August 22, 2023
Concentrated spending
Why are US households spending more money on fewer products and showing increasing divergence from each other in their selections?
Source: BodnarPhoto
Consumer spending habits don’t stay the same for very long. Many popular products of the past are no longer on supermarket shelves, and inflation has shaken up consumers’ selections. But one shopping pattern has emerged in recent years that may have wider implications for how economists look at the economy as a whole.
In a paper in the American Economic Journal: Macroeconomics, authors Brent Neiman and Joseph Vavra show that over the last 15 years, households spent an increasing share of their budgets on fewer products.
The authors' findings explain why overall spending concentration in the economy has dropped at the same time that household spending concentration has risen.
The researchers used Homescan data from AC Nielsen that covered roughly 170,000 households from 2004 to 2016 to measure shopping behavior.
The data revealed that the share of individual households’ purchases within narrow categories, such as “Coffee” and “Cosmetics,” became more concentrated. One explanation for this trend could be the rise of better products, or so-called “superstar” products, that dominate the market because of their superiority.
But further analysis of aggregate spending paints a different picture, showing that total spending on “superstar” products actually became more evenly distributed across various products over the period. That is, at a more aggregated level, no one type of product stood out as increasingly more popular than others in terms of spending.
This divergence between household and aggregate concentration suggested to the authors that households were increasingly buying different products from each other. Moreover, this pattern held across a range of household characteristics.
Diverging trends in spending
The chart below shows that the Household Herfindahl index—a measure of household spending concentration—increased steadily from 2004 to 2016. At the same time, the Aggregate Herfindahl index—a measure of product spending concentration at the aggregate level—declined.
“It seemed like this divergence across different people's consumption baskets could have been driven by increasing inequality—rich and poor households increasingly consuming different goods from each other—or an urban–rural divide—people in cities consuming different goods from people in more rural areas,” Vavra told the AEA in an interview. “But it turns out that divergences based on those characteristics are completely dwarfed by the divergence that we're seeing within these groups.”
Why, then, are household spending habits becoming more divergent and concentrated?
To answer that question—and get a sense of the magnitude of this phenomenon—the researchers built a model capable of including households that can purchase multiple products and can differ in which products they buy.
“We allow for different households to have different spending baskets, but in a way where we can input how correlated those baskets are with each other,” Vavra said. “The model allows situations all the way from every household consumption basket being completely randomly different to the other extreme of every household basket being perfectly correlated.”
The authors’ model shows that one of the most important drivers of concentration is increased product variety.
When there are more options, households are able to find products that are better suited to their tastes, which leads them to purchase a more concentrated bundle. But since households naturally have different tastes, they diverge from each other. As a result, aggregate purchases become more fragmented.
Looking at simple measures of concentration and inferring that that tells you something concrete about market power is pretty hard to do. You really need to take the details of individual spending patterns and individual behavior seriously.
Joseph Vavra
While economists are often concerned that increasing concentration hurts consumers, the authors' framework suggests that this increased spending concentration is actually not a good indicator of market power. They say that firms are still competing, while doing a better job of satisfying different tastes, benefiting consumers overall.
“Looking at simple measures of concentration and inferring that that tells you something concrete about market power is pretty hard to do,” Vavra said. “You really need to take the details of individual spending patterns and individual behavior seriously.”
♦
“The Rise of Niche Consumption” appears in the July 2023 issue of the American Economic Journal: Macroeconomics.