Per capita consumption of fluid cow’s milk has declined steadily in North America and Europe. In the United States, that well-known decline accelerated in percentage terms in the past two decades as the market for plant-based substitutes surged. This paper documents and explains significant impacts of plant-based products on the retail prices and quantities of fluid cow’s milk products in the United States, using a discrete-choice model of milk product demand with random coefficients. Demand estimates based on household purchase and store scanner data from 2018 to 2020 show that plant-based milk is more substitutable for organic and lactose-free cow’s milk products than for conventional cow’s milk. Based on our demand estimates, if plant-based milk products had not been available in the consumer choice set, the retail quantity of organic cow’s milk would have been 14.5%, lactose-free cow’s milk 10.2% higher, and conventional cow’s milk by 7.7% higher. The absence of plant-based alternatives would increase annual retail purchases of fluid cow’s milk by 1.46 gallons for a representative survey of households. In these data, each gallon of plant-based milk displaced 0.68 gallons of cow’s milk. While plant-based milk continues to affect the demand for fluid cow’s milk products, especially the higher-priced organic and lactose-free products, evidence shows that it has not been the primary driver in the recent accelerated decline in the consumption of fluid cow’s milk. 

We exploit the variation in the rollout of refrigerated almond milk across U.S. retail stores from 2008 through 2010 to estimate the causal effects on cow’s milk and soymilk. Using monthly retail store scanner data, we find that introducing almond milk reduced retail quantities of soymilk (3.6%) and lactose-free cow’s milk (3%), with smaller effects on organic (2%) and conventional cow’s milk (1%). Effects are larger in stores that adopt earlier or reach higher almond milk shares. Our econometric estimates indicate significant displacement of per capita consumption of soymilk and lactose-free cow’s milk, but not for conventional cow’s milk.

This paper examines the responses of alcohol sales to temperature shocks and evaluates how a warming climate influences alcohol sales across the United States. Using retail store sales data from 2006 to 2023 combined with product-level alcohol content from a barcode database, we convert weekly sales volume into ethanol-equivalent volume for each store. We find that alcohol sales, on average, rise with temperature up to roughly 90℉, beyond which the positive effect becomes smaller. The effect varies sharply by region: Warmer regions show much weaker responses to high temperatures than cooler regions, reflecting adaptation to hotter conditions. Applying these estimates to climate projections, national alcohol sales are projected to remain unchanged by mid-century under a high-emissions scenario, as increases in cooler regions offset declines in warmer regions. By the end of the century, sales are projected to decline by 2.4%, with larger declines in warmer regions (4.7%).