Abstinence from child labor and profit seeking

Opponents of child labor suggest that demand-driven mechanisms can eliminate such practices. If consumers prefer adult-labor products, they suggest that labeling will drive out child-labor products. I show that under Bertrand competition, even with complete information, this is unlikely. When products are differentiated only by labor type, such a scheme requires at least three firms and child-labor costs at least as large as adult-labor costs. With additional lines of differentiation, child labor is eliminated only when the adult-labor costs are smaller. Thus, while labeling may reduce child labor, it is unlikely to eliminate it.