Actually, they DID test the space shuttle tanks in Mississippi. See here: http://en.wikipedia.org/wiki/MPTA-ET. They don't test every tank there, but they tested the early ones and they test tanks that have certain kinds of problems, such as leaks.

As for building the tanks in Florida, the Kennedy VAB in Florida is 8 acres. The Michoud facility in New Orleans is vastly larger: more than 43 acres. Saying "because they built the VAB in Florida they could have built a manufacturing facility for large scale boosters" shows little knowledge of how the tanks are produced and how much space it takes to produce them. There was an existing 43 acre facility laid out to produce large scale industrial products in LA that the government already owned. The only additional cost of putting production there was barge traffic to Mississippi/Florida. Compare that to the cost of building a new 43-acre environmentally controlled facility from scratch in Florida, and I don't think that's a trivial decision. You can convince me, but you haven't thus far.

I'll freely admit that most of the Michoud space is not currently used. Original production plans were to produce one or two tanks per month, which grossly overestimated the Space Shuttle's turnaround efficiency. In the end, much of the Michoud capacity was not needed. But if you want to make the argument that Michoud was selected wholly or even primarily as a political favor, you have more homework to do.

I can't speak to the other cases, but they built the first stage of the Saturn V in Louisiana because... 1. They needed a very large industrial space. The factory is a converted aircraft production facility built during WWII. 2. They needed deep water access to ship the giant vehicles. External Tanks and the Saturn S-IC cannot be shipped by road or rail. 3. The selection was tied to accessibility of both Kennedy Space Center in Florida (for launching) and Stennis Space Center in Mississippi (for testing). Once you start to pick a few locations, logistics become non-trivial and your next choice becomes more constrained. I can't assure you that graft had nothing to do with it (can you assure me that any part of any government program in history didn't have some back-deal component?), but the locations were not selected purely for political effect.

I totally agree; I too was left very concerned after reading his thoughtful and thorough analysis. And I'm really disappointed that (practically) no one on Slashdot took a look at it. Especially after reading his analysis of confirmation bias and so on, the knee-jerk responses on this thread are particularly disheartening.

The law review article upon which the linked story is written is 80 pages long. It is, as best I can tell, totally consistent with known science (it doesn't postulate "black holes destroying the universe" or any such nonsense). It is an attempt to do 3 things:

1. Ask how a court ought to address a science experiment that could, by some very unlikely chance, destroy the earth. He uses LHC as an example, but also suggests Strong AI and nanotechnology as possible future examples.

2. Analyze how a non-expert court can, or should, evaluate highly technical and possibly controversial scientific claims for and against the safety of a bleeding-edge research project.

3. Analyze how logical or cognitive errors could realistically lead a scientist to accidentally or intentionally understate or mischaracterize the risks of her research.

Anyone on this board droning on about "Shut-up-the-LHC-can't-destroy-the-world" either DRTFA or totally misunderstood it. I will now quote the author:

My motivation in writing is certainly not to engender fear. I have no apprehension to share...

It is part of our 21st Century reality that we must take seriously a number of surreal planetary disaster scenarios. In that sense, the synthetic-black-hole disaster is not unique. For some time now, we have been confronted with the possibility of nuclear war and global climate change. In the future, we may have to remove still more scenarios from the science fiction category and place them on a list of real worries. Someday, we may need to seriously consider catastrophic threats from nanotechnology, genetic engineering, or artificial intelligence. Each one of these human-made global disaster scenarios involves incredibly complex questions of science, engineering, and mathematics. Courts must develop tools to deal meaningfully with such complexity. Otherwise, the wildly expanding sphere of human knowledge will overwhelm the institution of the courts and undo the rule of law—just when we need it most.

If he had chosen anthropogenic global climate change as his topic of analysis, I think there would have been a more interesting debate on Slashdot, but apparently any mention of "LHC" in the same breath as "black hole" causes some sort of hysterical allergic reaction in some people.