November 26, 2018
Gravitec Development is undertaking a major exploration of supermarket per-capita expenditures. This email is another set of notes that discuss our findings.
We all know that supermarket PCW is inversely related to household size; as the household size goes up, the PCW goes down. In fact, household size is the best single predictor of supermarket PCW, and that has been true for as long as we have been researching the topic. But how strong is the relationship?
Gravitec Development used the data from the 2017 BLS Diary Survey to perform a weighted linear multiple regression based on household size and income, the two largest predictors of supermarket PCW. The stronger relationship was with household size, which yielded a slope of about -6. (See the graph below). Interestingly, the slope was not as strong as in 2008 (-8) or 2009 (-9). This was true for both supermarket PCW and “Food Away” (food purchased for consumption at home).
In fact, household size now only accounts for about 50% of the variance, a measurable drop from our results of previous years.
These figures suggest the household size is not as important as in past years, although it is still the single best predictor of supermarket PCW. The flattening of the household size slope may be the result of the incredible growth of deep discounters, such as dollar stores, Aldi, and other chains. A small family today no longer needs to buy “bargain-sized” packages in order to reap the rewards of lower price.
Keep watching our columns! Or visit our web-site to find our pages on Project PCW.