It seems like there could have been much more interesting information extracted from this dataset than they actually got. Instead we're getting gems like the fact that people who make more money wind up spending more money on food, and "the people who are spending the most money on food overall devote more money to dining out". Families with children spend more money on food than single people.

The analysis goes on to ask interesting questions but not answer them. "how did people in Austin spend $12,447 on food and drink last year, while people in Miami spent half as much? What are they buying in Austin? And what aren't they eating in Miami?" Perhaps because Miami has a lot of old people and old people spend a lot less on food than young people? The data exists to figure out whether this would be sufficient to account for it, but they don't bother doing the analysis.

Oh, and the one graph I really wanted to see was "proportion of food budget spent on going out" vs "income" which I suspect might show some interesting behaviour. But the graph wasn't there.