Canice Prendergast on How Prices Can Improve a Food Fight (and Help the Poor)
What this covers
Canice Prendergast of the University of Chicago's Booth School of Business describes how Feeding America, the national food bank network, replaced its centralized allocation queue with an artificial-currency auction system. Partnering with Russ Roberts in conversation, Prendergast walks through how 215 local food banks now bid on food using shares proportional to the poverty levels in their service areas, enabling them to express their actual needs while preserving need-based fairness. The core innovation involves a sealed-bid auction clearing twice daily, with crucial mechanisms: when banks spend shares, that money redistributes back to all banks at day's end, creating a general-equilibrium effect where bidding up expensive items drives down staples, and negative prices allow banks to receive compensation for taking undesirable items.
The conversation ranges across the practical details of implementation and the economic principles underlying it. Prendergast explains how repeated game dynamics allow banks to substitute across time and food types, stabilizing an artificial currency that would otherwise lose value. He details the system's performance: within seven months, Feeding America began receiving 60–70 million additional pounds annually—a gain largely sustained even through the 2008 financial crisis. The design deliberately avoided last-second bidding tactics that would disadvantage small rural banks, and allowed fractional bidding and credit systems to level access. Prendergast connects the reform to Hayek's information problem and contrasts it with in-kind welfare versus cash assistance debates. He also addresses the food banking community's division over paternalism—whether to restrict offerings to nutritious foods or permit choice—showing how the market mechanism uniquely lets individual banks sort along this dimension. Throughout, he emphasizes that the appeals committee set up for exceptional cases has never convened, suggesting the system achieved legitimacy despite initial skepticism.
Prendergast argues that replacing a centralized, need-based queue with an artificial-currency auction let local food banks express their true preferences, which simultaneously improved allocation efficiency, preserved need-based fairness, and increased total food supply.
- Allocating 'shares' in proportion to local poverty preserves need-based distribution while enabling choice
- Redistributing spent shares to all food banks creates a general-equilibrium effect that makes staples cheap for food-poor banks
- Adding negative prices and allowing banks to resell their own surplus increased liquidity and supply
Decentralized market mechanisms outperformed top-down allocation by incorporating local knowledge and incentives.
- The food bank reform is an application of what Hayek called the information problem—the difficulty of knowing what people actually want—which connects to debates about in-kind welfare versus giving cash, as argued by Milton Friedman and increasingly adopted in international development.
“all of this is an application of what Hayek called the information problem--that it's very difficult to know what people actually want.”
- The reform did not create a 'real' market but transformed a fully top-down allocation into a hybrid that decentralized the assignment mechanism by adding market elements and incentives, which—unlike many partial markets that fail because they lack supporting institutions—proved fairly successful here.
“It's still a top-down solution but they decentralized a big chunk of it... So it's not a "real" market, but it's got market elements, which is often the case in these kind of markets.”
Democratic adoption requirement forced the system to serve dispersed smaller banks, not concentrated powerful ones.
- Because the change was adopted by a one-food-bank-one-vote ballot among all 215 food banks rather than mandated by the committee, the resulting system had to be attractive to the numerous small food banks and could not cater to the few large, powerful ones that handle most of the food volume.
“the one foodbank, one vote meant that whatever system you did choose had to be very attractive to the smaller food banks. It couldn't cater to the larger, more powerful ones. Correct?”