In a paper marked “preliminary and incomplete,” Brent Neiman and Joseph Vavra write,
Our key findings, that household product concentration and cross-household variance are both rising, is pervasive and robust: it holds in every geographic market in Nielsen data and so is not simply the result of growing differences across regions. Similarly, while household concentration levels differ with some observable demographic characteristics, the rise in household concentration and cross-household variance holds within income, race, education, age, and other demographic groups. The pervasiveness of the patterns suggest a key role for marketing, production or search technologies, or broad-based increases in preference heterogeneity. We demonstrate the salience of product innovation and the expansion of available varieties for the increase in concentration and segmentation of consumption: these trends hold within most retailers but are strongest in those with the largest increases in offered UPCs. In addition, when we compute concentration growth separately for existing and new products, we find that household concentration increases much more in the bundle of new products. That is, the products accounting for rising shares of consumption within narrow categories are predominantly new products not previously consumed by the household. Over the last decade there have been large increases in the set of products available within particular retailers, and our results suggest this enhances the scope for households to sort into particular product niches.
Thirty years ago, we already knew that households could be sorted into different consumption groups. Looking at just the period 2004-2015, the authors write,
Increases in household-product Herfindahls are even larger, rising by roughly 10 percent over the same time span. Overall these increases in household product concentration are nearly monotonic across time, highly statistically significant, and robust to a variety of specification changes.
Amazon bets on the effect by disaggregating delivery and constantly expanding the order book. Bote this is about shopping efficiency, consumers being limited by participating in deliveries and inventory management.
Given a strong incentive for suppliers at every level of the supply chain to increase their product diversity (allows better extraction of surplus through discrimination, confusion, etc.) and the noisy and sticky behavior of consumers (again at all supply chain levels), is this trend any surprise?
The counteracting force is the “menu costs” (i.e. costs of managing more products, marketing, IT) of suppliers, but it has been constantly decreasing with improved IT, so the trend is expected.
In general, this is yet another trend favoring the suppliers. The weakening of the consumers (or even small firms consuming from other, larger ones) is seen across the statistics of the last thirty years.