Can we learn from failure? Examining US experiences with nuclear repository siting

The United States may well have spent more money in the effort to develop nuclear waste sites than any other nation on earth, but save for the case of a facility in New Mexico that handles only military wastes, every single effort to site a new repository in the last several decades has ended in failure. Given that the siting efforts to date have generally ignored the advice of social scientists, this track record makes it difficult to know whether that advice could lead to outcomes that affected parties would consider ‘successes’ in the US context, but there are important opportunities to learn from what nearly all parties consider to have been ‘failures’. One particularly telling case involves efforts to develop a repository for lower-level radioactive wastes in New York State. There, as in many other US cases, the relevant officials chose to focus on an arbitrarily narrowed subset of work on socio-economic impacts that the agency considered ‘credible’, ignoring social science findings on stigma, public controversy, and so-called ‘special’ impacts related to the radioactive character of the wastes. The officials also appear to have erred in insisting on a top-down, ‘technical’ approach to screening, and in ignoring the very real impacts created by their own actions. Ultimately, the process led to a degree of outcry that was remarkable even in the context of nuclear waste facilities; the citizen blockades and round-the-clock outposts at the boundaries of the selected areas were sufficiently well-organized to prevent even preliminary on-site analyses, meaning that the socio-economic impacts that were excluded from consideration proved to be not just ‘credible’ but vivid, bringing the entire process to a halt, while none of the ‘credible’ biophysical impacts ever took place. The case suggests that essentially universal patterns should be relatively easy to predict, but that unfortunate choices can make the problems of siting even worse.