Measuring Willingness to Pay for Environmental Attributes in Seafood

We investigate whether consumers are willing to pay for sustainability in seafood. To do this, we estimate a logit random utility model (RUM) of seafood purchases using a product-level scanner dataset from a quasi-experimental setting that includes data both before and after the implementation of a seafood advisory and sustainability label. Each seafood product is defined as a bundle of attributes, including price, species, and sustainability rating. The sustainability rating is communicated to consumers through the use of a color-based traffic light label system, where a color rating is assigned to each seafood stock-keeping unit. Combining a structural demand model with a difference-in-differences approach allows us to take advantage of the implementation of the labeling treatment in a subset of stores in the local retail chain to estimate consumers' willingness to pay (WTP) for green, yellow, and red sustainability labels. We find that the addition of a yellow sustainability label negatively impacts consumer's WTP for seafood products, however this simple average effect does not fully capture many independent underlying mechanisms, such as consumer preferences for wild-caught versus farmed products, and the color-distribution of available labeled products within a species, which are empirically explored. Additional results from a second stage generalized least squares regression of RUM product fixed effects on product characteristics indicate that consumers prefer selective harvest methods, wild caught seafood, and U.S. caught seafood.