Predicting consumer demand responses to carbon labels

Providing carbon footprint labels for all food products is a daunting and potentially infeasible project. Knowing how consumers substitute away from high carbon goods and what they choose as substitutes is essential for understanding which goods are likely to result in meaningful reductions in carbon emissions. This paper proposes a model to systematically estimate how consumers will respond to information from a carbon footprint label. Our model uses consumers' value of their individual carbon footprint with own- and cross-price elasticities of demand data on carbon emissions from life cycle analysis to simulate shifts in consumer demand for 42 food products and a non-food composite, and subsequent changes in carbon emissions from different labeling schemes. Our simulation results have several findings, including: (1) carbon labels can reduce emissions, but labeling only some items could lead to perverse impacts where consumers substitute away from labeled goods to unlabeled goods with a higher carbon footprint; (2) carbon labels can inform consumers such that their previous beliefs about carbon footprints matter; and (3) carbon labels on alcohol and meat would achieve the largest decreases in carbon emissions among the 42 food products studied.