This post looks at CBO’s rationale for rejecting the WIFIA FCRA amendment language in S.3591, a 2020 Senate bill. This is worth revisiting in some detail for two reasons. The first is that the soon-to-be-reintroduced HR 8127 and current HR 2671 House bills use very similar language in their FCRA amendments as S.3591 did [1]. If CBO scores that language again at some point in the process, presumably they’ll apply the same rationale.
The second, more fundamental reason, is to review the New Approach in a context of the amendment proponents’ intentions and CBO’s basic reasoning.
The analysis begins on page 3 of CBO’s 11/20/20 Cost Estimate (emphasis added):
S. 3591 would limit the criteria used to determine the budgetary treatment of the anticipated net cost or savings of loans or loan guarantees made under EPA’s Water Infrastructure Finance and Innovation Act (WIFIA) program.
This is indisputably correct on the surface, because that’s the way the amendment language is written. But did that language accurately reflect fundamental intent? Was the amendment intended to limit any additional FCRA criteria being applied by WIFIA for loans to federally involved projects? Or just definitively get rid of the seriously flawed current Criteria? I think it was the latter, primarily because by that point the issue had become (and remains) contentious, and the amendment’s proponents likely wanted to stake out an unambiguous position. That’s understandable, especially because FCRA law and principles seem so clear that’s it’s hard to imagine why additional criteria would be needed other than to keep the issue in a bureaucratic rabbit hole. It took me a year of thinking about this issue to come up with a possible, if vanishingly rare, scenario that appears to require additional FCRA criteria, as described in the in the New Approach post.
If getting rid of the current Criteria was and still is the sole goal, the first task of the New Approach is to show amendment proponents why additional criteria are (at least theoretically) necessary and how they can be incorporated in WIFIA’s statute in a way that won’t impede truly eligible applicants. Admitting to CBO and other oversight agencies that additional criteria are necessary is not only objectively valid but a critical first step in making progress on this issue by establishing some common ground.
Now back to the CBO side. The paragraph continues:
Under the bill, any budgetary impacts would be recorded on an accrual basis if the borrower is a nonfederal entity and would be repaying the obligation with nonfederal funds. This provision would allow the costs or savings for loans to federal projects that meet those two criteria to be recorded on an accrual basis.
The two criteria, a non-federal borrower and a non-federal repayment source, are all that FCRA law and principles would seem to require in an honest world. The 1967 Report makes it crystal clear that the purpose of FCRA is very narrow: to get non-federal loan repayment cash inflows out of the cash budget due to their distorting effect and gaming potential. How many criteria are required to determine that a large investment-grade loan to a sophisticated non-federal borrower with an extensively analyzed non-federal repayment source is exactly what the 1967 Report was talking about? Not many. Absent the current Criteria, I think CBO would find it as difficult as the amendment’s proponents likely did to imagine exactly what additional criteria are needed. In fact, the difficulty of the exercise is probably what led OMB down the rabbit hole in the belief that there had to be something there that could be used to limit statutory eligibility.
The final sentence in the paragraph looks like an innocuous technical reminder:
That budgetary treatment [FCRA accrual] is not allowed for federal projects under current law.
In fact, it’s quite misleading, though probably unintentionally so. FCRA accrual treatment isn’t allowed for anything other than intangible federal financial assets, certainly not manifestly tangible federal projects. I think what’s meant here is that FCRA treatment is not allowed for federal loans that finance federal projects. But that’s also misleading. As a matter of fundamental statutory eligibility and policy purpose, a federal loan program like WIFIA should never be financing a federal project in the first place. To suggest that FCRA treatment is the program’s sole restriction on loans to finance federal assets, and that inadequate WIFIA FCRA criteria will therefore open the floodgates to all sorts of intra-agency mischief, is simply nonsense.
CBO knows this. What we’re seeing in the above sentence is an echo of the current Criteria’s rabbit hole, where questions are ‘criteria’, FCRA budgeting is applied to projects, a federally involved project is almost always a ‘federal project’ in accordance with undisclosed consolidation principles, consolidation is the only relevant topic for FCRA in the 1967 Report, law can be amended by footnoted diktats, and on and on.
This points to the next task of the New Approach: to demonstrate to CBO that the current Criteria are seriously and fundamentally flawed by proposing criteria that are in fact clearly based on FCRA law and foundational principles. CBO will never openly admit it, of course, but they don’t have to. Raising doubts about the validity of the current Criteria should be enough for CBO to quietly move on from their 2020 scoring, especially if a better alternative is proposed.
The next relevant paragraph again echoes the Criteria, but two specific phrases can be used as the entry point of New Approach concepts:
However, the status of a borrower as a nonfederal entity repaying a loan with nonfederal funds is not a sufficient basis for the loan or loan guarantee to receive FCRA treatment under current law. In directing this budgetary treatment under S. 3591, EPA could make loans and loan guarantees for federal projects or assets and record the costs on an accrual basis—which would be reflected in a subsidy cost—rather than on a cash basis, thus understating the initial funding required for those commitments.
The New Approach agrees with CBO that a non-federal borrower and non-federal repayment source are a necessary, but not sufficient basis, for FCRA classification. The applicant has also to demonstrate that its proposed loan is free from crookedness and imbecility [2]. That shouldn’t be hard. If it is, rejection is probably warranted for reasons that go far beyond FCRA budgeting.
The demonstrations of value and independent decision-making will also confirm that the asset being financed by the WIFIA loan is non-federal because why else would a non-crooked, non-imbecilic non-federal borrower be paying for it?
It is as simple as that. If an honest and intelligent non-federal entity spends $100 million on a federally involved project in its own interest, there will be a $100 million non-federal interest in that project. It’s that asset WIFIA and CWIFP loans are financing, not some sort of mysterious ‘federal’ Chesire cat in the rabbit hole.
The final and most important task of the New Approach is to focus CBO and other oversight analysis on the intrinsically non-federal nature of an asset that a non-federal borrower pays for. That will bring us back to the real-world where FCRA classification, with a few minor confirmations, is as straightforward as the 1967 Report and FCRA law intended it to be.
The Numbers
If CBO is persuaded that the additional criteria proposed in the New Approach are sufficient for FCRA budgeting purposes, their cost estimate should go back to the accrual basis normally applied to WIFIA and other loan programs [3]. The amendments themselves shouldn’t cost anything, as they simply enable the utilization of funding already provided by Congress and I don’t think CBO has ever scored that on the basis it wouldn’t be used.
Still, a couple of things are worth noting. CBO’s 2020 cost estimate assumed that if S.3591 was enacted, there’d be restored eligibility for $150 million per year of WIFIA loans to federally involved projects, which seems low. Now that CWIFP is being implemented and has funding for dams and levees, the estimate looks unrealistically low, based on the potential borrowers I’ve written about here and especially here. Might CWIFP’s implementation change something in CBO’s cost estimates for new WIFIA law? I can’t think of anything at this point, but perhaps the change in WIFIA’s scope since 2020 ought to be kept in mind.
Interestingly, CBO had a view of what would have happened to the current Criteria if S.3591 had been enacted. Here’s their footnote 5:
The Further Consolidated Appropriations Act, 2020, provided subsidy budget authority for the WIFIA loan program but required EPA to develop criteria to determine project eligibility and apply those criteria for projects selected in the 2020 cohort. CBO estimates that [the current Criteria] would effectively prohibit EPA from selecting projects for WIFIA loans that would receive accrual treatment under S. 3591 but cash treatment under current law. CBO expects those criteria to remain in place through fiscal year 2021. As a result, CBO estimates that the provision would have no effect on direct spending for funds provided in that fiscal year.
They seem to be thinking that the Criteria can remain in force even after new WIFIA law is enacted, at least for a period. I wrote about this scenario in the last section of the New Approach post. But what’s CBO’s basis for assuming that the Criteria go away after one year? Automatically, or are they assuming some additional future action? I wish they’d explained their thinking — perhaps they’d be on board with the kind of graceful exit recommended in the New Approach?_____________________________________________________________________________________________
Notes
[1] HR 8127 and HR 2671 have the same language:
‘‘If the recipient of financial assistance for a project under this subtitle is an eligible entity other than a Federal entity, agency, or instrumentality, and the dedicated sources of repayment of that financial assistance are non-Federal revenue sources, such financial assistance shall, for purposes of budgetary treatment under the Federal Credit Reform Act of 1990 (2 U.S.C. 661 et seq.)—
‘‘(1) be deemed to be non-Federal; and
‘‘(2) be treated as a direct loan or loan guarantee (as such terms are defined, respectively, in
Act).’’.
The Senate bill which CBO scored, S.3591, differs only in referring to “the project or asset for which financial assistance is being provided” in place of “such financial assistance“. As discussed above, this older language is incorrect (FCRA doesn’t apply to projects, only loans), probably echoing (as CBO’s statement did) the Criteria’s basic error. The fact that it’s fixed in the newer House bills is positive, of course. I doubt that change in itself would lead to a different conclusion, since presumably CBO will again apply the same Criteria-based framework, other things being equal. But the change (and especially why it was considered necessary) is a good opener for a discussion that’s very much on the right path.
[2] It’s worth noting that the relevant crookedness here is fundamentally related to an exercise of federal sovereign power, apparently an important factor for CBO’s general scoring principles. That should be highlighted to help establish some common ground outside the current Criteria.
[3] Unfortunately, that will probably include CBO’s previously used assumption that a WIFIA loan for 49% of the assets being financed will necessarily entail the issuance of additional tax-exempt debt for the 51% balance, leading to a hit on federal tax revenue. While that’s demonstrably not true for most EPA WIFIA loans to date, it might be more valid for CWIFP’s lower-rated and more contingent projects. In any case, a separate issue for another day.