I just came across a video that was also shared on Facebook, that claims that a recent US Department of Agriculture study found that 10% of food stamp purchases go to soda, or $8.5 billion annually. They also say that

the #1 item purchased through food stamps (SNAP) is soda.

Is this true?

  • 42
    As soda is often the cheapest drink sold, cheaper than bottled water, why would that be strange? – Agent_L Jan 18 '17 at 9:49
  • 13
    Concerning Soda vs bottles water: how drinkable does tap-water tend to be in the US? – Layna Jan 18 '17 at 12:16
  • 2
    Does anyone know what's the "street" cash-rate for food stamps? In Education of a Speculator Niederhofer (quoting some hobo) gives 50% but that sounds very profitable for the buyer. – TheMathemagician Jan 18 '17 at 13:14
  • 12
    @Layna entirely drinkable if by "drinkable" you mean potable. If you include aesthetic factors like temperature, taste, amount of sugar and carbonation, your results will vary. – ReasonablySkeptical Jan 18 '17 at 13:58
  • 15
    @Layna If you live in Flint Michigan, "not very". Some reports put the quality at ~90% in compliance with EPA standards, but unregulated chemical contamination is widespread. Some people are fine with tap water, and some are very concerned about it - and some people have good reason to be concerned, whether they are or not. You could say "it's complicated". – BrianH Jan 18 '17 at 18:35

TL;DR: Some parts of the claim are true, some are false. The first version of a New York Times story on this report contained an error, which seems to have spread (although the NYT story has been updated with the correct number).

Soft drinks were the #1 ranked individual commodity purchased by SNAP households in a 2011 dataset of SNAP transactions at one US grocery retailer (compared to #2 among non-SNAP households), but they constitute 5% of SNAP expenditures (compared to 4% in non-SNAP households), not 10%.

The $8.5 billion number is not accurate. A better estimate would be about $3.86 billion for 2011 (5.44% of $71 billion in SNAP benefits redeemed), if the spending patterns at this one grocery retailer generalize to all SNAP spending. I don't know where the $8.5 billion number came from, it's not in the NYT article.

Here is the USDA report. Note that although the report is from 2016, the data is from 2011. Also, it is based on a dataset from a single grocery retailer; it includes $6.7 billion worth of SNAP household expenditures from that year, less than a tenth of the > $71 billion total SNAP benefits redeemed that year. The dollar amounts in the tables below show the dollar amount of transactions in the dataset, not of all SNAP transactions that year.

The report says:

Across all households, more money was spent on soft drinks than any other item. SNAP households spent somewhat more on soft drinks than non-SNAP households (5 versus 4 percent).

It also notes that 9.3% of SNAP expenditures are on "Sweetened Beverages", a broader category which also includes many beverages not considered "soft drinks": juice drinks (containing 50% juice or less), juice blends and smoothies, sports and energy drinks, sweetened instant tea and hot cocoa mixes, and similar items (see report, appendix B-19, B-20). (Compared to 7.1% in non-SNAP households.) "Sweetened Beverages" is the #2 category by expenditures among SNAP households.

Also see the following tables, which compare the top food items and top food categories purchased by SNAP and non-SNAP households:

Table 5

Table 6

The erroneous 10% number seems to have come from the New York Times, who in a January 13, 2017 story In the Shopping Cart of a Food Stamp Household: Lots of Soda mistakenly wrote:

The findings show that the No. 1 purchases by SNAP households are soft drinks, which accounted for about 10 percent of the dollars they spent on food.

The error was pointed out on Twitter. The current version of the NYT story has the correct 5% number. A correction notice at the bottom of the article says:

Correction: January 18, 2017

An article on Saturday about the grocery spending habits of households receiving food stamps misstated the portion of spending used to purchase soft drinks. Households receiving food stamps spent 9.3 percent of their bills on “sweetened beverages,” a category that includes soft drinks, juices, and energy drinks, among others; they did not spend that percentage on soft drinks alone.

Some more criticisms of the NYT reporting on this story can be found here, here, here, and here.

  • 2
    mind that many people consider all non-alcoholic drinks "soft drinks", so the 10% number using that criteria might actually be low as the 9.3 you mention doesn't include things like coffee, milk, and bottled water. – jwenting Jan 18 '17 at 6:58
  • 12
    @jwenting The claim in the question is about "soda". Who calls coffee, milk, and bottled water "soda"? – ff524 Jan 18 '17 at 7:00
  • 2
    @Nzall maybe, but many people don't separate them when talking "soft drinks" so would take the total of them all. That's my point... – jwenting Jan 18 '17 at 9:00
  • 7
    @jwenting What "many people" would do is not the point of this question. This question asks: "a video on facebook claims that a report of the DoA found that 10% of food stamps purchases are soda. Is this factual?" It does not matter what most people consider soda, it matters what the DoA report considers soda. – Nzall Jan 18 '17 at 9:08
  • 6
    @Nzall I think it does matter whether the definition used in the DoA report matches up with what the video author and the people spreading the claim on facebook consider soda and what they are trying to cause people who see the claim to believe about food stamp recipients' spending habits, so "what most people consider soda" is absolutely relevant. – Random832 Jan 18 '17 at 16:00

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .