Military research is not a corporate subsidy.

This statistic bugs me:

From 1943 to 1999 the U.S. government paid nearly $151 billion, in 1999 dollars, in subsidies for wind, solar and nuclear power, Marshall Goldberg of the Renewable Energy Policy Project, a research organization in Washington, wrote in a July 2000 report. Of this total, 96.3 percent went to nuclear power, the report said.

Note the starting date of that statistic. 1943. The subsidy they are claiming we gave the nuclear industry was clearly the Manhattan project. It was not until 1957 that we built our first civilian nuclear reactor. That whole period until then was almost entirely military research. Even in the time since then much of this subsidy was military research that happened to be relevant to civilian applications. This research more than payed for itself. The United States Navy operates more nuclear reactors than any other organization in the world. You are not going to power an aircraft carrier on solar and wind, so this is a damn good use of money. Also, a lot of the remaining subsidy is entirely payed for by taxes on nuclear power plants! This was particularly true of the work at Yucca mountain. You can't seriously call money taken from an industry to pay for related infrastructure a subsidy.

Even if this statistic was accurate, it is still meaningless. There are about 100 civilian nuclear reactors in America. This, clearly grossly inflated, subsidy number only works out to about a billion and a half dollars per reactor. With a mere 150 Billion dollars in government subsidies we manage to produce a fifth of our electricity. The several billion we have spent on solar still only produces four hundredths of a percent of our electricity. With a payoff as low as that it is no wonder that the government chose to support nuclear instead of solar...

I don't buy into the case that nuclear energy is uneconomical. It rests entirely on the assumption that new reactors will be many times more expensive than the old reactors. The reactors we built in the fifties through eighties have managed to not kill a single member of the public with radiation poisoning while costing a small fraction of the 10 Billion that people are projecting new nuclear will cost. We could just build duplicates of these reactors at similar prices to what was payed back then and do just fine.

This discrepancy is magnified by the fact they project the cost of the first new reactor we build. Yes the first reactor we build might be 10 Billion dollars because we have to reinvent the wheel. However all that design work can pretty much be copied exactly for dozens of future reactors. Any reasonable projection of cost for the tenth, hundredth, or thousandth reactor is significantly lower than what costs were for the existing reactors we operate. Standardized designs alone would be enough to ensure that is true.