Economists Address a Century Old Blind-Spot in Pricing Theory
Economists have studied product menus and market segmentation, but new research examines what happens when firms use both strategies simultaneously to examine when personalized pricing hurts, and helps, consumers.
Most modern consumers are familiar with the 'menus' of quality offered by companies and sort themselves by willingness to pay, be it a 'basic' tier software subscription or a first-class flight. Armed with ever richer data, companies then divide those same customers into groups that face different prices. Economists have studied each practice for a century, but little was known about what happens when firms do both simultaneously.
But does this kind of price differentiation necessarily leave consumers worse off? A Cowles Foundation Discussion paper by Dirk Bergemann, Tibor Heumann, and Michael C. Wang (forthcoming in the American Economic Review) shows that the answer depends on the structure of the market.
Premium customers pay less than their full willingness to pay because they can always trade down to the basic tier. The better the basic product, the more the seller must concede to premium buyers to keep them from switching. Consumer welfare therefore hinges on the quality offered at the bottom of the menu.
But if a firm concentrates price-sensitive buyers into one segment and makes them a larger share of that submarket, it strengthens the seller's incentive to serve them well. Any premium customers in that are mixed in to this segment then capture larger rents.
"The benefits of segmentation depend critically on demand and cost elasticities, with no segmentation being optimal when aggregate demand is sufficiently elastic."
The findings complicate the idea that personalized pricing is inherently harmful. Segmentation can either increase or decrease consumer welfare. What matters is not simply how much information firms possess, but how consumer demand and firms’ costs interact.
This result challenges two simple policy views. One view treats personalized pricing and market segmentation as necessarily harmful. The other argues that segmentation must help consumers because it lets firms serve more buyers. The paper shows that neither view is correct, but rather depends on the market. For regulators, blanket restrictions on market segmentation may sometimes prevent practices that actually benefit consumers.
The elasticity test offers a clear guide. In markets with highly responsive demand and relatively rigid quality provision, segmentation has no potential consumer benefit under the paper’s conditions. Restrictions on personalized pricing or data-based segmentation have a strong consumer-welfare basis in these markets, since segmentation only rearranges consumers. That conclusion changes when demand is less responsive or when the seller can adjust quality easily. A ban on segmentation can then prevent arrangements that improve quality for price-sensitive consumers.