The ability of the lawyers to drag out the approval and construction phase of the reactors thru litigation drives the costs up to the point that the reactors are not economical.
Even this is not the primary factor, or even a
major factor. Yes, litigation and the bad U.S. environmental permitting doesn't
help, but the same kind of economic issues were seen in
most countries, save those that had very active nuclear construction programs (like France--and, again,
terrorist attacks on reactors). In the U.S., the economic issues were mainly due to tightening regulations in the
early to
mid 1970s (for the very good reason that the existing regulations grossly underestimated certain kinds of risk and produced dangerous plants) and the high degree of customization for each U.S. nuclear reactor, driven in large part by...well, the decentralization problem, where each reactor was a custom order by a utility rather than simply the result of the AEC going around and saying "we're going to build PWR Standard #1 here, here, here..." which is basically what happened in the countries that had more successful nuclear programs.
This is why I always come back around to the decentralization problem. The fundamental issue is that no one in the U.S. system has both the authority
and the will to
make people use nuclear power primarily. The utilities and state regulatory bodies, which have the authority, just want to produce cheap power. The AEC/DoE has the will (sort of their bag), but no authority to make people use it. If you
make people build nuclear plants, then obviously you'll have more nuclear plants, but more importantly you'll help with the economic issues. If you want nuclear power to be bigger in the U.S., this is the
key problem to solve, and it really needs PoDs that have nothing directly to do with nuclear power.