The Department of Energy, which gave us debacles like the Solyndra and Abound Solar bankruptcies, has put out a report claiming that its green-energy loan programs are turning a profit for taxpayers — a $5 billion profit, in fact. This claim was credulously reported by Bloomberg Businessweek, but don’t pop the Champagne yet: Former acting CBO director Donald Marron, now of the Tax Policy Center, takes apart this claim on the TPC’s blog. The government’s “profit” calculations don’t even take into account the major expenses:

DOE’s report does not address this issue, except in a footnote in a table (cut and pasted above) revealing that its $810 million of “interest earned” was “calculated without respect to Treasury’s borrowing cost.” In other words, DOE reports gross interest received, not the net interest taxpayers have earned after subtracting Treasury borrowing costs. The incomplete figures in the table seem to suggest that DOE has eked out a $30 million profit on its lending ($810 million in interest less $780 million in loan losses). But when we account for Treasury borrowing costs, taxpayers are actually well behind.

The report does not allow us to say just how far behind. We do know, however, that DOE loans are typically made at small, sometimes zero, spreads above Treasury rates. So a large portion of DOE’s “interest earned” must have been offset by borrowing costs. That puts taxpayer losses in the hundreds of millions of dollars.

The same concern applies to DOE’s statement that interest payments on these loans will eventually top $5 billion.

Taxpayers aren’t making money off these government loan guarantees, period.

And that’s even before we consider that a lot of projected government “profits” are often the result of an accounting illusion. As I mentioned before, unlike private companies, the government ignores the market risks of loans, making it look as if the agency is turning a profit when, in fact, it isn’t. The CBO issued a report on this issue here, I wrote about it here, and Marron has a paper on this here.

I would take issue with one of Marron’s points:

DOE’s lending programs should not be evaluated solely or even primarily based on their profitability or lack thereof. What matters is their overall social impact. How much are they advancing new technologies? How much are they reducing future pollution? Have they created jobs and economic growth? And are any gains worth the taxpayer subsidies? Those are the questions we should be trying to answer.

I disagree with a lot of that.

First, the government shouldn’t be in the business of lending money to private companies or giving them outright subsidies — whether the companies are small or large, whether they export or not, or whether they produce green energy or dirty energy, they shouldn’t be getting help. I realize that this argument isn’t convincing to most lawmakers or technocrats (even some free-market ones). But it doesn’t make it less true.

Moreover, the social impact that these programs make is almost impossible to measure properly. It is especially hard, and certainly not in the interest of most of those involved, to account for the unseen victims and all the hidden costs of these programs. Both lawmakers defending the programs and its beneficiaries churn out measures about job creation, profitability, and economic growth. It doesn’t matter how much these numbers are debunked or who does the debunking (the GAO, or the Ex-Im’s own IG), special interests and politicians continue to shamelessly use these numbers.

This is exactly what goes on with the DOE loan programs. You may remember that the agency claimed that the 1705 loan programs created 2,378 permanent jobs thanks to about $16 billions in loan guarantees. Taxpayers are left to realize that this works out to $6,731,034 in taxpayer exposure per job.​ They also aren’t told that these job numbers should be taken with a heavy dose of skepticism because the methods used don’t recognize that many of the businesses would exist otherwise, double-count jobs, and fail to differentiate among part-time, temporary, or seasonal jobs. If we have learned anything in the last few weeks, is that we should be highly skeptical of scoring coming from the government, so easy is it to game.

Finally, the federal government should never be in the business of lending money to private businesses because it allocates money based on politics rather than sound economics. As any student of public-choice economics knows, even with good intentions such a program is likely to end badly.

If you look at the rationale for most loan-guarantee programs, we’re told that the government needs to step in because the private lenders wouldn’t otherwise because there’s too much risk. First, it seems problematic that our lawmakers are willing to put our money on the line to lend money to borrowers that capitalists won’t tolerate (especially when so many Americans are struggling to pay their own bills in the first place). The private capital market’s refusal to lend money to risky borrowers is a feature, not a bug. Second, the data also reveals that most of the money actually goes to companies that already have access to capital anyway. All the government is doing is giving some firms an unfair leg up over the competition by lowering their borrowing costs, creating some market distortions, moral hazard, and lost competitiveness.

Case in point: The 1705 loan program that Marron suggests we should assess mostly benefited large and profitable companies. According to the DOE data, the top ten recipients of loans under the 1705 program were all solar generation companies, which received a combined $12.2 billion in loan guarantees and 76 percent of the overall amount guaranteed.

Among them: NextEra Energy Resources, LLC, a Fortune 200 company; Abengoa Solar Inc., a Spanish multinational company; and Prologis, a global real-estate investment trust. Utility firms like NRG received three separate loans in the top-ten recipient list.

The data also shows that nearly 90 percent of the loans guaranteed by the federal government since 2009 went to subsidize lower-risk power plants, which in many cases were backed by big companies with vast resources. This includes loans such as the $90 million guarantee granted to Cogentrix, a subsidiary of Goldman Sachs. That would be the same Goldman Sachs that ranks No. 74 on the list of America’s Fortune 500 companies — think they need taxpayers’ help?

As with many recipients of Ex-Im Bank benefits, even the recipients admitted they didn’t need the money. Quoted in the New York Times at the time, David W. Crane, NRG’s chief executive, explained, “I have never seen anything that I have had to do in my 20 years in the power industry that involved less risk than these projects . . . It is just filling the desert with panels.”

Again, this isn’t unique to the DOE programs or to the Ex-Im Bank; this is just how Washington allocates money. The best way to put an end to the deception is to stop giving money to private companies.

This article originally appeared at National Review Online.