This finding seems counterintuitive, as one would expect high-income families to be spending more while dining out. Our methodology sheds no direct light on that question, but we may infer that the home-owning upper income families have a frequency of eating out that is less than that of middle- and low-income families. As the per- centage of high-income families in the general population is relatively low, our results indicate that the restaurant industry is not receiving the benefit of their wealth and their residence in the neighborhood. At the same time, low- and middle-income families may be spending less per visit at a restaurant, but a larger proportion of them may be eat- ing out and doing so frequently, resulting in higher levels of restaurant revenues.