some of these substances, FDA might well require submission of a formal food additive petition or data equivalent to that required for such a petition in order to demonstrate that any added substances were GRAS. The costs associated with such a submission would depend on the amount and type of data required to demonstrate safety at the anticipated level of exposure. In 1997, FDA estimated the costs of data requirements for food additive petitions using surveys from four food companies (FDA 1997b). Adjusted to 1998 dollars using implicit GDP price deflator reported by the Council of Economic Advisors (1999), these costs ranged from $174,787 to $1,359,456. Comparing these costs to the costs estimated in table A.7 for EPA data requirements under FIFRA and FFDCA, the additional cost associated with the EPA process would be in the range of $1.5 to $2.6 million. In the event FDA were to require an environmental assessment (EA) under NEPA in conjunction with the food additive petition, the costs of developing data to support the EA might fall in the range of $2500 to $50,000 (FDA 1997a), which would reduce the additional cost of the EPA process by the same amount
To assess the extent to which regulation might discourage R&D related to crops that would have small potential markets and R&D by smaller entities, I compare the potential costs of meeting regulatory requirements with the fixed costs of breeding a new variety. I ignore the variable costs of producing seed for commercial sale, which depend on the size of the market. I consider only the costs of developing a crop to the point where it would be ready to scale up production for commercial sale. The costs of regulatory compliance are a form of fixed cost in that regulatory approval is needed before commercialization. It is thus appropriate to compare regulatory compliance costs with the fixed costs of breeding a new variety.
The costs of crop breeding depend on the costs of running a breeding operation, the time required to develop a new variety sufficiently for market introduction, and the success rate of new varieties. A simple model can be used to indicate how those factors influence the cost of developing a new variety. Let C denote the cost of running a breeding facility for a year, including direct costs and overhead; for simplicity, C is assumed to be constant (in real terms). Let T be the expected time required to develop a variety to the point where it is marketable. Let N denote the average number of years between successful introductions of new varieties, so that an average of 1/N new varieties are introduced every year. Let r be the (real) interest rate. The average cost of a new