of certainty about the magnitude of effect varies tremendously, with some effects being very clearly associated with sources and others not. Indeed, some effects cannot be quantified at all. Hammitt mentioned that costs in the NRC (2010) energy report were calculated with uncertainty ranges (see Table 5-1 in Chapter 5).

Several factors contribute to this uncertainty, not the least of which is the heterogeneous nature of the food landscape. Several participants commented at numerous times on the tremendous variation in production that exists across both space and time, which creates both analytical and data challenges. Another contributing factor is the complexity of the food system and the challenge of teasing apart pathways. As just one example, while almost the entire workshop discussion was on food products, an audience member pointed out that the same systems that produce animal food products also produce nonfood by-products (e.g., hides, fats, pharmaceutical products). He asked how costs and benefits should be allocated between food products versus nonfood by-products. Hammitt replied that there is “no non-arbitrary way” to allocate external costs (and benefits) associated with raising animals across all of the various products. Rather than considering total costs and how to allocate those costs, he suggested considering marginal changes that would occur if animals were raised in a different way. John Antle opined that if externalities were addressed at the production level (e.g., by taxing or otherwise imposing measures that impact industry decisions about production), they would be appropriately valued into the system at that level and reflected in the cost of product and by-product production. Marty Heller agreed with Antle that the problem could be approached from that sort of “system expansion perspective,” whereby all of those other components (i.e., by-product production pathways) are incorporated into the analytical model, but cautioned that an analysis of impacts of the food system in particular may still require some sort of allocation of costs among food versus nonfood products.


For some effects, lack of sufficient data may also contribute to uncertainty about the magnitude of those effects. There were many calls throughout the workshop for more data and research. For example, there was a call during the public health effects break-out group for more research on the health effects of exposure to hormones in animal food products. As another example, during her presentation Anna Alberini listed four understudied areas that she thinks represent fertile ground for new research on food-related valuation: (1) willingness to pay (WTP) to reduce dietary cardiovascular and diabetes risks; (2) WTP to reduce endocrine disruption risks (i.e., which

The National Academies of Sciences, Engineering, and Medicine
500 Fifth St. N.W. | Washington, D.C. 20001

Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement