[I urge readers to stick their head in a vise before reading this.]

head.jpgI have previously discussed the non-job-creating $50 billion in nuclear loan guarantees the Senate put into the stimulus (see here). For the record it was Sen. Robert Bennett (R-Utah), which I point out merely because “R-Utah” perfectly describes thinking behind this farce.

Not only won’t these loans generate any jobs in Obama’s first term, but as Peter Bradford, former member of the Nuclear Regulatory Commission, explained to me, it could actually kill jobs. How?

The capital markets are not swimming in credit. If you have billions in taxpayer backed loans for your project, even for a project that will take years to finalize and see actual jobs, you may well suck up money that might be otherwise be available for, say, wind projects that are shovel ready now. Bradford called the nuke loans “straw ready.”

Grist thanks its sponsors. Become one.

Worse, utilities that actually use these loans to build a nuclear plant would face an all but certain drop in their credit rating — see here. That means we are setting ourselves up to take over more trouble assets, since the Congressional Budget Office estimates the likely default rate of these loans at over 50 percent. If you liked nationalizing banks and insurance companies, you’ll love nationalizing nuclear utilities!

But here is where it gets particularly farcical: The loans only got snuck into the bill by budget gimmickry that replicates the high-leverage, fraudulent risk analysis that got us into the subprime mortgage and credit default swap mess. Some leading nuclear energy experts explained this to me Tuesday, and I will do my best to explain it to you.

Grist thanks its sponsors. Become one.

[I must warn you again that continuing to read this post puts you at great risk of uncontrolled cranial expansion.]

The Washington Post explained (not quite completely) last week:

Bennett’s amendment took $500 million away from $10 billion initially allotted to a new loan guarantee program for renewable energy and electric transmission projects and moved it to an existing loan guarantee program established under the Energy Policy Act of 2005. The existing program covers a much wider variety of energy projects, including “advanced nuclear” power plants, plants that “gasify” coal or turn it into liquid form, and plants that capture and bury carbon dioxide, a greenhouse gas produced by coal power plants.

Moving the money allows the government to stretch its loan guarantees further. Because of different accounting methods used in the two programs, a $500 million appropriation would permit approximately $5 billion in loan guarantees under the renewable program but $50 billion under the broader, existing program.

Yes, the $500 million switch cost the nation $5 billion in renewable and transmission loans but somehow gained $50 billion in nuclear loans. Does this mean nuclear power plants are 10 times less risky? Does this mean that nuclear power plants have a 1 percent default rate?


The Congressional Budget Office itself explained in a 2003 report [PDF]:

CBO considers the risk of default on such a loan guarantee to be very high-well above 50 percent. The key factor accounting for this risk is that we expect that the plant would be uneconomic to operate because of its high construction costs, relative to other electricity generation sources.


But wait. The CBO does believe that there is some recoverable value in a defaulted plant. Don’t ask my why, since I have no idea how they get value from some uneconomic, half-built plant that is probably the subject of major lawsuits (see “Nuclear meltdown in Finland”, here). I’m just a physicist, after all, and they are economists:

CBO estimates that the net present value of amounts recovered by the government on its loan guarantee from continued plant operations following a default and the project’s technical and regulatory risk would result in a subsidy cost of 30 percent.

I would note that the July 2008 report by the Government Accountability Office on the Loan Guarantee Program assumed “a default rate of 50.85 percent and a recovery rate of 50 percent, which result in a loss rate of 25.42 percent when multiplied together.”

So why is $50 billion in loans being scored as having a $500 million cost, when, in fact, CBO says that the subsidy is closer to 30 percent ($15 billion) and GAO says it is 25 percent ($12.5 billion)? This is where you have to enter the Alice-in-Wonderland world [or is that the Bernie-Madoff world] of budgetary scoring.

The LGP is built around the requirement/assumption that the industry getting these loans will pay, upfront, the equivalent value of the subsidy. I kid you not. The GAO explains:

The subsidy cost, as defined by the Federal Credit Reform Act of 1990, is the government’s estimated net long-term cost, in present value terms, of direct or guaranteed loans over the entire period the loans are outstanding (not including administrative costs). In calculating the subsidy cost for a guaranteed loan program, agencies estimate (1) payments from the government to cover interest subsidies, defaults, delinquencies, or other payments, and (2) payments to the government, including fees, penalties, and recoveries on defaults. Under FCRA, DOE would estimate the expected subsidy costs before issuing loan guarantees and is generally required to annually update, or reestimate, this cost to reflect actual loan performance and changes in expected future loan performance. To the extent that DOE underestimates subsidy costs and does not collect enough fees from borrowers, taxpayers will ultimately make up the difference.

The reason $50 billion in loans are being scored as $500 million is because the CBO is assuming that the DOE will only underestimate the subsidy (i.e. the default rate and recovery-on-default rate) by 1 percent. Don’t believe the CBO could be that credulous? Here’s the GAO:

The Congressional Budget Office (CBO) has estimated that DOE will charge companies fees at least one percent lower than costs, on average.

Does anybody on the entire planet outside of CBO think that? The GAO doesn’t appear to. And who pays if the DOE screws up:

To the extent that DOE underestimates the costs of the program and does not collect sufficient fees from borrowers to cover the true costs, taxpayers will ultimately bear the costs of shortfalls.

But it is much worse than all that.

The nuclear industry has no intention whatsoever of paying, upfront, 25 to 30 percent of the loan to cover the subsidy. Remember, as it is currently written, the loan program is only for a maximum of 80 percent of the cost of the plant. That means, for, say, a $10 billion, 1200 MW plant, the nuclear industry would have to put up the $2 billion not covered by the loan plus up to $3 billion for the subsidy. Not gonna happen.

The nuclear industry will be working hard over the coming year to insert language into legislation, most likely whatever energy bill comes out of Congress, that forces the tax payer to cover the cost of t
he subsidy. And I suspect they’ll try to get the loan guarantees to cover 100 percent of the cost.

Again, if you liked nationalizing banks and insurance companies, you’ll love nationalizing nuclear utilities!

One last thing. I’m also told that the nuclear industry may try to actually borrow the money for the loan itself from the Treasury — i.e. you and me — but I just can’t contemplate that possibility without risking a Scanners-like fate.

The only good news is that there is some small chance this $50B in loans may be dropped in the conference.

You may remove your vise.

This post was created for ClimateProgress.org, a project of the Center for American Progress Action Fund.