The literature available on socio-economic impact assessments of GM crops shows a very real problem that regulatory systems will likely face when dealing with regulatory decisions involving socio-economics. That is, the fact that there will be a host of competing socio-economic impact claims with regard to a specific technology.
In my experience, it can be even get to the point where the decision maker may face a set of assessments based on the same or similar set of farmers -even the same source database- and then have the assessments reach contrastingly different results as a consequence of the method used and the implementing teams that has conducted the assessment.
After all, when doing an assessment, be it a an economic or a socio/cultural/anthropological assessment especially in an ex ante (before adoption) assessment where there will be a multitude of assumptions used to forecast or provide estimations (be it Social Impact Assessments, Sustainable Livelihoods or Economic Impact Assessments) there is quite a bit of subjectivity involved in the assumptions and scientific tools used to forecast or estimate potential impacts. In essence we have a scientific process that has a lot of art in its implementation. I do not want to discuss those situations where there is clearly a conflict of interest and/or unethical behavior.
The regulatory/decision making process will require having clearly defined rules and decision making standards to ensure that competing claims are properly evaluated and decided upon. Here it is important to note that it is important to have as much detail as needed in implementing regulations, but we have to be cognizant that too much detail can increase the likelihood of regulations becoming inflexible and thus prescriptive. This reduces the opportunities to include innovative assessment alternatives which may make the process more efficient.