Worse than this? Yes, but in a different way.
In 2010, the $2b Shepherds Flat wind farm project in Oregon received a federal Section 1603 cash grant of about $500m, among other subsidies. What did federal taxpayers get out of this? The usual neoliberal narrative about the environment, climate change, jobs and all that, as was the style at the time. The private equity developers got a more tangible return — about a 30% annual ROI, according to a WH analysis, which was likely realized as a nice profit in 2021 when the project was sold to another private equity firm.
There was some criticism about the degree to which this project was subsidized by federal and state sources. But that was memory-holed soon enough. There was no scandal. It was (and is) well understood that government grants rarely result in national public goods and mostly end up as transfer payments to specific interest groups. Those on the receiving end grab everything they can, in accordance with the law and prevailing societal norms, and no one expects they’d be ‘any better than they should be’.
Around the same time, Solyndra received a $535m DOE loan. That was about the same amount of one-way federal money as went to the wind project, but the different form matters. If Solyndra had instead received a $535m grant, and subsequently failed completely, there’d have been criticism, but no scandal. As a loan, however, there was an expectation of repayment based on some sort of creditworthiness, presumably attested to by the promoters and carefully assessed by federal experts. The fact that the ‘loan’ effectively turned into a worthless ‘grant’ so quickly suggested that deception was involved in the transaction, either by those doing the deceiving or those allowing themselves to be deceived. Or both. No one went to jail, no federal employees were fired, but the appearance of deception was enough for a scandal.
EPA WIFIA and Mandatory Spending
In this context, let’s look at EPA WIFIA’s mandatory spending. This off-budget cost is now about 9% of the current portfolio, or 9 times higher than the frequently touted 1% in discretionary appropriations that Congress was led to believe that the loans cost. Still, the dollar amounts involved aren’t that high relative to other federal surprises — maybe $2 or $3 billion. Water infrastructure is not controversial. And the money likely ended up in lower water rates for some folks (admittedly, regardless of need) as opposed to mansions or private jets.
However, there is another perspective. In retrospect, it’s obvious that cancellable/resettable loan commitments to Aa3/AA- water agencies with excellent financing alternatives in the tax-exempt bond market would be used as ‘free interest rate options’, and that this behavior would cause high levels of positive interest rate re-estimates covered by mandatory spending at WIFIA. But how different is that from saying, in retrospect, the Solyndra ‘loan’ obviously wasn’t a loan at all? It’s an unnecessary observation — the issue is why it wasn’t obvious at the time the loan was made.
Maybe WIFIA’s high mandatory spending should have been obvious as soon as the program started receiving applications from big Aa3/AA- water agencies who didn’t need the money? Or when they used short-term financing for construction funding, keeping the loan commitment as a free hedge? Or when they demanded interest rate resets and had the political influence to get them?
The ‘free option’ aspect of a WIFIA loan commitment was obvious to me while I was an SGE at EPA back in 2019. A little later, when I crunched the limited, publicly available numbers, the scale of re-estimates was also predictable, with relative precision, depending on interest rates at the time the loan commitment was drawn. [1]
An Origin Story
Still, the mechanisms of ‘interest rate re-estimates’ and ‘free options’ are not exactly easy to understand, so we have to ask, ‘obvious to whom?’ I assume that the big Aa3/AA- water agencies definitely knew from the start how WIFIA loan commitments could be used for interest rate management. In fact, it’s possible that this understanding was key to the enactment of the program in the first place. I can imagine that some sharp public finance advisors saw the rate lock feature in the TIFIA and thought ‘this has limited value in transportation projects, but it could be immensely valuable if available to US municipal water agencies’. [2] Next stop — water lobbying groups like AWWA and NACWA, who probably didn’t understand the purpose very well, but were of course responsive to what their members were asking for. And then to Congress, where the ‘narrative’ about water infrastructure and trivial cost in discretionary funding was all that mattered.
There were a few bumps in the road. SRF lobbying groups objected to the competition but were assuaged by some right-of-first-refusal language that would not be relevant in actual use. CBO/JCT, likely following the strict logic of the narrative, thought that WIFIA loans would unleash a backlog of projects and increase the issuance of tax-exempt debt, and so scored the legislation that way — they were put in their place at some point, and the final statute had no restrictions on the borrower’s use of tax-exempt debt. I originally wondered why the tax-exempt bond market didn’t object to the competition — but now I see that the water agencies are big issuers in that market and could be allowed a few indulgences in an ‘occasionally useful’ option.
Bureaucratic Excuses, ‘Above the Waterline’
So far in this ‘origin story’, most of the players were just doing their jobs, as that is understood in a narrative neoliberal world. Or, like CBO/JCT or Congressional staff, so disconnected from the full picture of the program’s operations and purpose that they can’t really be blamed. But now we come to program implementation by EPA staff and program oversight, primarily by OMB. Both of these federal bureaucracies would have been at the frontline of in-depth loan application evaluation and familiar with detailed program numbers. It is with these that questions about the obviousness of WIFIA’s actual purpose begin to matter.
On one level, they have defenses. For EPA staff, the statutory requirements of the program focused almost exclusively on eligibility and creditworthiness. By these standards, loan applications from big water agencies could not be questioned. Likewise, estimates of the cost of loan commitments were precisely defined by FCRA law — the (trivial) credit risk was duly reserved for with discretionary appropriations, and re-estimates automatically funded by mandatory appropriations. The process was purely mechanical and no doubt followed to the letter.
More broadly, the implementation of a loan program goes beyond loan application processing and servicing. It is naturally assumed by federal staff that the policy objective and public good outcomes of a loan program have been baked in the legislative and rule framework. Part of the EPA’s job was to grow the program wherever possible within this framework as a way of achieving ‘more public good’. Of course, perennial bureaucratic ambition would also play a part, but this is not a problem if everything was done within the rules. WIFIA staff did in fact do a lot of refinements that made program loans even more useful to big, highly rated water agencies with large, long-term capex programs (amortization sculpting, master agreements, the resets, etc.), but these were likely the ones permitted or at least not prohibited by the statutory framework. In effect, these refinements were their only avenue to WIFIA growth and sustainability, not necessarily intentional ‘aiding and abetting’ of a select group (though I’m sure the AWWA and NACWA cheerleading wasn’t unwelcome).
Even OMB has — again, on one level — defenses against Solyndra-type criticisms when it comes to WIFIA’s mandatory spending. OMB vets the forward-looking projections of creditworthiness in loan applications — for which they were criticized in the Solyndra deal but an easy job for WIFIA loans. However, AFAIK, they don’t ‘project’ the eventual budgetary consequences of anything else. As long as FCRA law and their own budget procedures are scrupulously followed, and the results accurately reported, I think OMB doesn’t need to go further. This seems especially true for the somewhat abstract and years-delayed result of loan commitments used for interest rate management — the consequent mandatory spending only shows up in scale once the commitments are eventually drawn and re-estimates triggered towards the end of a long-term construction phase. Hard to ask questions about this — and easy to ignore.
In short, for both EPA and OMB staff, obligated in a strict sense to look only at results ‘above the waterline’ of WIFIA activity, the emergence of huge mandatory spending amounts in the last two years is not likely to be a source of a Solyndra-type scandal based on accusations of deception or gross negligence. Again, everyone appeared to be ‘doing their jobs’ and the unfortunate results are, well, fairly typical for federal programs. $2 or $3 billion? It could have been worse.
Mandatory Spending as a Symptom of a Deeper Issue
Now let’s consider another, but more fundamental perspective. What if WIFIA’s mandatory spending should be seen, not as an important thing in itself, but as a symptom of a deeper failure in US federal water finance policy? And that implementing and sustaining that failure involved real deception?
Here is the key point: the value of WIFIA’s interest rate management features (e.g., the rate lock and reset) relies primarily on the fact that the program borrowers have excellent financing alternatives. That way, they can cancel a loan commitment as soon as it’s ‘out-of-the-money relative to those alternatives — in the case of tax-exempt bonds, which trade at near-UST rates for such highly rated issuers, that’s a pretty low bar. The credible threat of cancellation (with all the political noise) is what drives the reset feature.
This generous scheme wasn’t difficult for EPA to implement since the central features were statutory, as copied from TIFIA. It appeared ‘costless’ and I think many at the program did in fact believe that some sort of magic was being performed by WIFIA loans. High re-estimates were understood by the senior folks, but as discussed above, these delayed and arcane consequences could be ignored. [3]
But there was a serious difficulty, right from the start. How to characterize what the WIFIA program was actually doing?
Part 2 will discuss the nature and cost of the EPA WIFIA’s narrative camouflage
___________________________________________________________________________________________
Notes
[1] Readers may ask why I didn’t raise the issue more vocally at the time. In fact, I talked to everyone who might listen, both in the program and in various oversight and even advocacy groups. I have the email receipts, and of course all the posts and detailed analyses I did from about 2020-2023. But it is true I didn’t try and raise it publicly in articles until very recently. My reasoning was that the issue would provide a spur to the development of more economic features for WIFIA loans (e.g., by amendment bills), as a natural progression from a ‘start up’ phase (where anything that creates loan volume might be considered acceptable) to a better and more sustainable model for federal infrastructure finance. I thought that existing stakeholders would see this also in a self-interested light, if explained carefully enough, and go along with it. Now I see that I was wrong — the CWIFP non-FCRA policy riders, WIFIA’s unnecessary funding and the lack of any amendments in the FY 2026 spending bills were the last straws. EPA WIFIA is not reformable.
[2] Anecdotal data point from circa 2019: While an SGE at EPA, I was at a lunch with two Goldman Sachs executives. The discussion was largely about the new WIFIA program but only in vague terms — except when the older and somewhat more flamboyant executive said, in sotto voce, ‘well, of course, the only real value is the rate lock’, as if to impart an arcane secret, more out of insider vanity than anything else. I saw the younger executive, a more disciplined-looking neoliberal apparatchik, stiffen a bit at this. I already knew the basic mechanisms of the loan program at that point, so the information wasn’t a surprise, but I do remember wondering about the incident. Now it seems easier to understand.
[3] This is true, in my direct knowledge. The conversations were uncomfortable, with a hint of warning not to pursue the topic any further.