value to be included in that 
category. For example, the nutritive value of fruits and vegetables 
varies extensively from one to the other. Offering incentives for 
purchasing all fruits and vegetables would include frozen or canned 
items that may contain added sugar or salt, thereby making them less 
healthy. 

In addition, availability of the targeted foods may be limited in some 
areas, depending on the foods selected. Some stakeholders we 
interviewed said small retailers serving large urban areas may not 
stock certain foods, such as fresh produce. State officials we 
interviewed commented that if fruits and vegetables were the targeted 
foods, they should be eligible in any form--canned, frozen, and fresh-
-because fresh produce may not be equally available throughout the year 
or throughout the country. 

Amount of the Incentive Will Affect Participant Response and Program 
Costs: 

Although financial incentives could take many forms, the amount would 
need to be large enough to motivate FSP participants to change their 
purchasing patterns by buying more of the targeted foods.[Footnote 27] 
The optimal value of an incentive to increase purchases of targeted 
foods is not known at this time, but as we have previously noted, ERS 
estimated in a study that a 10 percent reduction in the price of fruits 
and vegetables would result in a modest increase in fruit and vegetable 
consumption by the average FSP participant, and that a 20 percent price 
reduction would have a slightly larger effect.[Footnote 28] California 
proposed providing a larger incentive as part of its Healthy Food 
Purchase Pilot Program, pending the availability of state funds, equal 
to 30 percent or 40 percent of the amount that participants spend for 
fresh produce. 

The amount of the incentive would also affect the cost of the program. 
ERS estimated that with a FSP caseload of 25.7 million participants at 
the time of the study, the annual cost of a 10 percent price reduction 
would range from $300 million, if the price reduction was restricted to 
fresh produce, to $500 million, if all fruits and vegetables were 
discounted.[Footnote 29] The cost of a financial incentive program 
could be contained somewhat by setting a limit on the amount of 
targeted food purchases that would be eligible for the incentive. The 
ERS study did not estimate how much of the targeted foods participants 
would buy if there were a limit on the amount of fruit and vegetable 
purchases that would be discounted. Depending on the amount of funding 
available for such a program, policymakers may have to choose between 
offering a small and potentially ineffective incentive to a greater 
number of participants and offering a more substantial incentive to a 
smaller number of participants. 

Informing Participants Could Add to Caseworker and Nutrition Educator 
Responsibilities: 

To motivate participants to change their purchasing habits in response 
to financial incentives, participants must first be informed of the 
availability of incentives. Informing participants may increase the 
workload of the local assistance offices. Caseworkers from local 
assistance offices could inform new FSP participants when they apply 
for benefits and inform existing participants when they visit the 
office to reapply for benefits. Providing information through the mail 
is one way to advertise the incentive to existing participants; 
however, processes would need to be in place to correct home addresses 
if mail is returned. Information could also be posted on FNS and state 
FSP Web sites. Other marketing options include advertising in print, 
broadcast, and electronic mediums. Another marketing idea, suggested by 
a few retailers, is to use shelf tags to identify the targeted foods. 
However, use of this strategy would vary by retailer. 

Educating participants on the nutrition and health benefits of buying 
more of the targeted foods could enhance the effectiveness of financial 
incentives, according to some research and stakeholders. An evaluation 
of the effects of providing WIC participants with farmers' market 
coupons found that those who received both coupons and nutrition 
education showed the greatest changes in attitudes about fruit and 
vegetable consumption and intake. Expert panelists said that having an 
education component is crucial to the success of an incentive program. 
In addition, a retailer we interviewed emphasized that if the goal of 
this program is to change purchasing patterns, participants need to 
understand why the program is important to their families and their 
health. Officials from one state also suggested it would be helpful to 
provide nutrition education activities, such as food demonstrations, 
healthy recipe books, cooking classes, and giveaways (e.g., cooking 
utensils). However, the FSNE program does not currently reach all FSP 
participants. Educating all FSP participants would be challenging and 
would likely increase the workload of nutrition educators. 

Decisions about Program Monitoring and Evaluation Would Affect USDA's 
Ability to Protect Program Integrity and Determine the Effect of 
Incentives on Participant Food Choices: 

Program monitoring of day-to-day activities would be necessary to 
ensure that providing financial incentives for targeted foods to FSP 
participants does not erode program integrity. According to FNS 
officials, any new incentive program would pose an increased risk of 
fraud. For example, a FSP participant and retailer could agree to 
fraudulently claim that the participant spent a certain amount of their 
EBT benefits on the targeted foods. In this scenario, the retailer 
would ring up the targeted food purchases, the participant would swipe 
their EBT card for the purchase amount, and the retailer would give the 
participant a portion of the purchase amount in cash and pocket the 
difference. FNS officials suggested that limiting the amount of 
targeted food purchases that would trigger the incentive may reduce the 
risk of fraud and the cost of the incentive program (e.g., participants 
can receive incentives for purchasing up to $50 of the targeted foods 
per month). FNS indicated that it would need to incorporate any new 
program into its existing oversight activities by monitoring the total 
amount spent on the targeted foods and developing criteria to flag 
suspicious transactions. Since individual food purchases cannot be 
monitored through the existing FSP EBT system and USDA has a limited 
number of fraud investigators nationwide, any additional oversight 
responsibility would create a new challenge, according to FNS 
officials. 

Piloting the program prior to full implementation and evaluating the 
impact of the program are important steps to determining its 
effectiveness.[Footnote 30] Two important considerations in evaluating 
a financial incentive program are the research design and how results 
would be measured. 

* Research design: Ideally, impact evaluations allow a reasonable 
comparison of a participant group that received the financial 
incentives with a similar group that did not receive the incentives to 
accurately gauge the effect of the intervention. Appropriate sample 
sizes would need to be determined to allow meaningful comparisons of 
these two groups. However, alternative research designs could be used 
if randomly assigning participants to comparison groups is not 
possible. For example, researchers could use statistical techniques to 
isolate the effect of financial incentives from other factors that 
affect purchasing behavior. One expert panelist commented that 
isolating the impact of this price change from other store promotions 
may be difficult given the many factors involved. Any research design 
would have to take into account an appropriate time frame for the 
pilot, allowing time for retailers to change their operations to be 
able to implement the pilot and for participants to change their 
purchasing habits in response to the incentives. In light of these 
concerns, when ERS was consulted by California on its proposed pilot, 
ERS suggested that the evaluation include a before-and-after review of 
the purchases made by participants and nonparticipants, and that pilot 
counties be randomly selected. 

* Measuring results: Researchers could measure changes in participants' 
food purchases, consumption, or health outcomes to determine the 
effects of a financial incentive program. Measuring health outcomes 
would require tracking participants over an extended period. 
California's legislation authorizing a pilot incentive program requires 
researchers to assess whether the pilot program resulted in FSP 
participants' increased purchases of fresh fruits and vegetables. FNS 
officials suggested that a study could compare FSP participants' 
purchases with federal dietary guideline recommendations before and 
after an incentive program was piloted. If the purpose of the program 
were to encourage a diet that meets federal dietary guidelines and 
improves the health of participants, the effect on participant's 
consumption and health could also be examined. 

Options for measuring changes in food purchasing behavior include 
collecting participant-reported information or collecting food purchase 
records from retailers.[Footnote 31] Understanding whether participants 
increased their purchases of targeted foods would require gathering 
information on how the participants spent their entire food budget, 
since participants may pay for food several ways (e.g., cash, TANF cash 
benefits, EBT, or credit cards) and shop at multiple stores. 
Researchers would need this information to know if more targeted foods 
were being purchased or if participants were substituting the new 
targeted benefits for money they were already spending on those foods. 
Gathering this information may be challenging, depending on how the 
financial incentives are delivered. For example, FNS officials said one 
disadvantage of using paper vouchers is that the redeemed vouchers may 
not be linked to the rest of the family's purchases or household 
demographic information. Using participant-reported information on 
purchases to measure effectiveness can be problematic if participants 
cannot recall the information or if they otherwise misreport purchases. 
Using purchase data to analyze changes in purchases would be more 
accurate but could be difficult to collect and analyze. For example, 
collecting purchase data would likely require researchers to enter into 
data-sharing agreements with multiple retailers, which would need to 
include provisions to protect participants' privacy. Data would need to 
be collected from multiple retailers on all of the food items purchased 
by participants using all payment types. Also, food purchases may not 
accurately reflect consumption. 

Financial Incentives Could Be Implemented Electronically or Using Paper 
Methods, but Each Option Has Potential Implications for Ease of 
Implementation, Program Integrity, and Cost: 

An incentive program--delivered in the form of an additional FSP 
allotment tied to the purchase of certain foods targeted for promotion-
-could be implemented through electronic or paper methods, and each of 
the options would have implications for ease of implementation, program 
integrity, and cost. Electronic options include providing additional 
FSP benefits to the EBT cards currently used by FSP participants or to 
a separate card on the basis of the amount participants spend on the 
foods targeted for promotion. Alternatively, paper options for 
delivering incentives include providing vouchers to be spent on the 
targeted foods or coupons that would discount the prices of those 
foods. 

Electronic Delivery of Incentives to Existing EBT Cards May Be Easy for 
Participants to Use, but Would Require Several Changes to the EBT 
System: 

Incentives provided electronically by using existing EBT cards would 
build on the current checkout technology and process, may not require 
as much time to check out as using a separate card, and could minimize 
the risk of fraud. Online EBT for processing FSP transactions is 
operational nationwide and national standards have been developed to 
promote consistent transaction processing and allow interoperability, 
which means that participants may use their EBT cards at any FNS- 
authorized retailer in any state. According to expert panelists, 
financial incentives could be added to existing EBT cards without 
requiring an additional card swipe and PIN entry, thereby minimizing 
checkout time. Providing incentives electronically on EBT cards could 
also minimize fraud risk by requiring entry of a PIN to complete 
transactions. 

Expert panelists preferred this option and cited its advantages in ease 
of use for participants and program integrity. With this option, the 
retailer would electronically or manually separate and subtotal the 
targeted food purchases and the purchases of other eligible items, and 
the participant would swipe the EBT card and enter the PIN once into 
the POS device, as shown in figure 5. The purchase amount of the 
targeted foods and the purchase amount of other FSP-eligible items 
would be communicated to the EBT contractor who would calculate the 
financial incentive amount and either immediately add it to the 
participant's EBT account or periodically add to the account after the 
purchase, such as at the end of each business day. Participants could 
leave the store with a receipt showing the amount of the financial 
incentive they just received or can expect to receive. 

Figure 5: Incentive Amount Added to Existing EBT Card on the Basis of 
the Amount Spent on Targeted Foods: 

[See PDF for image] 

This figure is an illustration of how an incentive amount is added to 
existing EBT card on the basis of the amount spent on targeted foods, 
as follows: 

One Card: 
Customer-retailer interaction: 
* Retailer rings up $50 worth of groceries, including $10 in “targeted 
foods” and $40 in other eligible items. 
* Customer swipes EBT card and enters PIN to access account, with $10 
of the $50 purchase identified as a targeted foods transaction and $40 
identified as an other eligible foods transaction. 
* Customer receives $50 worth of groceries. 

One EBT food account: Includes targeted food transaction type and other 
eligible items transaction type; 
EBT contractor: 
* State EBT contractor verifies retailer’s license number and 
customer's account number, PIN, and account balance, then authorizes 
transaction. 
* On a real-time or periodic basis, state EBT contractor calculates the 
incentive (e.g., 20 percent) and adds the incentive (e.g., $2) to 
participant’s benefit account. 

Federal Reserve Bank: 
* State EBT contractor orders $50 reimbursement for retailer from FNS’s 
Federal Reserve Bank account. 

Retailer's bank: 
* Retailer is reimbursed $50 as part of the daily settlement process at 
end of business day. 

Sources: GAO, Nova Development Corporation. 

[End of figure] 

This method may be relatively easy for a participant to use, but expert 
panelists agreed that their preferred option would be a significant 
undertaking and would require several changes to the EBT system, as 
described in the following text: 

* Identifying purchases of targeted foods: To identify and subtotal 
purchases of the targeted foods at the POS, retailers with scanning 
systems would need to update the software used on their electronic cash 
registers to include an additional field that would flag particular 
items as the targeted foods eligible for incentives. Programming the 
additional field would enable the cash register system to automatically 
ring up and subtotal the targeted food purchase when the items' product 
codes are scanned. Maintaining an updated list of the eligible items 
could be challenging because UPC and PLU codes can change when new 
products are added or certain foods are included in special promotions. 
Also, fresh produce vendors use varying PLU codes for the same fruits 
and vegetables. Smaller retailers without scanning technology and a 
computerized price file system would need to manually sort out and 
subtotal up to three piles of food: one for the targeted foods that are 
eligible for the incentive, one for other FSP-eligible items, and one 
for non-FSP-eligible items. 

Successfully implementing an incentive program depends in part on the 
level of guidance retailers receive regarding foods that qualify for 
the incentives. Generally, there are two ways that retailers could 
identify which foods sold in their stores are the targeted foods 
eligible for incentives: that is, (1) FNS or states could develop a 
detailed list of the targeted foods and associated product codes or (2) 
FNS or states could provide a broad definition of the targeted foods 
and retailers could decide which of the foods in their stores would 
qualify. Retailer representatives and state officials we interviewed 
preferred that the foods eligible for incentives be defined at the 
federal level to assist retailers operating in multiple states and to 
minimize the burden on states. Retailer representatives on the expert 
panel said it would help retailers implement the program if FNS 
provided a detailed list of the targeted foods and associated product 
codes. However, FNS officials on the panel said FNS lacks the resources 
to develop and maintain a detailed list and would prefer to provide a 
more general definition of the foods targeted for promotion. 

* Tracking purchases of targeted foods: The total amount spent on the 
targeted foods must be tracked separately from other purchases to 
calculate and apply the incentives to participants' EBT accounts. To 
process the transactions, expert panelists agreed that the total amount 
spent on the targeted foods should be communicated through the EBT 
system, rather than requiring retailers to complete additional 
reporting. A few retailers commented that any ongoing reporting on 
individual purchases in addition to the reporting that takes place 
during EBT transactions at the POS would be burdensome, especially for 
smaller retailers that may not have detailed accounting procedures. 
Tracking the total amount spent on the targeted foods separately from 
other FSP-eligible items would require software modifications to 
retailers' POS devices. 

* Calculating and applying the incentive: To apply incentives to 
participants' EBT accounts, EBT contractors would need to develop 
software for calculating and applying the incentives and to modify 
their software to include a new transaction type that tracks the amount 
spent on the targeted foods. The interface between the EBT contractors' 
systems and state information systems that contain benefit issuance 
data may need to be modified. EBT contractors would also need to modify 
their retailer and participant complaint processes, modify automated 
telephone systems, and train customer service representatives to 
support the new benefit. 

Modifying and updating the retailer cash register systems and POS 
devices could be challenging with different types of equipment and 
varying levels of technical support. Smaller independent grocery stores 
that install and maintain their own equipment may face significant 
challenges in making technical changes if they do not have ongoing 
technical support. Some smaller retailers with outdated equipment may 
need new equipment to use updated software. However, supermarkets, 
which account for about 85 percent of FSP redemptions nationwide, may 
not experience as much difficulty in modifying their systems because of 
in-house information technology support. Table 2 summarizes the changes 
that would need to be made by each affected group to implement the 
option preferred by expert panelists. Some stakeholders, such as third- 
party processors that help process EBT transactions, may also need to 
make operational changes to implement this option. This broad list of 
required changes points to the types of costs involved in developing an 
incentive program. Expert panelists did not provide information on how 
much it would cost to implement these changes because the costs would 
vary, depending on the many program design decisions that would need to 
be made; however, the costs could be significant. FNS officials said 
that they believe this option may be more costly to implement than a 
paper voucher system. 

Table 2: Required Changes to Implement the Option Preferred by Expert 
Panelists: 

Stakeholder: FNS; 
Required changes, by role: 
* Monitor whether eligible foods were purchased and if the bonus is 
being calculated and applied properly; 
* Modify retailer monitoring systems to identify potential fraud; 
* Develop and impose regulations on both system standards and client 
and retailer sanctions related to administration of this program; 
* Provide information and training to federal, state, and local 
investigative staff and participating retailers. 

Stakeholder: States; 
Required changes, by role: 
* Modify the state information system to include up-to-date information 
on the amount of benefits participants have in their EBT accounts, if 
needed. 

Stakeholder: EBT contractors; 
Required changes, by role: 
* Develop EBT transaction standards to ensure that the transactions are 
implemented the same way across the country; 
* Develop software to calculate and apply the incentive for online and 
manual transactions; 
* Revise manuals and procedures; 
* Modify claims and error processes; 
* Modify the process for how to handle cross-state transactions (one 
EBT contractor serves the state where the participant has been provided 
benefits, while another EBT contractor serves a different state where 
the purchase occurred); 
* Modify retailer and customer complaint processes. Update customer 
support systems to support questions about this new benefit. Modify 
automated telephone systems and train customer service representatives; 
* Modify the software for the POS devices that the EBT contractors 
provide to retailers on behalf of the states, to include an additional 
targeted foods transaction field. Install the updated software and 
replace outdated devices. 

Stakeholder: Retailers; 
Required changes, by role: 
* Update POS device software to include a new targeted foods 
transaction type; 
* Modify software used on the electronic cash register systems to 
include an additional flag to allow targeted foods to be identified and 
subtotaled so that the subtotal of targeted foods can either 
automatically be communicated to the POS device or manually entered in 
by the retailer; 
* Flag the targeted foods in the information systems, and keep the 
database up to date, as with identifying WIC-or FSP-eligible items; 
* Develop new procedures for cashiers, and train cashiers on the new 
system. 

Source: GAO. 

[End of table] 

Incentives Could Be Delivered Electronically to Separate Cards, but 
Implementation Could Add Cost and Complexity: 

Several types of cards separate from participants' existing EBT cards 
could be used to deliver financial incentives.[Footnote 32] Two 
alternatives include issuing participants a new magnetic strip or smart 
EBT card that would be used for purchases of the targeted foods. 
[Footnote 33] Under this type of system, a FSP participant would 
receive additional benefits in a new targeted foods account and be 
issued a new card to access the benefits. At check out, the participant 
would swipe the new card and enter a PIN to purchase the targeted foods 
and swipe their existing EBT card and enter that PIN for the other FSP- 
eligible items. An amount equal to a percentage of what a participant 
spends on the targeted foods could be added to a targeted food account, 
as shown in figure 6. 

Figure 6: Incentive Amount Added to a Separate Targeted Foods Card on 
the Basis of the Amount Spent on Targeted Foods: 

[See PDF for image] 

This figure is an illustration of how an incentive amount is added to a 
separate targeted foods card on the basis of the amount spent on 
targeted foods, as follows: 

Two Cards: 
Customer-retailer interaction: 
* Retailer rings up $50 worth of groceries, including $10 in “targeted 
foods” and $40 in other eligible items. 
* Customer swipes the regular EBT card and enters PIN for the $40 in 
other eligible items. 
* Customer swipes the “targeted foods” card and enters PIN to access 
account. 
* Customer receives $50 worth of groceries. 

Two accounts: Includes targeted food account and other eligible items 
account; 
EBT contractor: 
* State EBT contractor verifies retailer’s license number and 
customer's account number, PIN, and account balance, then authorizes 
transaction. 
* Customer receives a $2 incentive added to their “targeted foods” 
account ($2 is a 20 percent incentive for spending $10 on “targeted 
foods”). 

Federal Reserve Bank/Retailer's bank: 
* Retailer reimbursed for $50 as part of the daily settlement process 
at end of business day. The state EBT contractor orders reimbursement 
for retailer from FNS’s Federal Reserve Bank account. 

Sources: GAO, Nova Development Corporation. 

[End of figure] 

A few stakeholders we interviewed suggested using smart cards to 
deliver incentives because these cards have the capacity to record and 
track purchases of individual items, which would be useful for 
monitoring and evaluation purposes.[Footnote 34] The computer chip on 
the smart card could contain the product codes of the specific targeted 
items. The grocery store clerk would scan the items at check out, and 
the system could match each item's product code to a list of approved 
targeted food product codes. This system would help ensure that only 
authorized food items are purchased and receive the financial 
incentives. The smart card transaction would require participants to 
enter a PIN, and the card could calculate and apply the incentive 
amount. 

Using separate cards to deliver incentives would be a more costly and 
complicated option than using existing EBT cards, according to expert 
panelists and some stakeholders we interviewed. State administrative 
costs could increase to purchase and replace the new cards, 
particularly smart cards, which cost more than magnetic strip EBT 
cards. Using a separate magnetic strip EBT card would require the same 
technical changes as using existing EBT cards; however, using smart 
cards would require greater changes to the EBT system. Currently, only 
the total amount spent on FSP-eligible items is tracked through the EBT 
system. Changing the FSP EBT system to one that communicates 
information on the individual items purchased would require significant 
and costly modifications to retailer and EBT contractor information 
systems. In addition, using a separate card may be more complicated for 
participants and grocery store clerks to implement and add time at the 
register to process the additional transactions, according to some 
stakeholders we interviewed. 

Paper Delivery of Incentives Would Not Require Changes to the EBT 
System, but Could Increase State Administrative Costs and Fraud Risk: 

Delivering incentives using paper methods, such as vouchers or coupons, 
may be easier and less costly for some retailers to pilot because these 
methods would not require changes to the EBT system, according to a few 
of the retailers we interviewed. For example, one retailer suggested 
that the state provide participants with paper coupons that would 
discount the prices of certain foods because it would eliminate the 
need for retailers to modify their checkout systems, and retailers 
could use a system similar to the current process for redeeming 
manufacturers' coupons. A retailer representative that works with small 
grocery stores suggested that the easiest and least expensive method to 
pilot an incentive program would be for the state to provide 
participants with the coupons and reimburse retailers when the coupons 
are redeemed, since none of the EBT-related systems would need to be 
changed. Figure 7 illustrates one way that financial incentives could 
be delivered using paper vouchers. In this example, participants are 
provided with paper vouchers valued at $10, which are to be spent on 
the targeted foods, rather than with a bonus that is based on the 
amount spent on the targeted foods. 

Figure 7: Incentive Provided to Participants Using Paper Vouchers as a 
Fixed Amount to Be Spent on Targeted Foods: 

[See PDF for image] 

This figure is an illustration of how an incentive amount is provided 
to participants using paper vouchers as a fixed amount to be spent on 
targeted foods, as follows: 

One card and paper vouchers: 
Customer-retailer interaction: 
* Retailer rings up $50 worth of groceries, including $10 in “targeted 
foods” and $40 in other eligible items. 
* Customer pays for the targeted foods with the vouchers received from 
the state. 
* Customer pays for other eligible items with their EBT card. 
* Customer receives $50 worth of groceries. 
* The retailer mails redeemed vouchers to state or EBT contractor, who 
tracks and reports on the amount of the vouchers used by participant. 

Two accounts: Includes targeted food account and other eligible items 
account; 
EBT contractor: 
* For EBT: The state EBT contractor verifies retailer’s license number 
and customer’s account number and PIN, and account balance, then 
authorizes transaction. 
* For vouchers: The state or EBT contractor collects vouchers and 
orders reimbursement for vouchers. 

Federal Reserve Bank/Retailer's bank: 
* The state EBT contractor orders reimbursement for purchases made on 
EBT card for retailer from FNS’s Federal Reserve Bank account. 

Sources: GAO, Nova Development Corporation. 

[End of figure] 

Paper vouchers have been used successfully to facilitate the purchase 
of fruits and vegetables in other programs. FNS's farmers' market 
programs and the WIC fruit and vegetable pilot program provided paper 
vouchers to WIC participants and elderly individuals so that they could 
purchase fruits and vegetables. For example, California piloted a 
program that provided vouchers for fruit and vegetable purchases for 
low-income women participating in the WIC program. A study evaluating 
the California WIC fruit and vegetable pilot reported that low-income 
women used the supplement almost fully and purchased a wide variety of 
fresh fruits and vegetables with the vouchers. Stakeholders we 
interviewed also identified projects funded with nonfederal dollars 
that provide supplemental coupons to FSP participants to encourage them 
to purchase fresh produce at farmers markets. For example, New York 
City's Department of Health and Mental Hygiene is partnering with 
farmers' markets to distribute Health Bucks--$2 coupons for fresh 
produce--to EBT customers who spend $5 in local produce at the markets. 

However, there are potential disadvantages to using paper methods to 
deliver incentives. Participants might find paper options more 
burdensome because they cannot carry over balances and may have to 
spend the entire amount of the voucher in one shopping trip. In 
addition, participants may not use the financial incentives if they 
perceive that a stigma is associated with using paper vouchers at the 
grocery store. Although paper methods would spare retailers from making 
software modifications, some of the burden would be shifted to states 
to develop systems to collect the coupons and reimburse retailers, 
which could increase state administrative costs. Although any incentive 
program could affect fraud risk, expert panelists and stakeholders we 
interviewed said paper options could increase the risk of fraud because 
coupons can be exchanged for cash or used by ineligible individuals 
more easily. 

Concluding Observations: 

Changing individuals' eating habits is difficult, given the many 
factors that influence food choices. However, the pilot projects 
authorized in the recent Farm Bill provide FNS with a unique 
opportunity to test whether financial incentives would help low-income 
Americans purchase and consume more of the foods that contribute to a 
healthy diet. Although research has found that financial incentives and 
nutrition education both show promise, not enough is known about the 
costs and long-term effectiveness of these approaches, either alone or 
in combination. Moreover, the financial incentives studied included 
coupons. Incentives using the current EBT technology, although 
technically feasible, are largely untested at this point. Also, the 
success of efforts to improve dietary intake may hinge in part on the 
availability of targeted foods, which may be limited in certain areas, 
such as small stores in large urban areas. Little is known about 
efforts to increase access to healthy foods. Only by testing these 
approaches, either alone or in combination, can the costs of such a 
program be estimated. Depending on the approaches selected for testing, 
estimates would need to include the costs of developing and 
administering the program, changing retailer checkout systems, training 
employees, educating participants, and providing increased benefit 
amounts. Program outcomes, such as changes to participant purchasing 
and consumption patterns, will also be challenging to assess and will 
not be useful unless similar outcome measures are developed that can be 
used across different projects to allow comparison. Evaluating the long-
term effect of the changes on the health of program participants will 
likely prove to be even more difficult to assess. However, given the 
evidence linking poor nutrition to adverse health outcomes, finding 
cost-effective approaches to improve nutritional intake is important. 
Unless careful attention is paid to estimating both program and 
administrative costs as well as the effects, Congress and USDA will not 
have the information required to judge whether the benefits outweigh 
the costs. 

Agency Comments and Our Evaluation: 

We provided a draft of this report to the U.S. Department of 
Agriculture for review and comment. On June 20, 2008, we met with FNS 
officials and an ERS official to obtain their comments. In general, the 
officials agreed with our findings and concluding observations. 
However, the officials made some suggestions to enhance and clarify our 
report findings, which we incorporated. First, FNS officials suggested 
that we further emphasize the potential for fraud inherent in a 
financial incentive program and the challenge FNS believes it would 
face in incorporating any additional oversight responsibilities. FNS 
officials believe that any new incentive program that involves the 
issuance of an additional benefit once targeted foods are purchased 
would open up a new avenue for food stamp trafficking. Second, FNS 
officials believe that the preferred option for delivering financial 
incentives through the EBT system would be costly to implement, perhaps 
even more costly than a paper voucher system. We believe that each 
option for delivering the incentive would have cost implications; 
however, without piloting such a program, it is not clear which 
delivery method would be most costly. Third, FNS officials suggested 
that we incorporate additional references to FNS reports that directly 
relate to the topic of our report. Finally, FNS officials suggested we 
incorporate full information about the assumptions and limitations of 
ERS estimates of the cost of delivering financial incentives to FSP 
participants on the basis of their purchases of fruits and vegetables. 
FNS officials identified the following issues that may affect the 
accuracy of the estimates: the age of the data, assumptions about 
participants' purchasing decisions, recent increases in food prices, 
continuing increases in FSP participation, and whether the program 
would affect FSP participation. We acknowledged the limitations of the 
ERS estimates in the report. FNS also provided us with technical 
comments, which we incorporated where appropriate. 

We are sending copies of this report to the appropriate congressional 
committees, the Secretary of Agriculture, and other interested parties. 
We will also make copies available to others upon request. In addition, 
the report will be available at no charge on GAO's Web site at 
[hyperlink http://www.gao.gov]. 

If you or your staff have any questions regarding this report, please 
contact me at (202) 512-7215 or brownke@gao.gov. Contact points for our 
Offices of Congressional Relations and Public Affairs may be found on 
the last page of this report. GAO staff who made major contributors to 
this report are listed in appendix II. 

Sincerely yours, 

Signed by: 

Kay Brown: 
Director, Education, Workforce, and Income Security Issues: 

[End of section] 

Appendix I: Objectives, Scope, and Methodology: 

The objectives of this report were to identify (1) what is known about 
the effectiveness of financial incentives and other approaches intended 
to increase purchases of targeted foods that could contribute to a 
healthy diet; (2) the key factors to consider in designing a program 
that provides Food Stamp Program (FSP) participants with financial 
incentives to purchase certain foods; and (3) options available to the 
U.S. Department of Agriculture's (USDA) Food and Nutrition Service 
(FNS) for implementing financial incentives, and the advantages and 
challenges involved in implementing such options. 

To address the first objective, we searched relevant databases, such as 
ProQuest, National Technical Information Service, Wilson Social Science 
Abstracts, and Sociological Abstracts. We also consulted with USDA's 
Economic Research Service (ERS) and FNS staff. We limited the scope of 
our work by looking at studies published since 2000. Through this 
process, we identified more than 100 studies published from January 
2000 through February 2008. We further narrowed the scope of our work 
to those findings from studies published as reviews or summaries of 
original research, ERS-published research, and studies of Special 
Supplemental Nutrition Program for Women, Infants, and Children (WIC) 
farmers' market programs. We conducted detailed reviews of the 16 
studies that met these criteria. Therefore, our results are not an 
exhaustive or historical treatment of the topic. The studies reviewed 
included both randomized controlled trials and nonrandomized controlled 
trials. Our reviews entailed an assessment of each study's research 
methodology, including its data quality, research design, and analytic 
techniques, as well as a summary of each study's major findings and 
conclusions. We also assessed the extent to which each study's data and 
methods support its findings and conclusions. 

To address the second and third objectives, we interviewed 
representatives of stakeholder groups that would be affected by 
implementation of an incentive program, including USDA officials; 
officials from four states with WIC Electronic Benefit Transfer (EBT) 
(Ohio, Texas, Washington, and Wyoming); officials from one state 
(California) that passed legislation authorizing the provision of 
financial incentives to FSP participants for purchasing targeted foods; 
the three EBT contractors serving almost all of the states; retailer 
associations and retailers representing large chain and smaller 
independent grocery stores; food and nutrition advocacy groups; and 
researchers. We also facilitated group discussions about the options 
for delivering incentives and implementation issues with state 
officials, retailers, and industry representatives at three national 
conferences. 

To further explore targeted food incentive options and implementation 
issues, we convened a panel of 17 experts representing USDA, states, 
retailers, EBT contractors, and manufacturers of retailer check out 
systems. On November 6, 2007, we held a half-day meeting with these 
panelists at our headquarters office in Washington, D.C. Before the day 
of the meeting, we provided each panel member with a set of discussion 
questions that covered specific topics, including the definition of 
foods targeted for promotion, incentive delivery options, tracking 
purchases, applying the incentive, and monitoring and evaluation. Each 
panelist was provided with time to respond to each question and discuss 
the impact of other panelists' responses on the panelist's stakeholder 
group interests. We transcribed responses and made verifications to 
ensure that we had accurately captured panel member statements. We 
identified the panelists through external parties who work on the 
issues covered in this report. We selected individuals who represent 
the different stakeholder groups that would likely be involved in 
developing a targeted foods incentive program for the FSP, if such a 
program was initiated. The panel included representatives from USDA; 
two states (Maryland and Washington); four retailers or retailer 
associations (Krasdale Foods, Kroger, National Grocers Association, and 
Safeway); the three EBT contractors serving almost all states 
(Affiliated Computer Systems, eFunds, and JPMorganChase); and three 
other industry representatives (Electronic Funds Transfer Association, 
IBM, and Verifone). 

We conducted this performance audit from May 2007 through July 2008 in 
accordance with generally accepted government auditing standards. Those 
standards require that we plan and perform the audit to obtain 
sufficient, appropriate evidence to provide a reasonable basis for our 
findings and conclusions based on our audit objectives. We believe that 
the evidence obtained provides a reasonable basis for our findings and 
conclusions based on our audit objectives. 

[End of section] 

Appendix II: GAO Contact and Staff Acknowledgments: 

GAO Contact: 

Kay Brown, (202) 512-7215, brownke@gao.gov: 

Staff Acknowledgments: 

In addition to the contact named above, individuals making key 
contributions to this report include Kathy Larin (Assistant Director), 
Susannah Compton, Neil Doherty, Kevin Jackson, Avani Locke, Ashley 
McCall, Luann Moy, Andrew Nelson, Susan Offutt, Barbara Oliver, Scott 
Pettis, Cathy Roark, Max Sawicky, Daniel Schwimer, Jay Smale, Rosemary 
Torres-Lerma, and Kate van Gelder. 

[End of section] 

Related GAO Products: 

Electronic Payments: Many Programs Electronically Disburse Federal 
Benefits, and More Outreach Could Increase Use. [hyperlink, 
http://www.gao.gov/cgi-bin/getrpt?GAO-08-645]. Washington, D.C.: June 
23, 2008. 

Food Stamp Program: Use of Alternative Methods to Apply for and 
Maintain Benefits Could Be Enhanced by Additional Evaluation and 
Information on Promising Practices. [hyperlink, http://www.gao.gov/cgi-
bin/getrpt?GAO-07-573]. Washington, D.C.: May 3, 2007. 

Food Stamp Program: FNS Could Improve Guidance and Monitoring to Help 
Ensure Appropriate Use of Noncash Categorical Eligibility. [hyperlink, 
http://www.gao.gov/cgi-bin/getrpt?GAO-07-465]. Washington, D.C.: March 
28, 2007. 

Food Stamp Program: Payment Errors and Trafficking Have Declined 
despite Increased Program Participation. [hyperlink, 
http://www.gao.gov/cgi-bin/getrpt?GAO-07-422T]. Washington, D.C.: 
January 31, 2007. 

Food Stamp Trafficking: FNS Could Enhance Program Integrity by Better 
Targeting Stores Likely to Traffic and Increasing Penalties. 
[hyperlink, http://www.gao.gov/cgi-bin/getrpt?GAO-07-53]. Washington, 
D.C.: October 13, 2006. 

Food Stamp Program: States Have Made Progress Reducing Payment Errors, 
and Further Challenges Remain. [hyperlink, http://www.gao.gov/cgi-
bin/getrpt?GAO-05-245]. Washington, D.C.: May 5, 2005. 

Food Stamp Program: Farm Bill Options Ease Administrative Burden, but 
Opportunities Exist to Streamline Participant Reporting Rules among 
Programs. [hyperlink, http://www.gao.gov/cgi-bin/getrpt?GAO-04-916]. 
Washington, D.C.: September 16, 2004. 

Nutrition Education: USDA Provides Services through Multiple Programs, 
but Stronger Linkages among Efforts Are Needed. [hyperlink, 
http://www.gao.gov/cgi-bin/getrpt?GAO-04-528]. Washington, D.C.: April 
27, 2004. 

Food Stamp Program: Steps Have Been Taken to Increase Participation of 
Working Families, but Better Tracking of Efforts Is Needed. [hyperlink, 
http://www.gao.gov/cgi-bin/getrpt?GAO-04-346]. Washington, D.C.: March 
5, 2004. 

[End of section] 

Footnotes: 

[1] The act also includes a provision changing the name of the Food 
Stamp Program to the Supplemental Nutrition Assistance Program, 
effective October 1, 2008. 

[2] Current regulations define "eligible foods" as any food or food 
product intended for human consumption, except alcoholic beverages, 
tobacco, and hot foods and hot food products prepared for immediate 
consumption. Eligible foods are further defined as seeds and plants 
that are used to grow food for the personal consumption of eligible 
households, and meals prepared and delivered or served to eligible food 
stamp recipients as well as equipment for hunting and fishing in 
certain specified areas in Alaska (7 C.F.R. § 271.2). 

[3] TANF is a federal block grant program that provides federal funding 
to states and tribes to operate programs that deliver cash assistance 
and services, such as transportation and child care assistance, to 
needy families. 

[4] For more information on the use of EBT and other electronic means 
of distributing government benefits, see GAO, Electronic Payments: Many 
Programs Electronically Disburse Federal Benefits, and More Outreach 
Could Increase Use, GAO-08-645 (Washington, D.C.: June 23, 2008). 

[5] Bar codes on most prepackaged items adhere to a UPC format that 
enables retailers to uniquely identify the items. Food items with UPC 
codes are prepackaged, with a fixed weight, count, or volume, and are 
distributed by national suppliers or large regional suppliers. 
Retailers with scanning systems can also create their own unique bar 
codes for non-UPC items, referred to as PLU codes. Retailers use PLU 
codes for foods supplied by local independent suppliers and foods that 
are purchased by weight, such as fresh produce. 

[6] FNS defines a "supermarket" as a store with between $2 million and 
$5 million of annual retail sales and, typically, 10 or more checkout 
lanes with cash registers and bar code scanners. A "superstore" is a 
very large supermarket with $5 million or more of annual retail sales. 
Both supermarkets and superstores have staple food sales greater than 
or equal to 65 percent of their retail food sales. 

[7] The figure 2 data on the consumption of fruits, vegetables, grains, 
and fat reflect the percentage of the population meeting USDA's 2000 
Dietary Guidelines for Americans, which has since been revised. Sodium 
consumption data reflect the percentage of the population meeting a 
standard tracked by Healthy People 2010, a project managed by the 
Department of Health and Human Services. Calcium consumption data 
reflect the percentage of the population meeting the recommended 
adequate intakes for calcium developed by the Institute of Medicine of 
the National Academy of Sciences. 

[8] Mary Kay Fox and Nancy Cole, Nutrition and Health Characteristics 
of Low-Income Populations, Volume 1, Food Stamp Program Participants 
and Nonparticipants, U.S. Department of Agriculture, Economic Research 
Service, (Alexandria, Va.: December 2004). An analysis of more recent 
data to be published by FNS will report similar findings, according to 
FNS officials. The forthcoming FNS report is entitled Diet Quality of 
Americans by Food Stamp Program Participation: Data from the National 
Health and Nutrition Examination Survey - 1999-2004. 

[9] For more information on USDA nutrition education programs, see GAO, 
Nutrition Education: USDA Provides Services Through Multiple Programs, 
but Stronger Linkages Among Efforts Are Needed, [hyperlink, 
http://www.gao.gov/cgi-bin/getrpt?GAO-04-528] (Washington, D.C.: Apr. 
27, 2004). 

[10] Social marketing is a private sector marketing model that can be 
adapted to social services, which often makes use of television, radio 
ads, videos, and brochures. These materials by themselves do not 
constitute social marketing; rather, social marketing entails a 
comprehensive program in which these materials are employed as part of 
tactics to reach a target audience. Social marketing also emphasizes 
the importance of keeping the target audience and network partners 
involved in needs assessment, message development, and refinement of 
messages and delivery strategies. 

[11] Joanne Wall, et al., "Effectiveness of Monetary Incentives in 
Modifying Dietary Behavior: A Review," Nutrition Reviews (Health 
Module: 2006), 64, 12. These randomized controlled trials involved 
random assignment of study participants to two or more interventions, 
including a standard of comparison or control, and a comparison of 
outcomes after the participants received the interventions. For 
example, participants could be randomly selected to receive financial 
incentives and compared with an appropriate control group that did not 
receive the incentives. 

[12] See the following citation for the original study: Judith V. 
Anderson, et al., "5 a Day Fruit and Vegetable Intervention Improves 
Consumption in a Low Income Population," Journal of the American 
Dietary Association (2001), 101(2). The Commodity Supplemental Food 
Program provides food and administrative funds to supplement the diets 
of low-income pregnant, postpartum, and breastfeeding women; infants; 
children up to aged 6 years; and individuals over aged 60 years. 

[13] The Bureau of Supplemental Food Programs Food Delivery Systems 
Unit, New York State Department of Health, New York State WIC Program 
2006 Vegetable and Fruit Demonstration Project (February 2007). 

[14] Dena R. Herman, et al., "Effect of a Targeted Subsidy on Intake of 
Fruits and Vegetables Among Low-Income Women in the Special 
Supplemental Nutrition Program for Women, Infants, and Children," 
American Journal of Public Health (2008), 98. For more information, see 
an earlier evaluation of the same approach: Dena R. Herman, et al., 
"Choices Made by Low Income Women Provided with Economic Supplement for 
Fresh Fruits and Vegetables," Journal of the American Dietetic 
Association (2006), 106: 740-744. 

[15] The 2005 Dietary Guidelines for Americans recommends 2 cups of 
fruit (4 servings) and 2½ cups of vegetables (5 servings) per day for 
individuals consuming 2,000 calories per day. A serving of vegetables 
is 1 cup of raw leafy vegetables, ½ cup of cooked vegetables, or ½ cup 
of vegetable juice. A serving of fruit is 1 medium-sized fruit; ¼ cup 
of dried fruit; a ½ cup of fresh, frozen, or canned fruit; or ½ cup of 
fruit juice. 

[16] Joanne F. Guthrie, et al., Can Food Stamps Do More To Improve Food 
Choices? An Economic Perspective, U.S. Department of Agriculture, 
Economic Research Service (Alexandria, Va.: September 2007). To develop 
the estimates, ERS applied the estimated rate of change in fruit and 
vegetable expenditures in response to the price changes to what 
households in the poorest one-fifth of the population spent per person 
on fruits and vegetables in 2004, using data collected by the Bureau of 
Labor Statistics' Consumer Expenditure Survey. 

[17] Given the age of the data used and the assumptions underlying the 
analysis, the magnitudes of the estimates of how low-income households 
would respond to fruit and vegetable price reductions may be 
inaccurate, either understated or overstated. Food stamp participants 
and policies have undergone changes since 1988, which may influence the 
results of this type of analysis. In addition, no single study can 
provide great confidence in the application of its conclusions to a 
policy question. ERS has replicated this analysis with more recent 
data; however, the findings were not published prior to us completing 
our work. 

[18] Hayden Stewart and Noel Blisard, Are Lower Income Households 
Willing and Able to Budget for Fruits and Vegetables?, U.S. Department 
of Agriculture, Economic Research Service (Alexandria, Va.: January 
2008). 

[19] Alice Ammerman, et al., "The Efficacy of Behavioral Interventions 
to Modify Dietary Fat and Fruit and Vegetable Intake: A Review of the 
Evidence," Preventive Medicine (2002), 35 (1). For more details about 
this study, see Alice Ammerman, et al., Efficacy of interventions to 
modify dietary behavior related to cancer risk. Evidence Report/ 
Technology Assessment No. 25, Publication No. 01-E029 (Rockville, Md.: 
February 2001). 

[20] Donna Ciliska, et al., "Effectiveness of Community-Based 
Interventions to Increase Fruit and Vegetable Consumption," Journal of 
Nutrition Education, Health Module (2000), 32, 6. 

[21] FNS developed a report reviewing the research on the effectiveness 
of computer-tailored nutrition education. See Dawn Aldridge, 
Interactive Computer-Tailored Nutrition Education, U.S. Department of 
Agriculture, Food and Nutrition Service (Alexandria, Va.: 2006). 

[22] Joceline Pomerleau, et al., "Interventions Designed to Increase 
Adult Fruit and Vegetable Intake Can Be Effective: A Systematic Review 
of the Literature," Journal of Nutrition (2005), 135 (10). 

[23] Another study we reviewed found that approaches involving 
individual or group counseling were successful in changing dietary 
behaviors and most were conducted with individuals with existing risk 
factors for heart disease. See Deborah Bowen and Shirley Beresford, 
"Dietary Interventions to Prevent Disease," Annual Review of Public 
Health (2002), 23:255-86. 

[24] Jennifer D. Seymour, et al., "Impact of nutrition environmental 
interventions on point-of-purchase behavior in adults: A Review," 
Preventive Medicine (2004), Vol. 39. For more information about 
approaches implemented in a grocery store setting, see Karen Glanz and 
Amy Yaroch, "Strategies for increasing fruit and vegetable intake in 
grocery stores and communities: policy, pricing, and environmental 
change," Preventive Medicine (2004), Vol. 39. 

[25] J.N. Variyam and John Cawley, "Nutrition Labels and Obesity," 
National Bureau of Economic Research, Working Paper No. 11956 
(Cambridge, Massachusetts: 2006). 

[26] For more information, see U.S. Department of Agriculture, Food and 
Nutrition Service, Implications of Restricting the Use of Food Stamp 
Benefits (Alexandria, Va.: March 2007). 

[27] Similarly, research suggests that disincentives must be 
substantial enough to motivate individuals to avoid purchasing foods 
with low nutritional value. An ERS study found that adding a relatively 
low tax on salty snacks of 1 cent per pound and 1 percent of value 
would not appreciably reduce the consumption of salty snacks and would 
have little effect on diet quality or health outcomes. For more 
information, see Fred Kuchler, et al., Taxing Snack Foods: What to 
Expect for Diet and Tax Revenues, U.S. Department of Agriculture, 
Economic Research Service (Alexandria, Va.: August 2004). 

[28] ERS estimates indicate that a 10 percent reduction in the price of 
fruits and vegetables would increase the amount purchased by 6 to 7 
percent. Fruit and vegetable consumption of the average food stamp 
participant is estimated at 1.95 cups per day. A 10 percent reduction 
in fruit and vegetable prices, therefore, would raise consumption to an 
estimated 2.08 cups per day. A 20 percent reduction in the price of 
fruits and vegetables would raise fruit and vegetable consumption to 
2.20 cups per day. 

[29] As we have previously stated, given the age of the data used and 
the assumptions underlying the analysis, the magnitudes of the 
estimates may be inaccurate, either understated or overstated. 

[30] FNS developed a document that describes issues to consider when 
evaluating nutrition education approaches. See U.S. Department of 
Agriculture, Food and Nutrition Service, Nutrition Education: 
Principles of Sound Impact Evaluation (Alexandria, Va.: September 
2005). 

[31] A FNS-funded study found that it is technically feasible to 
collect bar-code data on the products purchased with food stamp 
benefits and to link those data with demographic information about the 
food stamp household. For details, see John Kirlin, et al., Feasibility 
Study of Capturing Food Data at Checkout, U.S. Department of 
Agriculture, Food and Nutrition Service (Alexandria, Va.: September 
1999). 

[32] For a detailed description of card-based alternatives, see U.S. 
Department of Agriculture, Food and Nutrition Service, Analysis of 
Alternatives for Implementing a Cash Value Voucher Program, prepared by 
the State Information Technology Consortium (Herndon, Va.: March 2007). 

[33] Providing participants with gift cards containing a financial 
incentive amount that is stored on the card is another option. While 
this option may require fewer technical changes to the checkout system 
because some retailers already have systems in place to process gift 
card transactions, store gift cards can only be used at a limited 
number of store locations. Ensuring that only allowable foods are 
purchased may be more difficult with a store gift card because, 
typically, there is no limitation on what can be purchased in the 
participating stores with the gift cards and neither a PIN nor a 
signature is required to complete a transaction. 

[34] Smart cards are embedded with a computer chip that can perform 
calculations and store significant amounts of data. These cards are 
currently used in four states to process WIC EBT transactions. 

[End of section] 

GAO's Mission: 

The Government Accountability Office, the audit, evaluation and 
investigative arm of Congress, exists to support Congress in meeting 
its constitutional responsibilities and to help improve the performance 
and accountability of the federal government for the American people. 
GAO examines the use of public funds; evaluates federal programs and 
policies; and provides analyses, recommendations, and other assistance 
to help Congress make informed oversight, policy, and funding 
decisions. GAO's commitment to good government is reflected in its core 
values of accountability, integrity, and reliability. 

Obtaining Copies of GAO Reports and Testimony: 

The fastest and easiest way to obtain copies of GAO documents at no 
cost is through GAO's Web site [hyperlink, http://www.gao.gov]. Each 
weekday, GAO posts newly released reports, testimony, and 
correspondence on its Web site. To have GAO e-mail you a list of newly 
posted products every afternoon, go to [hyperlink, http://www.gao.gov] 
and select "E-mail Updates." 

Order by Mail or Phone: 

The first copy of each printed report is free. Additional copies are $2 
each. A check or money order should be made out to the Superintendent 
of Documents. GAO also accepts VISA and Mastercard. Orders for 100 or 
more copies mailed to a single address are discounted 25 percent. 
Orders should be sent to: 

U.S. Government Accountability Office: 
441 G Street NW, Room LM: 
Washington, D.C. 20548: 

To order by Phone: 
Voice: (202) 512-6000: 
TDD: (202) 512-2537: 
Fax: (202) 512-6061: 

To Report Fraud, Waste, and Abuse in Federal Programs: 

Contact: 

Web site: [hyperlink, http://www.gao.gov/fraudnet/fraudnet.htm]: 
E-mail: fraudnet@gao.gov: 
Automated answering system: (800) 424-5454 or (202) 512-7470: 

Congressional Relations: 

Ralph Dawn, Managing Director, dawnr@gao.gov: 
(202) 512-4400: 
U.S. Government Accountability Office: 
441 G Street NW, Room 7125: 
Washington, D.C. 20548: 

Public Affairs: 

Chuck Young, Managing Director, youngc1@gao.gov: 
(202) 512-4800: 
U.S. Government Accountability Office: 
441 G Street NW, Room 7149: 
Washington, D.C. 20548: