Nutrition labeling: Information effects on consumer behavior and welfare

Policy makers use cost/benefit analyses to justify regulations altering the amount or presentation of health-related information to consumers. The current method of benefit analysis, the cost of avoided illness approach, is limited because it assumes that consumers hold the same risk assessments as scientists and that the social benefits associated with the provision of health-related information are adequately represented by changes in illness occurrences. Development of a benefit estimation method that relies on observed behavior is useful because demand shifts induced by an information change reflect consumer risk perceptions, consumer risk preferences, and consumer preferences for other (non-risk) attributes. The dissertation uses grocery store scanner data and an Almost Ideal Demand System framework to measure the market and welfare impacts of providing nutrient information.;The results confirm that nutrient labeling significantly affects consumer behavior and welfare. The value of labeling is largest for less educated individuals, although the education effect varies across products. The value of labeling is also lower for older individuals, consistent with results indicating that older individuals are less likely to use health-related information. The results provide evidence that consumers act as if they have nutrient (or health risk) budgets. Providing nutrient information does not necessarily lead to increased consumption of healthy foods. In fact, two effects may occur: consumers may reduce their net intake of unhealthy nutrients and/or consumers may switch consumption away from unhealthy products in those food categories where differences in other quality characteristics (e.g., taste) are relatively small between the more and less healthy products, and switch toward unhealthy products in categories where differences may be relatively large (i.e., a 'substitution effect'). When this substitution effect is large then changes in the overall consumption of healthy products (and the resulting health risks) may be small. Thus, nutrient labeling may increase social welfare without any significant change in health risk. For this case, the cost of avoided illness approach would underestimate the social benefits of providing nutrient information.