The National Environmental Policy Act is one of the most effective toothless laws ever written. It doesn't mandate any substantive outcome: There are no pollution thresholds, no land use requirements, and it creates no actions for money damages. Instead, it is a process oriented law. Any major federal action has to have an environmental impact statement before it can go forward, and that information, in turn, may influence policy. Environmental impacts must be considered, and because they are, at least around the edges, projects are often adjusted to reduce their environmental impacts. Citizens can bring suits to force agencies to carry out the environmental impact study process in good faith, even though they can't, by virtue of the act itself, compel any governmental action.
This legislative model was adopted in the House version of the 2007 fiscal year defense appropriations bill in at least two areas.
One mandate called for the Navy to evaluate nuclear propulsion in all new ships is considers building, noting the pressures of increasing oil prices and the logistics virtues of having ships, particularly those that operate in support of nuclear powered aircraft carriers, that aren't dependent upon oil.
The other mandate called for the Department of Defense to evaluate robotic options for all new major systems.
Neither proposal mandated what it supported. The proponents believe that if robotic and nuclear options are placed on the table, that they are likely to be adopted in many cases where they are appropriate, even without a Congressional mandate. But, located in an appropriations bill, an area where Congress really is supreme over the executive branch when it chooses to act, the suggestions may carry a great deal of weight.
What could this mean in practical terms?
It means that the CG(X), the new cruiser based on the DD(X) now repurposed with a missile defense role, will likely have a nuclear fueled option considered, and so will the replacement for the existing destroyer fleet which the DD(X) was so dismally designed to fill. It also means that proponents of diesel submarines for the U.S. Navy will face an even more daunting challenge than they already did.
If this mandate had come sooner, it might have favored UAVs like the Predator as an Army replacement to the Kiowa, rather than the ARH-70 helicopter, although this procurement battle may already be over. The push for more robots is also likely to influence looming procurement decisions in the areas of anti-land mine, and anti-water mine warfare, and perhaps in encouraging unmanned patrol boats, such as those now in used by the Navy of Singapore, and unmanned anti-submarine warfare platforms, perhaps in connection with the Littoral Combat Ship.
Of course, the biggest benefit of a mandate like this is not for systems we can already foresee, but systems for which a nuclear or robotic system hadn't been considered at all, but turns out to be quite useful -- perhaps moving supplies in a warehouse or logistics ship hold, for example.
Also, the law of unintended consequences can work in funny ways.
One reason that there is so much political support for buying Virginia Class submarines, is that there is a felt need to keep the nuclear propulsion expertise used to build them employed, so that the U.S. can keep that option open, and so that other nations don't steal them from the U.S. for their own ends. While some members of Congress believe that the submarine itself has virtue, I've heard the "corporate welfare" justification for that program more than once.
If those nuclear propulsion experts were employed building the CG(X), this might ease the pressure to keep building nuclear attack submarines, which might actually boost efforts to have the United States build small diesel submarines, which are sufficiently small that Navy procurement officials could say in good faith are not appropriate to outfit with nuclear propulsion.
On their face, these appropriations requirements have no real effect. The don't mandate anything, but a part of the process. But, it probably won't do much harm, and could do a lot of good.