Yes, WIFIA’s FCRA budgeting issue with federally involved projects is important. Everyone recognizes that. Now, what about this one?

Here is more information. But don’t talk about it, okay?

• FCRA Re-Estimates and the Anti-Deficiency Act

• The Cost of WIFIA’s FYE 2022 Portfolio (TL/DR Version)

• Estimating the Cash Cost of WIFIA’s FYE 2022 Portfolio

• The Economic Cost of WIFIA’s Portfolio at FYE 2022

The WIFIA loan program is currently divided into three areas, (1) the ‘original’ and apparently successful WIFIA for state & local water agencies, (2) the currently small and only slightly differentiated SWIFIA section, and (3) the Corp’s new co-authorized program, CWIFP.

What is the reason for this division? Simply a bureaucratic matter within a unified statutory framework? Loan product branding for different potential borrowers? Echos of legislative history?

Perhaps all of those, and others like them besides. None are as important as the fundamental differences emerging between the divisions:

Irreconcilable Differences

Here’s a controversial idea: Regardless of its original groundbreaking role, State & Local WIFIA is now holding back the development of SWIFIA and CWIFP:

• State & Local WIFIA loans will inevitably compete head-on with the municipal bond market. But that market can’t beat WIFIA’s free interest rate options. This is why Congress is so parsimonious when it comes to WIFIA appropriations — a small fraction of what TIFIA and CIFIA get, despite the water program’s rapid growth and apparent success.

• Why don’t State & Local WIFIA stakeholders fight harder to expand the program? Because they’re also (and even more so) stakeholders in the municipal bond market. In theory, a better approach might be for them to promote synergies between the two sources of infrastructure financing, expanding and improving both. But in the real world of entrenched monopolies, long-established relationships and zero-sum solutions, their prioritization of one over the other is a rational choice.

• Because State & Local WIFIA was the original version and has a track record of delivering results, their stakeholders will naturally be more organized, focused and energized than those of the new, as-yet unutilized divisions. Any proposals that could upset their part of the WIFIA applecart will meet effective opposition. The SRF WIN Act’s ambitious plans in 2018? They might have posed a threat to State & Local WIFIA’s détente with muni bonds, in addition to the usual resistance of any group to a dilution of its influence. Despite broad-based (but less organized?) support, the Act got reduced to a consolation prize.

• It goes beyond appropriations. Loan programs can expand their capabilities not just with increased funding, but by loan product development as well. State & Local WIFIA has done a good job (perhaps too good) in enhancing and expanding the scope of their interest rate-lock products. But its stakeholders won’t back product development that isn’t important to state and local water utilities and agencies. 55-year loan term for long-lived projects? Not usually relevant to our projects. Improving OMB’s FCRA non-federal criteria? Not our problem. Sub-Treasury rates to incentivize small SRFs? Our highly rated borrowers are quite happy with their free interest rate options. And so on.

Unstable Disequilibrium

Well, the federal government is filled with sub-optimal programmatic outcomes. Why should the trajectory of the WIFIA program be any different? It’s easy to imagine that State & Local WIFIA will soon plateau into a kind of captive US ExIm Bank for minor transfer payments to a sub-sector of water infrastructure, exactly to the extent allowed by the municipal bond market. SWIFIA and CWIFP will remain as its stunted and ineffective satellites, kept around for political convenience and to prevent other federal loan programs occupying the space, but not given the funding or loan capabilities to succeed.

Perhaps. But I don’t think so. Even for federal programs, a more-or-less permanent sub-optimal outcome requires the absence of destabilizing factors. There are at least two of these in the background:

• Most fundamentally, the growing needs of the sectors that SWIFIA and CWIFP serve. A few more years of failing infrastructure, economic difficulty and a changing climate can quickly become a perfect storm. Obviously unmet needs will coalesce and energize the sectors’ stakeholders to demand that their existing loan programs be given the required funding and capabilities, and they’ll be prepared to take on State & Local WIFIA’s stakeholders to get them. Now add in regional and political overtones (a ‘Cross of Muni Bonds’ speech?) and the issue goes beyond the technicalities of federal loan programs. I think some movement in this direction has already started.

• Most explosively, State & Local WIFIA’s exposure to huge funding losses. I’ve written about this frequently in other posts, so I won’t go into the details. The important point here is that State & Local WIFIA’s success is not real. Its $16 billion portfolio of executed loan commitments will almost certainly cost far more than the program’s Congressional appropriations to date, and taxpayers are the hook for the difference. If this unexpected cost becomes another Solyndra-type issue (and I think it will), then State & Local WIFIA’s influence will be significantly diminished, and SWIFIA and CWIFP stakeholders will be motivated to distance their programs from the mess. I don’t know how this type of destabilizing event will end, but a reversion to the status quo ante is not likely.

An Unnecessary Intra-Sectoral Conflict

There’s no need for all this. The current and potential problems with WIFIA’s divisions highlight the inadequacy of a purely sectoral approach to federal infrastructure loan programs. They’ll each have their own stakeholders that are quite rationally concerned with their own sector, without regard to the bigger picture. Worse, if a ‘sector’ is defined too broadly, there’ll be competing stakeholders within a single loan program. Without a non-sectoral authority to sort out the issues and develop equitable solutions, the strongest stakeholder in the program will prioritize its own interest, to the detriment of the others, as I think we’re seeing at WIFIA.

For now, SWIFIA and CWIFP stakeholders have no choice but to focus on their respective sectors. Paradoxically, however, one way to advance their causes may be to support a more unified approach to federal infrastructure loan programs. Cross-sectoral loan program stakeholders will naturally have an interest in SWIFIA and CWIFP growth and development. But perhaps in this case, given the relatively extreme degree of sub-optimality caused by sectoral differences, they’ll also begin to recognize their potential central importance.

In a prior post, I explained why adding a limited interest rate buydown provision to WIFIA’s statute would make sense for the Corps’ new loan program, CWIFP.  The limited buydown provision allows an infrastructure loan program to lower (within limits) a loan’s execution interest rate to what it would have been on the day the loan application was accepted.  It’s a potentially significant benefit for projects that have a long development phase during a time of volatile interest rates.  CIFIA (the primary focus of several earlier posts on this topic) has this provision, as does TIFIA, but WIFIA currently does not.

Adding the limited buydown to WIFIA might also make sense for another major stakeholder of the loan program – state revolving funds, or SRFs.  Obviously, WIFIA loans to SRFs are quite different than those to large infrastructure projects.  SRFs have a few special ‘SWIFIA’ provisions within WIFIA which are designed facilitate lending to the funds, and a dedicated (albeit small) amount of credit subsidy.  But the more fundamental difference is about policy objectives – while a WIFIA loan to a large infrastructure project is intended to accelerate its development and completion, the purpose of a WIFIA loan to an SRF is to encourage the fund to leverage its loan portfolio, effectively increasing the impact of its state and federal grants.

Encouraging Leverage at SRFs

Although all the classic reasons for leveraging a loan portfolio would seem to apply, most SRFs do not utilize much, if any, external leverage [1].  There doesn’t seem to be a fundamental reason for this, though a lack of administrative resources might be a common factor [2].  SRFs in more densely populated states receive a larger share of federal grants (per the CWA’s population-based allocation) than those in rural states and can accumulate a capital base that supports the staffing and transactional economies of scale required to issue and manage tax-exempt bonds.  In theory, since a WIFIA loan is in effect a single-lender private placement with an interest rate roughly comparable to a tax-exempt bond, it should be an easier starting point for smaller SRFs considering leverage.  In practice, however, the only WIFIA loans to SRFs to date have been for two large funds that already issue bonds.

Perhaps in acknowledgement that policy objectives for SRF leverage weren’t being realized under current law, the SRF WIN Act of 2018 attempted to make WIFIA loans more attractive to smaller SRFs. The Act included significant SRF-specific funding and various features, including sub-Treasury interest rates for smaller SRFs.  Despite broad-based support, there was serious opposition from major WIFIA stakeholders for whom the program was already working well enough.  I don’t know why these stakeholders perceived the expansion of a successful loan program’s capabilities as a zero-sum game. Doubtless the story is complex.  In any case, the opposition largely prevailed, and the Act was whittled down to the minor provisions enacted in SWIFIA.

Focusing on the First Step

In this context, the limited buydown might be seen as the type of feature that might have been included in the SRF WIN Act, though with a more subtle effect and a lower-key profile.  As with many things in life, the first step for an SRF to leverage its portfolio is probably the hardest.  At WIFIA, the process involves three steps over the course of at least two years – a letter of interest, an application, and loan execution.  Under current law, only the most difficult and costly part, an executed loan, has any certain value.  With a limited buydown feature, however, an accepted application itself has some potential value by setting an interest rate (albeit within limits and subject to the program’s agreement) for future leverage.  That’s valuable to help the SRF’s planning and decision-making processes with more specific numbers, and perhaps also to visualize how the leverage would work in more concrete terms.  If Treasury rates start rising, the application’s potential value becomes intrinsic and measurable – and perhaps significant.  If not, and the SRF decides to discontinue the process for whatever reason, there are no penalties for withdrawing an application.  The sunk cost at that point is limited to the application fee of $100k and relatively minor staff resources.

In effect, the limited buydown makes a successful WIFIA application a small but realizable goal that will potentially improve an SRF’s future leverage if the fund decides to go forward, but not cost very much if it does not. In turn, this makes the decision to take the first step — sending in an essentially costless letter of interest — more attractive because it has a potentially valuable near-term result that does not require a commitment to full loan execution. Yes, of course — the policy objective here is for more WIFIA loans to be executed and more SRF leverage to be in place. SRF WIN’s sub-Treasury rate on executed loans was a more direct approach to that. But the objective is also served by providing an incentive to get SRFs to start the process of considering leverage in the first place. I think adding the limited buydown provision to WIFIA can help accomplish that.

Limiting the Limited Buydown

It would be straightforward to simply cut and paste CIFIA’s limited buydown language and include it in the WIFIA statute. This has the advantage of utilizing a recently enacted precedent (CIFIA law was part of IIJA 2021) which in turn was based on the limited buydown language in WIFIA’s own precursor model, TIFIA. The cut and paste approach will likely work for CWIFP, which like CIFIA is intended for large, long-development infrastructure projects.

But for SRF policy objectives, it would probably make sense to add some limitations on eligibility for the provision. Large SRFs that are already leveraged with bonds or have the capital and staff resources to pursue a full WIFIA loan execution whenever they choose will naturally want to use the limited buydown as a potential enhancement for leverage they were going to do anyway. However, that’s obviously not consistent with the policy objective of encouraging smaller SRFs with limited resources to consider leverage that they might not have done otherwise. An SRF limited buydown provision should be designed with focus and clear additionality, in the same spirit as WIFIA’s provisions for small communities.

SRF WIN limited the availability of sub-Treasury execution interest rates to states that received less than 2% of federal SRF funding for the year. That’s definitely correlated to more rural states with smaller SRFs and could probably work as a rough filter for limited buydown eligibility. Other mechanisms could focus more directly on an SRF’s recent experience in leveraging or related current capabilities to exclude the ones that don’t need further inducement. There’s plenty of data available for a fine-tuned approach here and to characterize limited buydown eligibility as a data-driven technical refinement.

_____________________________________________________________________________________________

Notes

[1] A 2018 NRDC report, Go Back to The Well: States and the Federal Government Are Neglecting a Key Funding Source for Water Infrastructure, outlines the numbers on page 6: “Of the 28 states that have used bonding, just 12 are responsible for nearly 75 percent of the bond revenues generated…Only New York, Massachusetts, Ohio, and Indiana have regularly leveraged their SRFs through the sale of bonds…”

[2] A 2022 report from Duke University and EPIC, Uncommitted State Revolving Funds notes that a lack of administrative resources is an important reason why many SRFs have relatively high levels of uncommitted funds. If that’s the case for a SRF’s core loan-making function, then it’s likely even more true for leverage.

Federal loan programs for large-scale public infrastructure are focused on specific sectors.  And since the US doesn’t have a federal ‘Department of Infrastructure’, infrastructure loan programs are also situated in, and largely managed by, the non-financial federal agencies responsible for their sector.  Much of this makes sense in terms of specific policy objectives and for some aspects of program operational efficiency.

Infrastructure sectors have dedicated federal advocacy organizations and think tanks.  These are of course the primary source for proposing and developing loan programs designed for their respective sectors.  This natural arrangement has much to be said for it, too, especially with respect to energizing the sectoral stakeholders.

Common Elements Across Sectors

The current sectoral focus of federal infrastructure loan programs, however, should not be seen to reflect any fundamental uniqueness in their central activities. The programs have many common non-sectoral elements as well.  Most fundamentally, this is because they all make big, long-term secured loans to large creditworthy entities capable of building and operating large-scale projects. Four major programs also share the same basic statutory framework, defined originally for TIFIA and then more or less replicated for WIFIA and the Army Corps’ CWIFP, and more recently for CIFIA.  Their loans have similar financial features (fixed interest rate determination, prepayment flexibility, non-subordination, etc.) and are subject to the same FCRA budgetary treatment, OMB oversight and federal crosscutter requirements.  They are all funded by the US Treasury.

Program borrowers face common challenges in financing large-scale infrastructure projects, including:

• Long planning and construction periods during which interest rate risk on the project’s permanent long-term permanent financing must be hedged or otherwise managed.

• The need to source a significant amount of long-term debt and other capital, since program infrastructure loan amounts are statutorily limited to a maximum percentage of project cost — usually 49% for TIFIA, WIFIA and CWIFP, and 80% for CIFIA.  The balance of the project’s capitalization will need to come from other sources, primarily the tax-exempt bond market since most program borrowers are public-sector agencies or qualify for PABs. Other sources include SRFs, P3 equity investors and specialty lenders. The project’s lenders and investors will need to work with both the mandatory requirements and optional features of the federal program loan.

• Large-scale, long-lived projects are exposed to intrinsic, difficult-to-manage risks like climate change.  A project’s financing should be designed to help mitigate their impact whenever possible.

Large-scale projects have common stakeholders whose organizations range across infrastructure sectors – labor organizations, engineering & construction firms, state & local governments, legal and financial specialists, etc. In effect, these are common stakeholders of loan programs, too.

Common Issues and Opportunities

The extent of common elements across infrastructure loan programs, borrowers and stakeholders means that there are also many common issues and opportunities related to their loans.  Here’s an illustrative list, based on my own experience (links to relevant posts or articles), of various topics that appear to be relevant, to a greater or lesser extent, for large-scale federal infrastructure program loans: 

The chart below shows what I think to be each topic’s relative importance (reflected by color intensity) across four major loan programs:

A Unified Approach?

Some form of unified approach to solving issues or developing opportunities for infrastructure loan programs’ cross-sectoral capabilities would seem to be useful, for several reasons. The most obvious case is where one program has established a precedent or implemented a statutory refinement that would apply to the others. A less apparent but perhaps more significant reason arises from the fact that many issues and opportunities in program loans involve technical aspects of finance and debt markets. Sectoral agencies don’t have an in-depth level of relevant expertise here — it’s well outside their main mission. Yet all the programs make large loans to sophisticated borrowers who are simultaneously sourcing capital from major markets. A unified approach to the common financial aspects of program loans would benefit from classic scale economies and produce better results.

The most important reason for a unified approach to federal infrastructure loan programs, however, is more far-reaching than improving current implementation with precedents and expertise. The approach can provide a focal point for cross-sectoral stakeholders to view infrastructure loan programs outside of sectoral siloes, resulting in an additional — and more broadly-based — level of advocacy for expanding program capabilities.

This is the second post in the FCRA Non-Federal Series.

The 1967 Report of the President’s Commission on Budget Concepts is surprisingly interesting.  It’s certainly limited in some ways by its historical context.  But the writing is clear, the concepts are carefully derived from fundamental principles, and the recommendations are direct and far-reaching.  The Report appears to have successfully defined the way the federal budget works today, especially with respect to the FCRA concepts I’m familiar with.  As I noted in the prior post in this series, the Congressional directive’s instruction to follow the Report in conjunction with interpreting current FCRA law was a good call.

Two chapters in the Report are relevant to the FCRA non-federal issue.  Chapter 3, Coverage of the Budget, describes what should be included in the overall federal budget.  Chapter 5, Federal Credit Programs, outlines a separate section within the budget for federal direct loans and guarantees.

Coverage of the Budget

Chapter 3 begins with the statement of a principle that appears to guide the Commission’s approach to budget coverage but also may be important to specific FCRA non-federal solutions: Any federal activity that allocates the government’s resources must be subject to the internal ‘discipline’ of the budget if it’s not subject to external discipline.  The paragraph goes on to say that however clear that might be in theory, in practice it’s not easy to draw the line:

The next paragraph describes two examples (one of which is definitely historical) to illustrate the extremes of what should and should not be included in the federal budget. In a list of more ambiguous situations, the Report specifically mentions enterprises jointly owned by federal and private-sector participants (the context implies that the latter would also include non-federal public-sector participants):

After describing some other areas of ambiguity, the Report sums up an approach for ‘borderline’ situations – “when in doubt, include it”:

In the main section of the chapter, the Commission admits that it’s easy to fall into a rabbit hole when defining the budget’s theoretical boundary lines and therefore the Report’s practical scope is limited to a ‘few key agencies and programs’.  The rest of the chapter discusses those, none of which seem especially relevant to the FCRA non-federal issue:

General Observations About Chapter 3:

• If there is some ambiguity about federal budgeting for the activities of federally involved projects as a type of ‘enterprise owned jointly’ by federal and non-federal entities, then the Report’s ‘borderline inclusion’ recommendation would presumably apply directly.  But that’s not the case for the FCRA non-federal budgeting issue, which relates only to loan classification within the budget.  The two areas are fundamentally different [1].

• In contrast, there is no ambiguity about federal budget inclusion for WIFIA and all its activities, including its loans.  None whatsoever – WIFIA is a wholly federal program within a major federal agency and everything about it must be (and is) included in the federal budget.  The FCRA non-federal issue is solely about whether a program loan belongs either in the FCRA subsection or the general cash-based budget — one way or another, it will be included [2].

• Although Chapter 3’s inclusion concepts don’t appear to apply directly, the ‘discipline’ principle stated in the chapter’s first sentence can provide some guidance about the way the Commission might have looked at the FCRA non-federal issue.  If a program loan is not subject to external discipline, then some special consideration is presumably required to ensure that it will be subject to the internal discipline of the budget, as discussed further below.

Federal Credit Programs

Chapter 5 covers federal credit programs as a specialized area.  I was struck how closely the principles outlined in this chapter seem to have determined in some detail what FCRA law looked like 23 years later.

The Report admits that federal credit is a difficult area within their conceptual framework and recommends that program loans be put in a separate section within the unified budget.  The main objective of the separate section is to provide better information about the true cost of program loans in connection with their economic impact.  The Commission expected that ‘most’ federal program loans would belong in the separate section:

The chapter’s next bullet provides an accurate summary of how FCRA methodology would eventually work. The grant-like element in federal program loans belongs in the cash-based budget, while (implicitly here) the loan’s reversing cash flows do not:

Some types of federal program loans, however, should stay in the cash-based budget, primarily because they are effectively grants for which no repayment is expected or definable [3]:

The main part of Chapter 5 starts with the observation that a separate budget section for program loans is important because federal loan programs were steadily expanding ($30 billion in those days was apparently considered ‘real money’). Note that loan programs for large-scale public infrastructure aren’t mentioned in their list. I don’t think there were any at the time. Even today, such programs are a very small segment of total federal credit. However, their future expansion, including in sectors where federal involvement is frequent, is the same rationale for solving the FCRA non-federal issue:

Later in the main section, the Report provides an important insight into why the Commission considered federal program loans to be fundamentally different than other federal economic activities.  Unlike other cash transfers from the federal government, the borrower of a program loan has assumed an obligation to repay it:

General Observations About Chapter 5:

• The budget’s separate loan section (which eventually became FCRA) can be derived from two primary principles in the Report: (1) for overall inclusion, when an activity is not subject to external (that is, non-federal) discipline, and (2) for inclusion in the separate loan section, when there is a substantive and definable repayment obligation.

• Assuming that a program loan’s repayment obligation is subject to external discipline from a non-federal source, most of the loan’s cash flows should be excluded from a cash-based budget since they are not internally managed federal expenditures or revenues. The grant-like portions of the loan (the credit loss and funding subsidy amounts), however, are provided by the federal program and not subject to external discipline. Hence, that portion must be estimated and recorded as a net expenditure, as the Report described and as FCRA currently requires.

• If a loan’s repayment obligation, regardless of transactional form, is not substantive or definable, then it’s not subject to external discipline. In effect, all the loan’s cash flows then become internally managed federal expenditures and revenues, and the loan should be included only in the cash-based budget, not the separate loan section. This is consistent with the Report’s examples of loans that should be excluded from the separate section.

How Might the Commission Have Considered the FCRA Non-Federal Issue?

The above interpretation of how two of the Report’s fundamental principles define the separate budgetary treatments for loans is useful because it highlights one question that the Commission did not address but seems to be at the core of the FCRA non-federal issue: What if a program loan has substantive and definable repayment obligation that is not subject to external discipline — in other words, if the loan’s repayment obligation is from a federal source?

If this question was posed to the Commission in 1967, I think their response would have started with a description of two extremes, as they did at the beginning of Chapter 3. If the federal participant in a project is the direct recourse obligor or guarantor of a program loan, then obviously the loan completely lacks external discipline and should not be included in the separate section. At the opposite extreme, if the federal participant’s role in a project has no financial impact and the project loan’s sources of repayment are entirely outside the federal sphere (e.g., locally generated user-fees or taxes), then the loan is clearly fully subject to external discipline and ought to be included in the separate section.

The Commission would probably have continued by admitting that situations between the two extremes are ‘difficult’. And they likely would have supported the Congressional directive’s approach of developing budget classification criteria to be used by loan programs in these situations, at least until a FCRA statutory fix was available. But based on the Report’s emphasis on the budget’s unifying principles, and the successful application of their specific recommendations as the foundation of FCRA, I think they would have expected that any new criteria or statutory language for solving the FCRA non-federal issue would be narrowly focused and consistent with their concepts. It seems to me that this can be effectively accomplished by utilizing the two principles — external discipline and repayment obligation — that appear to determine the exclusion of certain loans from the separate loan section proposed by the Report.

_____________________________________________________________________________________________

Notes

[1] A non-capital, cash-based accounting system doesn’t really include a FASB-like consolidation concept, and I assume that federal participants simply record their investments in a project as an ‘expenditure’ and any return of cash as a ‘revenue’.  Is it more complicated than that? If there is a budgeting issue here that doesn’t pertain solely to the project loan, doesn’t it require a solution regardless of whether the project has applied to a federal loan program? If the issue pertains solely to the budgetary treatment of the project’s loan by a loan program, then (as the next bullet above describes), it’s already within the budget.

[2] Interest rate re-estimates have an interesting treatment under FCRA law. They’re not included in the program’s discretionary budget but automatically receive budget authority for mandatory appropriations — in a separate account. This treatment might have its own issues, but in fact the cost of re-estimates is included, albeit a bit obscurely, in the federal budget, reflecting the comprehensive scope of FCRA.

[3] It’s worth noting that if such grant-like loans were subject to FCRA treatment, they’d almost certainly require a 100% subsidy rate. In effect, FCRA would automatically put them back in the cash-based budget.