Again, because I am slow, I'm pointing you all to a post by Jonathan on the Volokh Conspiracy about the National Research Council's criticisms of the Office of Management and Budget's new guidelines for risk assessments conducted by federal agencies. The full report is here, but there's a shorter press release. The NRC report points to several problems with OMB's plan, including

• having internal inconsistencies,
  
• inappropriately suggesting separate risk assessment approaches for general v. "influential" assessments, given the difficulty of determining what assessment might be influential at the outset,
  
• having too limited an evaluation of what constitutes an "adverse health effect,"
  
• erroneously focusing on human health risk assessments while neglecting technological risk assessments,
  
• providing risk assessment standards that are "beyond the state of the science", and
  
• failing to address risk communication (as compared to the a proposal by the European Commission) and sensitive populations.

So OMB's plan is currently on hold. (Thanks to my research assistant VS, who forwarded this to me.)