Product Access, Retailer Supply, and the Economics of Shortages

Abstract

This dissertation examines the inequality in consumers' product access during times of shortage from three distinct perspectives. The first chapter studies the welfare cost and distributional impact of cleaning product stockouts during the early COVID-19 pandemic. Since stockouts are not directly observed in the data, I develop an algorithm to quantify the extent of stockouts across stores using weekly sales data. Demand analysis shows that product differentiation is large enough such that the stockouts of popular products has led to significant welfare loss for consumers. Two patterns of product availability emerged among retail chains: High-Availability Retailers (HARs) such as discount stores and warehouse clubs have more product varieties, greater inventories, and lower stockout rates than Low-Availability Retailers (LARs). The uneven distribution of HAR stores across markets drives the inequality in product access during times of scarcity, with shoppers in "availability deserts"---markets scarce in HAR stores---experiencing poorer product access and lower welfare. In the second chapter, I switch attention to the online marketplace where price gouging was rampant during demand shocks, and investigate the impact of anti-gouging laws on consumers' product access. These laws impose price caps to curb unfair price rise, but they may also reduce product supply, which leads to inefficient allocation among consumers. I focus on personal cleaning products sold on Amazon during the COVID-19 crisis, and find that during time periods characterized by significant price increase and inelastic supply, a binding price cap enhances the welfare of all income groups, which indicates that the welfare gain from price reduction dominates the cost of inefficient allocation. Conversely, in periods with less intense price gouging, price caps tend to benefit low-income consumers while hurting rich shoppers. The welfare impact of changing the cap depends on market characteristics that varies substantially over time, suggesting that static price caps are unlikely to yield optimal welfare results. In the final chapter, I analyze the inventory acquisition and distribution strategies of major retail chains during demand shocks. This complements Chapter 1 by exploring why HARs perform better in providing product access during times of shortages. Compared with LARs, HARs like discount stores have superior logistic capabilities to efficiently distribute inventories to their stores, which results in lower stockout rates, more consistent weekly sales, and less variation across establishments. HARs also procure a wider variety and larger quantities of products from upstream suppliers, ensuring that their stores have more abundant product offerings during times of crises. Nevertheless, the superior product availability at HAR stores is attributed to their intrinsically higher inventory levels and greater product diversity, rather than to a higher responsiveness to rising demand shocks in each market.