Public support for policies to improve the nutritional impact of the Supplemental Nutrition Assistance Program (SNAP)

Abstract

Objective

To determine public attitudes towards federal spending on nutrition assistance programmes and support for policies to improve the nutritional impact of the Supplemental Nutrition Assistance Program (SNAP).

Design

Participants answered survey questions by telephone assessing support for SNAP spending and proposed programme policy changes.

Setting

USA.

Subjects

Survey of 3024 adults selected by random digit dialling conducted in April 2012, including 418 SNAP participants.

Results

A majority (77 %; 95 % CI 75, 79 %) of all respondents supported maintaining or increasing SNAP benefits, with higher support among Democrats (88 %; 95 % CI 86, 90 %) than Republicans (61 %; 95 % CI 58, 65 %). The public supported policies to improve the nutritional impact of SNAP. Eighty-two per cent (95 % CI 80, 84 %) of respondents supported providing additional benefits to programme participants that can only be used on healthful foods. Sixty-nine per cent (95 % CI 67, 71 %) of respondents supported removing SNAP benefits for sugary drinks. A majority of SNAP participants (54 %; 95 % CI 48, 60 %) supported removing SNAP benefits for sugary drinks. Of the 46 % (95 % CI 40, 52 %) of SNAP participants who initially opposed removing sugary drinks, 45 % (95 % CI 36, 54 %) supported removing SNAP benefits for sugary drinks if the policy also included additional benefits to purchase healthful foods.

Conclusions

The US public broadly supports increasing or maintaining spending on SNAP. The majority of respondents, including SNAP participants, supported policies to improve the nutritional impact of SNAP by restricting the purchase of sugary drinks and incentivizing purchase of healthful foods with SNAP benefits.

Facing ongoing national fiscal budgetary challenges, policy makers in the USA are debating whether to cut spending on federal nutrition assistance programmes, including the Supplemental Nutrition Assistance Program (SNAP), formerly known as the Food Stamp Program^{( 1 , 2 )}. Simultaneously, the public health community is urgently seeking policy solutions to the obesity epidemic with the potential to both improve population health and reduce future obesity-related medical expenditures^{( 3 )}.

In December 2011, a record 46·5 million people, or approximately one in seven Americans, participated in SNAP^{( 4 )}. With a budget of $US 75 billion in 2011, SNAP is the largest of the fifteen federal nutrition assistance programmes. SNAP aims to alleviate hunger and improve the health of low-income individuals by providing resources to purchase food^{( 5 )}. As one of the major federal anti-poverty programmes, SNAP has provided a critical buffer against poverty for millions of adults and children^{( 6 )}. Despite the programme's success at reducing hunger and poverty, few efforts have been undertaken to leverage spending on SNAP to improve the health of programme participants. Recent evidence-based nutrition updates to the National School Lunch Program and the Special Supplemental Nutrition Program for Women, Infants, and Children demonstrate the potential for aligning SNAP benefits with the Dietary Guidelines for Americans^{( 3 , 7 , 8 )}.

A range of policies aimed at improving the nutritional impact of SNAP have been proposed or piloted, including incentivizing SNAP participants to purchase healthful foods such as fruits and vegetables, increasing the amount of SNAP benefits per household, limiting the use of SNAP benefits for unhealthful foods such as sugar-sweetened beverages or sugary drinks, and increasing nutrition education efforts targeting SNAP participants^{( 3 , 8 – 10 )}. In 2011, the US Department of Agriculture (USDA) rejected a request from New York City (NYC) to pilot test removing sugar-sweetened beverages from the products SNAP participants could purchase using benefit dollars^{( 11 )}. The USDA noted potential stigmatization of SNAP participants as one of the reasons why it rejected the proposal. However, support for the proposal among SNAP participants was not assessed as part of the agency's review process in making the decision.

To inform the ongoing SNAP policy debate, we assessed the opinion of a representative sample of US adults, including SNAP participants, regarding federal spending on SNAP and on policy proposals to improve the nutritional impact of SNAP.

Methods

Data are from a four-question survey added to an ongoing weekly Harris Interactive poll by researchers at the Harvard School of Public Health assessing attitudes regarding support for federal spending on nutrition assistance and farming programmes as well as support for policies to improve the nutritional impact of SNAP (Appendix). Harris Interactive conducted the random-digit-dialled telephone survey as part of the weekly Harris Poll National Quorum omnibus survey within the USA between 12 and 22 April 2012, among 3024 adults (aged 18 years and over), including 418 adults who reported that their household had received SNAP benefits in the previous 12 months. That survey was conducted over three waves with ∼1000 respondents each wave on 12–15 April, 13–16 April and 19–22 April. Cooperation and response rates in the Harris Poll National Quorum are not reported for every wave fielded, but the poll had an average cooperation rate of 36 % and an average response rate of 5 % during this period computed according to the American Association for Public Opinion Research standard definitions for cooperation (COOP3) and response (RR3) rates, assuming 31 % of telephone numbers with unknown eligibility would be eligible based on recent research^{( 12 , 13 )}. According to the American Association for Public Opinion Research, the cooperation rate is the number of complete interviews divided by all eligible households contacted, whereas the response rate is the number of complete interviews divided by all eligible households in the sample.

Survey responses were weighted for age, sex, race/ethnicity, education, region, number of adults in the household and number of telephone lines in the household where necessary to bring them into line with their actual proportions in the population. Although response rates to telephone surveys have declined in recent years, independent studies have found that weighted results from shorter-duration telephone surveys are similar to results from surveys with longer duration and higher response rates^{( 13 – 15 )}. In addition to possible non-response bias and sampling error, variation in responses may occur due to question wording or order when compared with other surveys.

Demographic variation in support for SNAP spending and support for SNAP nutrition policy proposals is presented based on weighted and unweighted responses, with 95 % confidence intervals based on weighted data. There were no qualitative differences between weighted and unweighted results. Differences in the proportion of respondents supporting SNAP spending and nutrition policy proposals by demographics were tested for significance adjusting for survey weighting using the Rao–Scott χ ² test. Data were analysed using the PROC SURVEYFREQ statement in the SAS^® statistical software package version 9·3.

Results

Seventy-seven per cent of the public believed that federal spending on SNAP should be increased (48 %) or maintained (29 %; Table 1). Only 21 % of the public believed that federal spending on SNAP should be decreased. Among other factors, support varied significantly by political party identification (P < 0·001); 88 % of Democrats believed that federal spending should be increased or maintained compared with 81 % of Independents and 61 % of Republicans. The majority of respondents supported increasing or maintaining spending on SNAP across all sociodemographic subgroups measured.

Table 1.

Public opinion regarding federal spending on SNAP by sociodemographic factors in a survey of 3024 US adults aged 18 years or older selected by random digit dialling conducted in April 2012

	Increase or maintain spending				Decrease spending
	Frequency	%	Weighted %	95 % CI for weighted %	Frequency	%	Weighted %	95 % CI for weighted %	P value*
Overall	2244	74	77	75, 79	700	23	21	19, 23
Political party identification
Republican	697	57	61	58, 65	490	40	36	33, 40	<0·0001
Democrat	1204	88	88	86, 90	128	9	10	8, 12
Independent	245	77	81	76, 86	56	18	16	11, 21
Region
Northeast	437	79	79	75, 83	108	19	19	15, 23	0·417
Midwest	511	75	76	72, 80	150	22	22	18, 26
South	795	71	75	72, 78	291	26	23	20, 26
West	501	75	80	76, 83	151	23	18	14, 21
Income
<$US 35 000	761	83	87	85, 90	127	14	11	9, 13	<0·0001
$US 35 000–49 999	281	77	77	72, 83	81	22	22	16, 28
$US 50 000–74 999	291	68	72	66, 77	125	29	26	21, 30
$US 75 000–99 999	226	71	69	63, 76	88	28	29	23, 35
≥$US 100 000	339	67	68	63, 73	157	31	31	26, 36
Don't know/refused	346	69	71	65, 76	122	24	23	18, 27
Education
≤High school	776	79	81	78, 84	173	18	16	14, 19	<0·0001
Some college	630	74	76	73, 80	206	24	22	19, 25
≥College	805	70	70	67, 74	314	27	27	24, 30
Race/ethnicity
White NH	1642	72	74	72, 76	591	26	24	22, 26	<0·0001
Black NH	272	91	91	87, 95	19	6	7	3, 10
Hispanic	142	83	84	77, 90	25	15	15	9, 21
Other/mixed race	147	74	77	69, 84	48	24	22	15, 29
Gender
Male	1077	71	74	71, 77	408	27	24	21, 27	0·001
Female	1167	78	80	77, 82	292	19	18	15, 20
Age (years)
18–34	253	79	81	77, 86	61	19	17	13, 21	<0·0001
35–44	237	72	73	68, 78	83	25	24	20, 29
45–54	440	75	77	74, 81	132	23	21	17, 24
55–64	577	78	79	76, 82	153	21	19	16, 22
65+	726	71	72	69, 75	260	25	23	21, 26

SNAP, Supplemental Nutrition Assistance Program; NH, non-Hispanic.

Proportions are presented based on both unweighted and weighted frequencies. Party identification as Democrat or Republican includes those who responded as Independents but reported leaning towards either party. Whites, blacks and other/mixed race are non-Hispanic; Hispanics are of any race. Responses to the SNAP spending question are not reported from respondents answering ‘don't know’ or refusing responses to demographic questions unless the ‘don't know/refused’ category was greater than 10 % of respondents to the demographic question.

*P value based on Rao–Scott χ ² test of association between demographic categories and attitudes towards spending (increased a lot, increased a little or kept the same v. decreased a little, decreased a lot or don't know) on SNAP adjusting for survey design.

Support for proposed policies to improve the nutritional impact of SNAP was also high across political party identification and SNAP participation status (Table 2). Of the four policies proposed, respondents were most supportive of a proposal to provide ‘…additional money to SNAP (Food Stamp) participants that can only be used on fruits, vegetables or other healthful foods’. Eighty-two per cent of all respondents supported this proposal, including 87 % of Democrats and 76 % of Republicans. SNAP participants also supported this proposal (86 %) more than any of the other three policies proposed. While still supported by the majority of respondents, support was lowest (65 %) for a proposal to provide SNAP participants ‘…more food stamp dollars to guarantee that they can afford a healthy diet’. This proposal to increase SNAP benefit levels had the largest gap in support across political party identification, with only 49 % of Republicans supporting the proposal compared with 77 % of Democrats.

Table 2.

Support for policy proposals to improve nutritional impact of SNAP by political party identification and SNAP status in a survey of 3024 US adults aged 18 years or older selected by random digit dialling conducted in April 2012

	Support for removing sugary drinks from list of approved SNAP products				Support for providing additional money to SNAP participants that can only be used on fruits, vegetables or other healthful foods				Support for providing SNAP participants with more food stamp dollars to guarantee that they can afford a healthy diet				Support for educating SNAP participants by providing nutrition or cooking classes
	%	Weighted %	95 % CI for weighted %	P value*	%	Weighted %	95 % CI for weighted %	P value	%	Weighted %	95 % CI for weighted %	P value	%	Weighted %	95 % CI for weighted %	P value
Overall	71	69	67, 71		81	82	80, 84		62	65	63, 67		73	74	72, 76
Political party identification
Republican	70	70	67, 74	0·034	73	76	73,79	<0·001	44	49	46, 53	<0·001	61	63	59, 66	<0·001
Democrat	72	69	66, 72		88	87	85, 90		77	77	75, 80		84	83	80, 86
Independent	71	70	64, 76		85	83	77, 88		63	67	60, 73		72	77	71, 82
SNAP participation
Participants (n 418)	57	54	48, 60	<0·001	85	86	81, 90	0·212	80	82	78, 87	<0·001	76	80	75, 84	0·056
Non-participants (n 2596)	73	73	71, 75		80	81	79, 83		58	61	58, 63		72	73	71, 75

SNAP, Supplemental Nutrition Assistance Program

Proportions are presented based on both unweighted and weighted frequencies. Party identification as Democrat or Republican includes those who responded as Independents but reported leaning towards either party. Responses to SNAP policy questions are not reported from the fewer than 10 % of respondents who answered ‘don't know’ or refused to respond to demographic questions.

*P values based on the Rao–Scott χ ² test of association between political party or SNAP participation and support (strongly support or somewhat support v. somewhat oppose, strongly oppose or don't know) for each policy adjusting for survey design. The frequency of ‘don't know’ responses ranged from 1 % to 3 % of all respondents for each SNAP policy question.

The proposal to remove ‘…sugary drinks (such as soda) from the list of products that can be purchased using SNAP (or Food Stamp) benefits’ was supported by 69 % of all respondents with no gap in support between Republicans (70 %) and Democrats (69 %). A majority of SNAP participants (54 %) supported removing sugary drinks from SNAP benefits. Respondents who did not support removing sugary drinks from SNAP benefits were asked whether they would support removing sugary drinks if the proposal was combined with a policy that would provide additional money to purchase fruits, vegetables and other healthy foods. Of the 46 % of SNAP participants who initially did not support removing sugary drinks (n 181), 45 % (95 % CI 36, 54 %) supported removing sugary drinks if the policy also included additional benefits to purchase healthful foods (data not shown).

Discussion

As Congress debates changes to SNAP and other components of US farm and nutrition policy, the present nationally representative survey found widespread public support for increased or maintained federal spending on SNAP. These results are very consistent with a survey conducted for the Food Research and Action Center in January 2012, which similarly found that 77 % per cent of US adults did not support cutting federal spending on SNAP as a way to reduce government spending^{( 16 )}. A June 2012 survey conducted for the National Journal also found that a majority (62 %) of respondents supported increasing or keeping SNAP spending the same^{( 17 )}. As research from the USDA has recently shown, SNAP provides a critical buffer against the experience of severe poverty, particularly for children^{( 6 )}.

In addition to confirming public support for maintaining SNAP benefits identified in previous surveys, the broad public support for removing sugary drinks from SNAP benefits identified in the present study highlights the need for a more comprehensive debate about the current ability of SNAP to fulfil its mandate to ‘provide for improved levels of nutrition among low-income households’^{( 18 )}. Although a recent Institute of Medicine report emphasized that improving the nutritional impact of SNAP is an essential component of a national strategy addressing the obesity epidemic due to the programme's scale, reach and level of federal investment in the programme, the Institute of Medicine committee concluded that in addition to practical considerations, ‘…limiting food choices for SNAP recipients may be viewed as patronizing and discriminatory to low-income consumers’^{( 3 )}. Similar concerns regarding potential stigmatization of SNAP participants were raised in the USDA's rejection of the request from NYC to pilot test removing sugary drinks from SNAP benefits^{( 11 )}. However, a 2011 in-person survey of 498 SNAP participants and applicants in NYC found that 49 % supported NYC's proposal to remove SNAP benefits for purchasing sugary drinks, compared with 16 % who did not care and 35 % who did not support the proposal^{( 19 )}. The current study is the first national survey of whether SNAP participants would support removing SNAP benefits for the purchase of sugary drinks. In contrast to concerns over patronizing SNAP participants and in line with findings from NYC, over half of SNAP participants surveyed in the present study supported the proposal. The percentage rises to three-quarters of SNAP participants supporting removing sugary drinks from benefits if the policy also included incentives to purchase healthful foods.

While the USDA rejected NYC's request to evaluate the impact of removing sugary drinks from SNAP benefits, it is currently running a small $US 20 million pilot programme testing the impact of incentivizing the purchase of fruits and vegetables in SNAP^{( 10 )}. Consistent public approval for incentivizing SNAP fruit and vegetable purchases across political parties and SNAP participation status supports increased investment in evaluation of this strategy. Additionally, given the demonstrated negative health impact of sugary drinks^{( 20 )} and public support identified in the present study, policy makers should consider a pilot programme that both incentivizes healthful SNAP purchases such as fruits and vegetables and removes sugary drinks from the list of products that can be purchased with SNAP benefits.

The present study provides decision makers with a clear statement of public support for continued federal investment in preventing hunger and severe poverty through the SNAP. As decision makers seek to improve the effectiveness and impact of federal spending, the findings from the study also document widespread support for policy proposals to align investments in SNAP with the broader goal of improving the health of Americans.

Appendix

Survey developed by researchers at the Harvard School of Public Health. Random-digit-dialled landline-only telephone survey of 3024 US adults aged 18 years or older conducted by Harris Interactive in three waves fielded 12–15 April, 13–16 April and 19–22 April 2012.

Base: all respondents

Q1 ‘Congress is currently debating support for farming and nutrition assistance programs included in the federal budget. I'm going to read some nutrition and farm assistance programs and for each one, I'd like you to tell me whether you think spending on that program should be increased a lot, increased a little, kept the same, decreased a little or decreased a lot.’ (READ EACH ITEM – ‘Should spending on this program be increased a lot, increased a little, kept the same, decreased a little or decreased a lot?’)

• 1.Increased a lot
• 2.Increased a little
• 3.Kept the same
• 4.Decreased a little
• 5.Decreased a lot
• 6.Don't know (v)

(1–6 below read in random order)

• 1.‘The Supplemental Nutrition Assistance Program, also known as the SNAP or Food Stamp program, which helps forty-six million people in the US afford food.’
• 2.‘The WIC program, which provides nutritious food to nine million pregnant or breastfeeding women, infants and children under 5 years old.’
• 3.‘The School Lunch and Breakfast Programs which provide low-cost or free meals to thirty-two million school children.’
• 4.‘Crop Insurance Programs, which pay farmers if their crops are lost due to weather or natural disasters.’
• 5.‘Payments to support farmers growing corn, wheat, soybeans, and other crops.’
• 6.‘Conservation programs that protect farmland and other natural resources.’

Base: all respondents

Q2 ‘How much do you agree or disagree that participants in the SNAP or Food Stamp program use their SNAP benefits to purchase healthy foods? Do you strongly agree, somewhat agree, somewhat disagree or strongly disagree?’

• 1.Strongly agree
• 2.Somewhat agree
• 3.Somewhat disagree
• 4.Strongly disagree
• 5.Don't know (v)

Base: all respondents

Q3 ‘Please tell me how much you would support or oppose the following SNAP (or Food Stamp) program policies to improve participants’ diets.’ (READ EACH ITEM – ‘Do you strongly support, somewhat support, somewhat oppose or strongly oppose this policy?’)

• 1.Strongly support
• 2.Somewhat support
• 3.Somewhat oppose
• 4.Strongly oppose
• 5.Don't know (v)

(1–4 below read in random order)

• 1.
  ‘Removing sugary drinks (such as soda) from the list of products that can be purchased using SNAP (or Food Stamp) benefits.’

  • 1a.
    (ASK ONLY OF RESPONDENTS WHO CHOSE ‘SOMEWHAT OPPOSE, STRONGLY OPPOSE OR DON'T KNOW ON Q3.1) ‘Providing additional money to SNAP participants that can only be used on fruits, vegetables or other healthful food in addition to the removal of sugary drinks from the list of products participants can purchase with food stamps.’
• 2.‘Providing additional money to SNAP (or Food Stamp) participants that can only be used on fruits, vegetables, or other healthful foods.’
• 3.‘Providing SNAP (or Food Stamp) participants with more food stamp dollars to guarantee that they can afford a healthy diet.’
• 4.‘Educating SNAP (or Food Stamp) participants by providing nutrition or cooking classes.’

Base: all respondents

Q4 ‘During the past 12 months, have you or a member of your immediate family received benefits from the SNAP (or Food Stamp) program?’

• 1.Yes
• 2.No
• 3.Don't know (v)
• 4.Decline to answer (v)