solar subsidies

Assessment of Incentives and Employment Impacts of Solar Industry Deployment

A mixed portfolio of energy options has allowed Americans to enjoy long-term economic growth and prosperity. The federal government has engaged directly in developing each energy resource in the mix, although the dollar value estimates of this federal support vary considerably. This report focuses on solar power, evaluating the diffusion of solar energy technology in the United States in the context of the technology adoption process and federal engagement in developing energy options.

Natural gas fracking well in Louisiana

Eric Lipton and Clifford Krauss are winning no friends in the renewable energy industry with their article "A Gold Rush of Subsidies in Clean Energy Search."  I understand some of the upset.  After all, even with the most positive interpretation possible of the negligence that led to vast amounts of publicly-owned oil and gas in the Gulf of Mexico being taken for zero royalties, we are still talking a loss to the Treasury of $53 billion over the next 25 years (and quite well described in this link in the Economist).  That is, shall we say, quite a lot of solar panels.  Yet, this legislative and policy implementation ineptitude has not gotten nearly the level of scorn as the Solyndra default; in fact, attempts to role back the resource giveaways have been blocked in Congress.   Lipton and Krauss have singled out a solar facility even though nuclear plants, coal with CCS, ethanol plants, or all sorts of other subsidized industrial infrastructure have similar subsidy stacking attributes, conflicts of interests, and policy shortfalls, and could have selected instead.  These are critical points to remember, because too often libertarian and fiscal conservative groups who weigh in forcefully on subsidies to solar or ethanol remain uncomfortably silent when polite conversation shifts to decades-old subsidies to fossil fuels.

But content-wise, I think Lipton and Krauss did a nice job.  Given the secrecy that often surrounds the acquisition of subsidies at all levels of government like treats on Halloween eve, I would not have expected any reporter to be able to get everything right.  But there were lots of things they did get right, some of which are rarely captured in subsidy reviews.  If their detractors are angry at their selection of solar subsidies, I'd encourage them to direct their energies to applying a similar approach to uncover the full range of subsidies to selected conventional energy facilities.

Let's leave the solar element aside for a moment, and focus instead on the article as a case study for what is wrong with the US' competitive marketplace today.  First, it is clear that absent a major investigation by skilled analysts (in this case, the NYT even brought in the energy group at Booz & Company to estimate some values), nobody outside of the boardroom knows all of the subsidies flowing to any particular industrial facility.  This is not a useful way to run what is supposed to be a market economy.  Second, industry claims on job creation or economic contributions to the local economy are not vetted at the outset of a project and verified over time.  Rarely are at least some of the subsidies held in escrow pending delivery of these promised benefits. Escrows or other forms of guarantees that promised benefits will be realized are commonly used in all sorts of less expensive transactions; they ought to be a central part of these deals as well.  Third, the authors point out some of the problems that can arise if investors have little of their capital at risk, or can pull that risk off the table too quickly.

All of this is bad policy, whether applied to solar or to nuclear.  It leaves taxpayers more or less in the dark on how much they've put into a facility, and how much of this was above the minimum incentive needed to get the project off the ground rather than simply a transfer to investors or plant owners in the form of reduced risk or increased return.  Finally, when so many government entities are offering subsidies, we can enter a twilight-zone like scenario where one jurisdiction is merely bidding against another for who can give out the bigger slice of incentive pie.  The original goal of providing the smallest boost possible for a risky project with significant public benefits to tip from non-viable to viable is lost.

The Lipton and Krauss article brings to the forefront the issue of subsidy stacking, a term that appears to go back at least to the 1980s (thanks to Ron Steenblik for this link), but is finally entering the realm of routine policy evaluation.  Too often subsidies are evaluated based on national numbers for a single incentive.  Looking at the pool of incentives to specific projects is extremely important, however, as it is at the project level that the subsidies alter production decisions and distort markets for competing providers of goods and services. 

A good way to look at the difficulty in shining the light on subsidies is to review the rebuttal on the NYT article put out by NRG, the current owner of the California Valley Solar Ranch project.  I've reprinted it in its entirety below, along with some of my own commentary.


For those of you who have read the New York Times’ November 11 (online) article, “A Gold Rush of Subsidies in Clean Energy Search,” below are some of the facts that the article got wrong or didn’t include at all. The California Valley Solar Ranch (CVSR) is an ambitious, important project with a national purpose and NRG is proud of our efforts to help put people back to work building a more sustainable energy future.

The NYT says: “When construction is complete, NRG is eligible to receive a $430 million check from the Treasury Department — part of a change made in 2009 that allows clean-energy projects to receive 30 percent of their cost as a cash grant upfront instead of taking other taxbreaks gradually over several years.”
The Facts: $380mm will go to repay the DOE loan and the American taxpayer, and only $50 million flows to NRG as the equity investor in the project.

Commentary:  While it is nice that most of subsidy "B" is being used to repay subsidy "A," it is correct that these are two separate (and generous) subsidies going to the project.  The terms of the DOE's Title 17 loans (section 609.10) normally give the borrower the lesser of 30 years or 90% of the expected life of the asset to repay the government.  There is nothing in the statutory language or final rule that mandates accelerated pay-off as NRG has stated they are required to do.  Since they received their loan just before the program expired, it is possible they accepted repayment terms like this.  Another possibility is that they are defining the payment of a credit subsidy fee associated with the loan commitment as though it were repayment of the actual loan.  This seems unlikely, since multiple sources say that Congress appropriated the funds (from taxapyers) to cover the credit subsidy fee on the section 1705 loans -- another subsidy to the recipients.  Because the agreements are sealed, so we have no way to know. DOE's credit is generally believed to be inexpensive relative to other funding sources, so it would be in NRG's interest to delay loan repayment as long as possible.

The NYT says: [Lawrence H. Summers warned of] “’double dipping’ [of government incentives] that was starting to take place. They said investors had little ‘skin in the game.’”
The Facts: The Government protected itself against this concern by ensuring that most of the money from the treasury grant received by the project has to be used to pay down the DOE loan. On the CVSR project, almost 90% of the proceeds from the cash grant will go to pay down the DOE loan.

The NYT says: “PG&E, and ultimately its electric customers, will pay NRG $150 to $180 a megawatt-hour, according to a person familiar with the project, who asked not to be identified because the price information was confidential.”
The Facts: The exact price is confidential, but the number quoted is significantly higher than the actual power price in the CVSR agreement is overstated in the article by a significant amount.

Commentary: Since the NYT estimate came from somebody who supposedly knew the project economics, it would be interesting to understand why NRG's figure and that person's figure diverge.  Furthermore, the plant is delivering clean kWh, so one would want to benchmark its levelized cost to consumers versus other options for clean power.  From a public policy standpoint, we'd want to compare options on a pre-subsidy basis, not post (as this figure is) to evaluate total societal investments for different options to generate clean power.  Such an evaluation needs to include demand-side reductions, because not using energy is also quite "clean".  Most federal subsidies, however, are targeted towards increased supply.

The NYT says: “By 2015, NRG expects to be earning at least $300 million a year in profits from all of its solar projects combined,”
The Facts: NRG expects to make pretax profit of approximately $49MM (as disclosed on slide 40 of our 3rd quarter earnings call presentation.) The $300MM referenced in the article is actually EBITDA: Earnings Before Interest, Taxes, Depreciation and Amortization, not profits. The interest that we pay on the projects generates revenue for the American

Commentary:  EBITDA is often used as a way to evaluate the economics of an investment independent of the influences of capital structure or tax laws.  That is why NRG used that value, and not the net profits number, in its call with investors.  Among the influences of tax accounting, for example, are very favorable depreciation schedules for solar (up to a 100% write-off in one year, depending on when the plant goes into service).  This favorable write-off is both a subsidy, and one that would result in large differences between EBITDA and net profits -- particularly in the earliest years of an investment. 

Fact: The economic analysis reflects a failure to understand the facts behind the deal. The economic analysis of equity returns confuses after-tax and pre-tax returns and levered and unlevered returns. The analysis should include not only the tax benefits of the project, but also the income taxes that will be paid by the project over time. The analysis provided only includes the tax benefits with none of the costs. It is the equivalent of saying someone’s take-home pay is equal to their annual salary plus the standard IRS tax deduction, while ignoring the income tax paid by that person against which their tax deduction is credited! All of us know it doesn’t work that way for individual taxpayers. And it doesn’t work that way for solar projects in the California desert.

The operations and maintenance assumption is too low by more than half. It does not take into account additional operating costs of the plant such as: replacing inverters, lease payments, operational insurance, and asset management costs - among others.

Commentary:  All of these factors are relevant, sort of.  The fact that the NYT used input from Booz & Company suggests they may have integrated some of this to a greater degree than NRG alleges, as Booz analyzes many, many projects.  As noted above, if you have very large tax subsidies, the after-tax return may be a less accurate representation of project economics and returns than looking at pre-tax returns.  In terms of leverage, one would want to look at returns based on the capital structure of the project (mostly inexpensive debt because of the loan guarantee) and compare it to the capital structure and capital costs without the government subsidies to benchmark the level of government support.  Yes, income taxes paid by the project are relevant -- though may be less significant than the large subsidies through the loan guarantees and grants in lieu of tax credits.

Fact: Our projects will also help to put our country back to work. The CVSR project creates 350 high-paying construction jobs and combined, NRG’s large-scale solar projects will result in 3,840 direct and indirect construction jobs. Ivanpah, the largest solar thermal project in the world, enabled by the DOE loan guarantee program is an engineering and technological marvel and it is being built by more than 600 American pipefitters, laborers and construction workers—many of whom haven’t worked for up to two years prior to this job. They are proud of what they are building, confident in the future that they are creating and thankful to be gainfully employed.

Commentary:  The workers should be proud of what they are building, and having people back at work is a good thing.  But there are many ways to create jobs, and if jobs are the primary policy driver we ought to be looking for the most efficient ways to accomplish that goal.  Public statements on job creation by the recipient company are not a reliable indicator of the efficiency of job creation.  It is useful to point out as well that the authorizing language for 1705 in ARRA added a requirement that much of the labor used to build the plant be at prevailing (union) wage levels under the Davis-Bacon Act.  This drives up the cost of construction, and may also reduce the number of people employed as a result.

Fact: These projects are not possible without the DOE support. No solar PV projects of this scale, $1 billion and above, have gone forward without DOE assistance; these projects involve more capital than the private sector will finance.

Commentary:  There are many reasons why this may be the case, not all of them supportive of an argument that the government should be financing them.  There may be high risk relative to other (even other clean) options, for example.  Or limited expectations that learning from the initial one or two plants will bring down installed costs for future facilities.  One nice element of Renewable Portfolio Standard approaches (assuming they are the only subsidy at play and not added to a slew of other subsidies) is that you can buy clean energy competitively.  Even if you need to pay a premium relative to conventional energy resources (perhaps even fair, given that conventional fuels cause large external damages not reflected in their prices), an RPS keeps the delivery risks of building and operating the new facility with the private sector.  Further, investors are forced to accept the smallest subsidy possible through a competitive bidding process -- a useful process for "discovering" the real amount they need to achieve target returns.  Other programs merely guess at this level, and through subsidy stacking often overshoot what's needed by a large margin.

Fact: NRG is investing more than $1 billion of our own capital as equity in these massive solar projects. It is important to remember that we are a business with a fiduciary obligation to our shareholders, and we do need to make a return on the investment. We are investing more than $1 billion of our own capital in solar projects to advance technologies, meet customers’ needs for renewable power, and create jobs. If we build and operate these projects in accordance with our
own expectations of cost and performance, we expect to realize solid equity returns on our investment on a risk-adjusted basis.

Commentary:  It is good that NRG is putting some of their own capital at risk, and $1 billion is a large number from the perspective of normal consumers.  However, in terms of industrial investments (this figure applies to multiple projects, not just CVSR), the gross number may not be that meaningful.  I'm more interested in what percentage of each project is long-term risk capital by NRG; and in ensuring that the subsidies that are paid out are given over time based on the successful completion and economic operation of a project rather than front-loaded in the project's life span.  Front-loading dilutes the incentives for careful investment and operation over the long-term.

Fact: These projects are good for the environment. NRG’s large-scale solar projects would displace 1.7 million metric tons of greenhouse gases annually and result in dramatically improved air quality compared to fossil fuel alternatives. Big solar projects represent the future for our company and for our society. Solar power represents a sustainable, inexhaustible, zero-emitting, domestic energy resource that can be installed on every rooftop and every parking lot in every city and hamlet across our great country. The International Energy Administration announced that, on our present course, climate change will be irreversible by the year 2017. The public and private sector need to work together to do more, not fewer, of these large renewable projects.

Commentary:  No argument that solar power is cleaner.  But as with jobs, society needs to buy the most of its desired product that it can with a limited budget.  One can agree with the premise that we need to reduce our carbon footprint dramatically (as I do) while not necessarily concluding that the particular investment evaluated by the NYT was an efficient way to do it.

Fact: These projects and government support are driving down the cost of solar energy. The cost of solar power is falling dramatically, in large part because innovation, deployment at scale and competition are forcing inefficiencies out of the supply chain, and government support for solar power across the federal, state and local level and across both political parties has been instrumental in achieving those declining costs.

Commentary:  There are many things that may be driving down the cost of solar energy.  If I had to pick the markets I think are most important for this to happen, it would be distributed solar PV and solar hot water.  It would not be large scale solar plants, though I recognize others may have a different view.  Scaling of production is a good thing, as is the pace of innovation.  But it is not clear that dumping so much money on one facility (NYT: "As NRG’s chief executive, David W. Crane, put it to Wall Street analysts early this year, the government’s largess was a once-in-a-generation opportunity, and 'we intend to do as much of this business as we can get our hands on'") is the way to get there.

Fact: The government loan guarantee program is a loan program. Loans received by these solar projects, such as the California Valley Solar Ranch, are indeed loans that will be repaid (with interest) for the benefit of the American taxpayer.

Commentary:  This is standard borrower rhetoric.  These firms have relationships with lots of banks, and borrow as well on capital markets through share issues.  They are choosing to borrow under this particular program because the terms are better than what they can get elsewhere.  This means that even if taxpayers get funds back, we will receive a below-market amount of interest, and sometimes even less than the Treasury's cost of borrowing.  And while we should take NRG's statement that this particular funding is "indeed loans that will be repaid..." as a sign of their goodwill, if the debt really were risk-free, they could have borrowed at attractive terms elsewhere.

Natural gas fracking well in Louisiana

Ben Gemen over at The Hill describes as "hardball" the effort by House Republicans to bring to light documents and decision making behind a $535 million loan guarantee for a new facility to be built by Solyndra. Hardball is appropriate for an unprecedented level of taxpayer exposure for investment into individual, privately-owned assets. 

I hope the Congressional effort is successful in teasing out real documents that would allow real oversight of DOE's loan guarantee program.  DOE's Loan Programs Office has been notoriously resistant to transparency, despite operating at a scale that already runs into the tens of billions of dollars of commitments ($39.8 billion as of July 2011).  Proposals to extend this federal-financing approach through a "clean energy" bank of sorts have been floated with figures in the hundreds of billions. This is not a program to be glossed over.

Gemen reports that

In a letter to [Rep. Cliff] Stearns Tuesday, OMB Deputy General Counsel William Richardson, Jr. notes that OMB "made available" 1,400 pages of emails and attachments this week, which follows hundreds of other pages made available earlier.

It is impossible from the outside to tell how material this information is, and whether the critical items on risk management, potential conflicts of interest, and likelihood of market success have been provided. 

Nonetheless, I'm guessing that the data dump on Solyndra is about 1,400 pages more than what has been provided for DOE's commitment to the Vogtle nuclear power project in Georgia.  At roughly $8.3 billion (which is really a direct loan, as program rules require a loan of this scale come from the Federal Financing Bank), DOE's support for Vogtle is nearly 16 times as large as what taxpayers are risking for Solyndra.  Surely if a subpoena is warranted for gathering data on the Solyndra project, it is warranted for Vogtle as well.

I have been critical of the weak structure of DOE's foray into large scale loan guarantees for energy infrastructure since the program's inception.  It is not as though the Department has a robust and successful history of this type of capital deployment.  Here are the my formal comments to DOE on the proposed program structure back in 2007.  They were ignored then; DOE focused mostly on loosening program rules during that round of comments based on input from interested investment banks, firms that did or would represent many of the beneficiaries of the guarantee program itself.  Some of these banks (e.g., Merrill and Lehman) imploded soon after due to the very types of conflicts of interest and weak oversight that remain a concern in the DOE lending initiative. 

Though DOE remains under pressure to push money out the door, and in the process continues to ignore most of these structural issues today, the problems -- and the financial exposure -- remain.  It would be a service to all taxpayers if Cliff Stearns extended his interest in fiscal prudence and oversight beyond a single project and to the program overall.