Food Spending by Age and Household Size

I received several emails and comments about my post a couple days ago on food spending by households with different incomes.  For example, over on twitter Adam Ozimek asked:

I'm happy to help provide additional information.  Here is total food spending by age and income.


As the figure shows, low income households all spend about the same on food regardless of age.  People aged 65-74 years tend to spend the least on food regardless of income until the highest income categories at which point the oldest respondents spend the least on food.  Households between the ages of 25 and 44 years tend to spend the most on food (holding constant factors such as household size, etc.)

How much of this food spending is away from home?  Here is how households allocate their food budget between away from home vs. at home by age and income.


Regardless of age, households with higher incomes tend to spend more of their food budget eating out than lower income households.  However, the youngest consumers tend to spend much more of their budget away from home than older consumers.  At the lowest income category, for example, people age 18-24 spend 38% of their food budget way from home whereas people age 65-74 only spend 25% of their food budget away from home. All high income households (above $160,000) allocate more than 40% of their food budget to away from home spending - the highest is by 25-34 year olds who spend 46% of their food budget away from home.  

What about household size?  It's fairly well known that there are economies of scale in household food spending (i.e., two people can eat more cheaply than one on a per-person basis).  For example, the SNAP (or food stamp) program provides up to $194/month for a one person household.  If every person was expected to spend the same, then one should give $194*2=$388 for a two person household.  But, that's not what the SNAP program does.  They only give up to $357/month for a two person household.  The program administrators didn't just decide this willy-nilly, but rather they observed in spending data (like the kind I'm using here) that spending doesn't increase 1:1 for each additional person in the household.  

In my data, for example, the estimated spending on food at home for a household of size one is $73.60/week, but the spending at home for a household of size two is far less than double ($73.60*2=$147.20) and is only $92.12/week.  The figure below shows spending on food at home and away from home for households with 1, 2, 3, and 4 members holding constant income, age, education, etc. Spending on food away from home is essentially flat.  Does that mean a four person household can eat out for the same as a two person person household?  Not necessarily.  It may mean that 4 person households are eating at McDonald's while 2 person household are eating at something a little higher end.