The problem with the WIFIA ‘criteria’ published in June 2020 is that they are not criteria, they’re questions that imply criteria being used by OMB off stage. Not a very transparent approach. This is discussed in detail in a prior post.

Well, why not fill in the blanks with explicit criteria developed from the sources specified by the Congressional directive — FCRA law and the 1967 Report? I came up with six based on the language & principles stated in those sources, and consistent with the CBO Report as well.

Using these six criteria as a guide, answers to WIFIA’s questions about federally involved projects can be much more straightforward — and informative. Which is the point, right?

• (A) There must be a substantive obligation to repay the program loan.

• (B) The substantive obligation to repay the loan must be from a non-federal entity using non-federal resources.

• (C) The substantive obligation to repay the loan must be the result of an independent decision by a non-federal entity and not directly or indirectly compelled by the federal government.

• (D) The non-federal entity obligated to repay the loan should also be the primary beneficiary of the capital improvements financed by the loan’s proceeds.  

• (E) FCRA treatment should be evaluated independently of the loan program’s other federal policy objectives, eligibility requirements or selection criteria. 

• (F) FCRA treatment for a specific loan should be consistent with the federal participant’s budgeting and tax treatment for the program loan and all equivalent non-federal debt of the project.

This is the fifth post in the FCRA Non-Federal Series

In June 2020, EPA, OMB and Treasury published WIFIA Criteria Pursuant to the Further Consolidated Appropriations Act, 2020 in the Federal Register.  At publication, these FCRA non-federal criteria became applicable to loans being considered by the WIFIA program, including the sections sharing WIFIA’s statutory framework, SWIFIA and CWIFP.

General Observations

Before getting into the specifics, here are a few general observations about the WIFIA criteria:

• Despite the Congressional directive language and the Federal Register title, these aren’t ‘criteria’, but 18 questions that imply criteria being applied by OMB offstage [1].  What was the reason for this somewhat non-transparent approach?  If I had to guess, I’d say it reflects some uncertainty at OMB and EPA about how the criteria are supposed to work, other than giving WIFIA the ability to reject a project application with practically any degree of federal involvement.  The GAO Report’s description of OMB’s thinking is consistent with this interpretation.  Perhaps the huge pressure to get something published by a deadline was also a factor?

• The only budget principle mentioned in the Federal Register publication is the ‘when in doubt, include it’ guidance from the 1967 Report.  As discussed in several prior posts, there’s no question that a loan from a federal loan program will be included in the federal budget – the loan’s FCRA treatment is the specific issue that the criteria should elucidate.  Yet many of the questions seem to be directed at determining whether the federal participant should be including more (all?) of the project’s non-federal cash flows in its federal budget.  If that’s an issue, it will exist regardless of a program loan and likely involve much larger cash flows (e.g., all the project’s non-recourse debt and equity capitalization, all the construction cash flows, all the O&M, etc.).  That seems to be a more fundamental budgeting problem – shouldn’t OMB address it directly with the federal participant?

• It’s hard to see how OMB’s questions fundamentally differ from the many questions and required information involved in any federal infrastructure loan application.  Much of the basic data sought will be the same for both.  And the program can always ask for more.  Information for a loan’s FCRA treatment is effectively statutorily required in a WIFIA application through the program’s mandated budgeting for credit subsidy allocation (§ 3901 (13)) and selection criteria (§ 3907 (b)(E)).  If more detail is required in the case of a federally involved project, this could have been accomplished with additional questions in the application materials and specific guidance in WIFIA’s program guide, just as it’s done for any other statutory standard.  Of course, the Congressional directive required publication of OMB’s criteria in the Federal Register, so the usual approach wasn’t possible in this case.  Still, there’s some dissonance here.  I can see why a statement of criteria and associated principles would belong in the Federal Register as a sort of ‘official pronouncement’.  But questions to elucidate the application of criteria in specific cases would seem to belong at the loan application process level.

Six Criteria for FCRA Treatment of a Program Loan to a Federally Involved Project

The FCRA non-federal issue is important and there’s nothing wrong with asking a federally involved project applicant to provide more information.  But in this series, we’re interested in how the FCRA non-federal issue can be efficiently resolved.  For that purpose, OMB’s questions don’t shed much light on how the information will be used to determine FCRA treatment in specific cases because the evaluation criteria are only implied, not stated.  Trying to guess the criteria OMB is using from the questions being asked is perhaps an interesting intellectual exercise, but I don’t see much practical value in it.

Instead, I’m going to approach OMB’s questions by providing the missing criteria myself, based on the reviews of the 1967 Report, the CBO Report, and the GAO Report in prior posts.  I think the budgeting principles that surfaced in those reviews are clear, consistent with each other, and precisely applicable to the FCRA non-federal issue.  The necessary criteria – I think there are six — seem to flow naturally from them.  If I’m wrong, at least this approach provides a more defined framework for future development than OMB’s questions alone.

Here they are, with reference to the principles behind them:

• (A) There must be a substantive obligation to repay the loan.  According to the 1967 Report, the distinguishing characteristic of a loan and the basis for its separate (i.e., eventually FCRA) treatment in the federal budget is the obligation to repay.  This obligation must be defined and substantial – if it’s not, the loan is effectively a grant and not eligible for FCRA treatment.

• (B) The substantive obligation to repay the program loan must be from a non-federal entity using non-federal resources.  In the 1967 Report, the core principle of budget inclusion for an activity is the fact that it is not ‘subject to economic disciplines of the marketplace’ and hence must be subject to the internal discipline of the federal budget.  For the FCRA non-federal issue, the term ‘marketplace’ is logically expanded to include non-federal economic actors like local or regional communities and their agencies.  The external discipline principle can be extrapolated to FCRA:  If a program loan’s repayment cash flows (net of subsidy) are not subject to the discipline of non-federal entities deciding to use their non-federal resources to pay them, then those cash flows are federal and must be subject to the internal discipline of the cash-based budget.  I think this is precisely what FCRA law meant in section 504 (1) in defining an eligible loan as “a disbursement of funds by the Government to a non-Federal borrower under a contract that requires the repayment of such funds”.  The substantive non-federal obligation to repay is the key factor, not the nominal characterization of the borrower.  The people drafting FCRA law would have had the 1967 Report right in front of them.  Perhaps they thought that the principles underlying their use of the term ‘non-Federal borrower’ in the definition were so clear that further explanation was not necessary.  If so, that would put FCRA’s ‘silence’ about program loans to federally involved projects in a different light.

• (C) The substantive obligation to repay must be an independent decision by the non-federal entity and not compelled in some way by the federal government.  If the use of federal sovereign power is subverting the external discipline that non-federal entities would otherwise bring to bear on an activity, then the activity becomes federal and should be subject to the internal discipline of the federal budget.  If the activity in question is a program loan, it shouldn’t be eligible for FCRA treatment.  This is consistent with the CBO Report’s scoring criteria about the exercise of federal sovereign power.  But for FCRA, it should be interpreted precisely in connection with the non-federal entity’s decision to repay the program loan.  If a federal participant uses sovereign power to help create or even enable a very useful project, that does not necessarily determine that a non-federal entity will agree to repay financing associated with it.  

• (D) The non-federal entity obligated to repay the loan is also the primary beneficiary of the capital improvements financed by the loan’s proceeds.  Unless there is an explicit statement of philanthropic or patriotic intent, it probably can be safely assumed that a rational non-federal entity will only agree to repay the program loan for its own benefit.  But I think for federal infrastructure loan programs, which primarily finance specific capital expenditures, confirmation of this can (and should) be made more explicit, especially in the case of large multi-part projects.  There should be a more-or-less direct connection between the use of loan proceeds for construction of those parts of the project benefiting a non-federal entity and the same entity’s obligation to repay the loan.   Notwithstanding any eminent domain power, scale economies, or efficiencies that a federal participant might bring to the project, the benefit-cost metrics of a large basic infrastructure project are necessarily limited by physical reality and the cost of construction.  Connecting the use of program loan proceeds and the non-federal entity’s benefits will help ensure that there’s some rough parity between the two, and that the non-federal entity’s decision to repay is substantive in terms of the allocation of real resources. 

• (E) FCRA treatment should be evaluated independently of the loan program’s other federal policy objectives, eligibility requirements or selection criteria.  The 1967 Report makes it clear that FCRA is a unique section of the federal budget and intended for a very specific purpose.  Its only policy objective is to ensure that a loan’s non-federal repayment cash flows don’t distort the cash-based federal budget.    FCRA decisions shouldn’t influence, or be influenced by, other loan program factors.  FCRA treatment shouldn’t be used as an indirect way to insert new selection criteria about a project’s overall level of federal support, change the program’s statutory eligibility requirements, or reinforce unrelated federal budgeting objectives like legislative scoring.  The 1967 Budget Commission very consciously put this separate section for federal credit in a silo – it’s meant to stay there.

• (F) FCRA treatment for a specific loan must be consistent with the federal participant’s budgeting treatment for all equivalent project non-federal debt.  Notwithstanding a FCRA decision’s siloed position within the program for a specific loan, it should be consistent with the federal budgeting by the federal participant for other project debt with equivalent terms (e.g., purpose, security, seniority, source of repayment, recourse to the project, etc.).   If the federal participant is properly including such equivalent debt in its cash-based budget, that indicates that the debt’s repayment is not subject to external discipline.  The program loan should not be differentiated in this respect just because it comes from a federal loan program and FCRA treatment should be disallowed.  Likewise, if the federal participant is properly excluding the project’s equivalent debt in its cash-based budget, a program loan should receive FCRA treatment.  To put it more generally, to avoid double counting and budget gaming, all the federal agencies involved in a project should have the same position on whether the repayment of the project’s debt is subject to external discipline and therefore may be excluded from the federal cash budget.  FCRA treatment for a program loan to the project (if one is applied for) is just an extension of this position for a federal lender.  If the project is issuing tax-exempt bonds (which can’t have a direct or indirect federal guarantee, per IRC 149 (b)), the IRS’s position should also be consistent with federal budgeting because the same basic principles apply in this case.

Interpreting OMB’s Questions with Six Stated Criteria

With these six FCRA criteria, we can interpret OMB’s questions in a focused way that will guide a potential WIFIA applicant to provide the information necessary for determining correct FCRA treatment for the proposed loan — not more, not less, and consistent with the Congressional directive. I sketch out the guidance as if it were to be in a program user guide, with reference to the above criteria.

IV. Initial Federal Asset Screening Questions

A. Is the project, in whole or in part, a project currently authorized by Congress for the Army Corps of Engineers or Bureau of Reclamation to construct [2]?

• If this applies, the applicant should provide information that demonstrates for the relevant part of the project that such Congressional authorization did not (1) significantly alter the basic conditions that the applicant would expect to find for its typical infrastructure projects (e.g., required long-term agreements, normal benefit-cost metrics, elements of control, etc.), and (2) require or compel the use of non-federal for the repayment of the proposed program loan. (Criteria A, B, and C)

B. Is the project, in whole or in part, a local cost share requirement for an Army Corps of Engineers or Bureau of Reclamation project [3]?

• The applicant should be aware that although WIFIA loans to support federal cost-sharing are authorized, and the reduction of federal resources for a project is a positive factor in WIFIA’s statutory selection process, FCRA treatment of the proposed loan is a completely separate decision based on the facts relevant to the specific FCRA criteria stated above. (Criterion E)

V. Federal Asset Screening Criteria

Structure of the Project

1. To what degree does the Federal Government comprise the WIFIA project’s user base?

• The applicant should provide information demonstrating that the applicant (including related parties) is the primary beneficiary of the completed construction and operations of that part of project proposed to be financed by the program loan. (Criterion D)

2. Does the project involve the use of the Federal Government’s sovereign power (excluding, e.g., National Environmental Policy Act (NEPA) review)?

• The applicant should provide information demonstrating that its obligation to repay the program loan and other related project financing will not be directly or indirectly compelled by the specific exercise of federal sovereign power by the project’s federal participant or any other federal agency. Note that compliance with existing, non-specific federal law (e.g., environmental regulations, Clean Water Act, NEPA, etc.) does not constitute compulsion for this purpose. Evidence includes confirmation that the decision will follow the participant’s standard processes and will be subject to voter referenda, public debate and comment, local council voting, positive benefit-cost review, etc. as normally applicable. (Criterion C)

3. Does the WIFIA project require the construction or acquisition of an asset for the special purpose of or use by the Federal Government?

• The applicant should provide information demonstrating that if the relevant part of the project has scale, design or construction features that appear to primarily serve a federal purpose, (1) the applicant’s primary project benefits are not significantly impacted, and (2) the features will not be a material use of program loan proceeds. (Criterion D)

4. To what degree does the Federal Government direct the contracting process for the WIFIA project?

• The applicant should provide information demonstrating that federal involvement in the contracting process and related agreements for the relevant part of the project did not materially influence its decision to seek and be obligated to repay a program loan for the project. Evidence would include a benefit-cost analysis or a standard project finance engineering review showing that the contracting was within normal industry parameters. (Criterion C)

5. Is there a specific authority provided to the WIFIA project by an Act of Congress without which the WIFIA project could not proceed?

• The applicant should provide information demonstrating that any specific authority granted to the relevant part of the project by Congress (1) was consistent with the implementation of non-specific federal policy objectives (e.g., for a region or infrastructure sector) and (2) did not compel the applicant to seek and be obligated to repay a program loan for the project. (Criterion C)

6. What is the Federal Government’s role in the governance of the project?  In other words, what is the role of the Federal Government in selecting management or overseeing the project (including, but not limited to, approval of contract scope and step-in rights, or as a member of a board of directors), both during construction as well as in terms of operations and ongoing maintenance?

• This question is similar to No. 4 above but applies to the post-completion, operational phase of the project. The applicant should provide information demonstrating that federal involvement in the future operations and management for the relevant part of the project did not materially influence the nature of the long-term agreements between the participant and the project with respect to control and redress in matters concerning end-user benefits, relative to the participant’s normal agreements for large-scale infrastructure projects. (Criteria C and D)

7. Is this project part of a larger Federally authorized project (not limited to but consistent with the initial screening criteria) and if so, does the project under consideration for a loan or loan guarantee constitute a useful segment—either a planning segment or a useful asset—as defined in the Capital Programming Guide (supplement to OMB Circular A–11)?

• If both factors are applicable, the applicant should provide information demonstrating that the relevant part of the project (1) will provide direct benefits to the applicant that are not significantly impacted by the project’s other uses or utility, and (2) that program loan proceeds are primarily used in connection with those direct benefits. (Criterion D)

Financing of the Project

8. Does the Federal Government provide resources for the WIFIA Federal loan repayment?

• The applicant should confirm that the program loan repayment will (1) come from entirely non-federal sources and (2) not directly or indirectly utilize federal resources specifically dedicated to the purpose of program loan repayment or a guarantee thereof. (Criteria A and B)

9. Will the WIFIA project meet the non-subordination requirement provided in 33 U.S.C. 3908(b)(6)?

• The applicant should provide information confirming that the federal participant and related federal agencies will not have any specific or extraordinary rights for seniority of claims over the secured debt obligations (including the program loan) of the relevant part of the project. Federal seniority of claims that arise in the normal course of the enforcement of non-specific laws (e.g., federal tax liens) and not in connection with any project-specific legislation for authorization or funding can be excluded for this purpose. The applicant should be aware that the program loan’s seniority with respect to other secured debt of the relevant part of the project is subject to WIFIA’s non-subordination requirement without regard to any federal involvement in the project. (Criteria C, E and F)

10. Does the WIFIA project depend on the Federal Government making other in-kind contributions (land, real estate, right-of-way, etc.)?

• The applicant should confirm that no federal in-kind contributions to the project will be utilized in, or available for, program loan repayment via sale of such contributions or other conversion of their value to money, including indebtedness. (Criteria A, B and D)

11. Is non-federal financing available for the project?

• The applicant should be aware that since a WIFIA program loan amount is limited to 49% of the cost of the relevant part of the project benefitting the applicant, the absence of other financing for the same purpose with essentially equivalent terms (e.g., security, repayment source, etc.) for a significant part of the 51% balance of cost will be considered unusual. If this is the case, the applicant should provide explanatory information. If the applicant is expecting to access other non-federal financing, details with respect to its relationship to the federal participant or other federal agencies should be provided, if known. For example, if the project is planning to issue tax-exempt bonds, the participant should provide information regarding the bonds’ compliance with IRC 149 (b), which prohibits a federal guarantee of tax-exempt debt. (Criteria B and F)

12. If the project is required to obtain an investment-grade rating opinion letter, per 33 U.S.C. 3901(4) and 3908(a)(3), to what extent does the rating opinion letter consider Federal support as a credit enhancement?

• Credit rating agencies primarily review a borrower’s ability to repay, and generally assume that a project will perform as designed. If the project relies on variable revenues (e.g., sale of output, non-monopoly user fees) the agencies will assess revenue risk. Most public infrastructure will rely on non-variable, contractual revenues (e.g., taxes, covenanted user rates) where the primary risk is the ability to pay. If the relevant part of the project is subject to revenue risk, the applicant should provide information demonstrating that the federal government is not providing any specific support for those revenues on a long-term basis (i.e., a fixed purchase agreement at a minimum price and volume). If project revenues are primarily sourced from taxes or user rates, the applicant should provide information confirming that the federal government is not specifically supporting or supplementing those revenues in connection with financing for the project, including the program loan. The analysis underlying rating agency opinion letters and/or formal ratings should be consistent with this information. (Criteria A and B)

Project Liabilities

13. To what degree will the Federal Government bear funding liabilities associated with the WIFIA project not otherwise appropriated by Congress or captured in the loan subsidy?

• The applicant should provide information demonstrating that any federal funding for the project that arises from the federal participant’s operational and managerial obligations, but not contemplated in the participant’s original authorization or appropriations, will not be used, directly or indirectly, to provide resources to the applicant for the repayment of the program loan. (Criteria A and B)

14. Is the risk to the Federal Government low relative to the private sector for the financing of the WIFIA project?

• The risk profile of the program loan (as evidenced by expected credit rating, proposed loan coverage ratios, etc.) should not be significantly different than other equivalent project debt from non-federal sources. If it is significantly higher or lower, this may indicate an insufficient degree of arms-length interaction between the federal participant and the federal lender. The applicant should provide (1) information demonstrating the similarity between the risk profiles of the program loan and equivalent project debt from non-federal sources, or (2) explanatory information about any significant dissimilarities. (Criteria A, B and C)

15. To what degree does the Federal Government own or is the Federal Government contractually obligated to complete, maintain, or repair damage to the WIFIA project?

• The applicant should provide information confirming that any such federal ownership or obligations will not result in federal resources being used for program loan repayment through the direct or indirect monetization of ownership equity (e.g., by sale) or the fulfillment of contractual obligations (e.g., by the payment of liquidated damages). (Criterion B)

16. Is the Federal Government liable for unforeseen costs (e.g., environmental impacts, damage from natural disasters, or cost overruns) either before, during or after completion of the WIFIA project?

• The applicant should provide information confirming that any such federal liability for unforeseen project costs (1) will not result in federal resources being used for program loan repayment and (2) does not significantly change the benefit-cost relationship between the expected benefits of, and payment for, the relevant part of the project, relative to industry norms for similar infrastructure project financings. (Criteria B and D)

_____________________________________________________________________________________________

Notes

[1] The Background section of the publication does say: “The 1967 Report notes the inherent difficulty in making a Federal or non-Federal determination in many cases and suggests a series of questions, which guide the criteria below…” Yes, that report does suggest asking questions — but it also provides an extensive discussion and clear statements of principles (especially in the Federal Credit Program chapter) to guide the right questions and interpret answers. General questions about an issue are of course necessary to develop criteria for decisions in specific cases. But specific questions directed towards specific cases don’t ‘guide the criteria’ that apply — the criteria should be developed and stated beforehand.

[2] The footnote attached to this question states flatly: ‘A project authorized by an Act of Congress to be built by the Army Corps of Engineers or Bureau of Reclamation is ineligible for WIFIA financing.’ I assume (perhaps incorrectly) that the ineligibility is not part of the WIFIA statute or another, more broadly applicable law because otherwise there would be a reference to it. More importantly, if it existed, such statutory ineligibility would simply render any FCRA budgeting issue moot — the application would be rejected on the threshold review. I suspect that the ineligibility here is based on OMB’s presumption that the application will fail the FCRA test because a project authorized by an Act of Congress must be a ‘federal project’ and therefore all of its debt must be ‘federal debt’ which is not FCRA-eligible because there’s a ‘federal borrower’. This presumption is not valid, as discussed above and in prior posts. Why can’t a specific case be judged on its own merits, without such presumptions? What does the Congressional act precisely specify? How is ‘project’ being defined here? What are the actual terms and sources of program loan repayment? And so on.

[3] The footnote attached to this question ends: Project applicants are encouraged to review all applicable statutory requirements before seeking WIFIA financing. Good idea — please point out WIFIA’s explicit statutory ineligibility for Congressionally authorized projects.

This is the fourth post in the FCRA Non-Federal Series

In July 2022, the GAO published a report on the FCRA non-federal issue, Transparency Needed for Evaluation of Potential Federal Involvement in Projects Seeking Loans.  Unlike the two prior reports discussed in this series, the GAO Report does not analyze budget principles but primarily describes the results of an investigation.  It was requested by the House’s Budget Committee in 2020 and the investigation was conducted from June 2021 to July 2022. 

The core of the GAO Report is a description of OMB’s concepts underlying the FCRA non-federal criteria for WIFIA published in 2020, per the Congressional directive.  Interestingly, Treasury involvement is not mentioned, even though in the directive it was named along with EPA and OMB as one of primary parties responsible for criteria development.  That doesn’t mean Treasury wasn’t involved in some way, though the relative lack of focus and precision about debt mechanics in the published criteria does suggest their input was limited.

The Report also briefly describes GAO investigation into how the FCRA non-federal issue is considered at five other federal infrastructure loan programs, though outside of WIFIA, it’s apparently non-existent or very infrequent.  That scope, however, is the context for GAO’s single recommendation, that criteria for the FCRA non-federal issue be developed and applied uniformly to all federal loan programs.  Since the GAO Report does not question or evaluate OMB’s thought processes or the resultant WIFIA criteria, the implication is that WIFIA’s currently published criteria should be the model for a universal application.

Factors Considered by OMB in the Development of the Criteria

There are five important paragraphs in the GAO Report on OMB’s description of its thinking in developing the FCRA non-federal criteria. I’ll quote them and make some general observations about each in sequence.

According to OMB staff, their classification of a borrower as federal or non-federal depends on whether a loan is sought for a federal or non-federal activity, rather than solely on whether the direct recipient of the loan is a federal or non-federal entity.  In other words, the fact that a loan applicant is not a federal entity does not automatically mean that the proposed project will be considered non-federal for the purposes of FCRA.

• 1967 Report makes it clear that the distinguishing characteristic of a federal program loan is the obligation to repay, not what it is ‘sought for’.  Repayment is the ‘activity’ that matters here.  I think it can be assumed (though it should also be confirmed) that if substantial non-federal entities are willing to commit to repay a large amount of money over decades, they are doing it out of non-federal self-interest – the project, or part thereof, that they’re paying for is (perhaps sadly) not a patriotic gift to the nation.  Their presumably rational decision to allocate their own non-federal resources for loan repayment is the external, non-federal discipline that the 1967 Report is looking for with respect to separate budget treatment.

• FCRA does not classify ‘projects’ – only federal program loans.  With that more precise language replacement, the sentence can be flipped around with equal truth: “In other words, the fact that a loan applicant is a federal entity (e.g., for CBO scoring purposes) does not automatically mean that the proposed program loan will be considered federal for the purposes of FCRA.” Again, it all depends on who’s repaying the loan.

Structure of the Project.  OMB staff stated that if the federal government is the primary recipient of a completed project or if it receives the majority of a benefit from a project completed by a non-federal borrower, OMB may determine the loan would not qualify for FCRA’s special budgetary treatment.

• On surface this makes complete sense – if the project’s primary beneficiary is federal, then it’s a federal project.  This echoes the CBO Report’s criteria and its examples describing facilities dedicated for use by federal agencies.

• But there’s a missing piece that ought to be made explicit.  If the project’s primary beneficiary is federal, then what is the ultimate source of the proposed program loan’s repayment?  Presumably, it will also be a federal entity, directly or (more likely) indirectly through contractual or lease payments, as in CBO’s examples.  If that’s the case, then a proposed program loan will fail the external discipline test and should not receive FCRA treatment.  But much more fundamentally, infrastructure loan programs are intended to support financing for large-scale public projects that benefit specific local or regional communities.  It’s hard to imagine that a large-scale infrastructure project that primarily benefitted only a federal agency would even meet threshold eligibility at an infrastructure loan program – much less that local or regional communities would pay for it.

• Still, the ‘Structure of the Project’ is important, though the key point about this is not mentioned.  In the real world, a large-scale project with federal participation but with substantial non-federal participants, owners, and stakeholders, and financed at least in part by non-recourse debt, will have a complex structure that can’t be easily classified as wholly one thing or another [1].  It depends on what ‘activity’ you’re concerned about.  Regarding FCRA treatment, that activity is specifically the repayment of a program loan and whether it is subject to external, non-federal discipline.  For the FCRA non-federal issue, consideration of the project’s structure needs to be focused on answering this precise question.

Financing of the Project.  OMB also assesses the resources available to a borrower for repayment and how dependent a project may be on federal support.  OMB staff stated that as part of credit program implementation, agencies are asked why a given objective cannot be achieved without federal support, what existing private sources of credit are available to support the project’s objective, and what other federal credit or noncredit programs exist to support the objective.

• The first sentence starts in a promising way with a mention of loan repayment (finally!) but then slides into an apparent implication that the overall degree of federal support for a project will determine the ‘resources available’ for that repayment.  In fact, they aren’t necessarily related.  In a real-world infrastructure project financing, the ‘borrower’ is likely to be a special purpose company e.g., ProjectCo LLC.  But the substantive source of repayment for the program loan and other non-recourse debt will be the non-federal resources of the beneficiaries of the project (or relevant parts thereof), likely local or regional communities.  The federal participant will most likely be involved at or below the ProjectCo development, management, and ownership level.  The relationship between the repaying beneficiaries and the federal participant needs to be examined, but assuming they’re honest and rational actors, it’s unlikely that the resources for loan repayment will be considered federal.  The degree of overall federal support for the project, outside repayment resources, might be a policy matter for the loan program to consider, but it isn’t a FCRA issue.

• The questions about loan program objectives are explicitly asked in OMB Circular A-129 and are applicable to all federal credit programs and loans.  But what they address is whether federal credit is necessary, not ‘federal support’ as a whole.  Certainly, the questions have nothing to do with FCRA.  If a loan program is making loans to highly rated, indisputably non-federal public agencies with excellent access to the tax-exempt bond market for financing essential infrastructure projects that they’ll do anyway, the loans will certainly receive FCRA treatment, but one might wonder how they’re consistent with Circular A-129 [2].  On the other hand, a program loan to a one-off, very large-scale, and barely investment-grade infrastructure project financing being developed by a consortium of regional authorities and including a federal participant will have great answers to the Circular’s questions but encounter the FCRA non-federal issue.  What is OMB’s point here?

• Perhaps OMB is thinking that the FCRA non-federal issue might be correlated to a project receiving a lot of federal support, in effect ‘reminding’ the loan program that there might be a Circular A-129 issue (e.g., the loan is not necessary) with that, even though the two are separate.  While that reminder might not be a bad idea in terms of overall policy objectives, I’m not sure it belongs in the specific FCRA criteria that Congress asked for or how it would work with other loan program eligibility or selection criteria.  For example, WIFIA explicitly references federal support in its selection criteria: §3907 (b)(2)(K) The extent to which assistance under this chapter reduces the contribution of Federal assistance to the project.  A federally involved project that would otherwise need to rely completely on federal resources but has decided to raise non-recourse debt, including a program loan, to be repaid from non-federal sources is clearly reducing ‘Federal assistance’ to the project.  In this case, the project’s FCRA non-federal issue is indicating an opportunity to reduce federal support, a statutory program objective.

Project Liabilities.  OMB staff stated that, in some cases, former or current federal involvement in projects may confer a significant level of control and future responsibility, or liability, to the federal government.  OMB staff stated that in these cases, future activity completed by a non-federal borrower on an asset would still be considered a federal project.

• OMB appears to be referring to the project’s non-debt liabilities – that is, recourse performance obligations of the project’s owners, developers, and the federal participant.  These can be substantial in a classic project financing, though they are usually off-set by contractual arrangements with other creditworthy parties – construction bonding, insurance, third-party O&M management, etc.  It’s possible (maybe likely in some cases) that the federal participant can uniquely absorb some project risks that the private-sector or sub-national governments can’t.  But presumably this is part of the federal participant’s policy implementation towards objectives that can be achieved by the project, especially if the project won’t happen otherwise.  If there is a cost, it will need to be authorized in the usual way and, if paid, show up in the participant’s part of the federal budget.  Certainly, this role will enhance or may even enable the project with respect to the benefits that non-federal entities are willing to pay for.  But isn’t that the point of federal involvement in the first place?

• More specifically, for the FCRA non-federal issue, the federal participant’s “level of control and future responsibility, or liability” needs to be put in the context of the project’s non-recourse debt and program loan repayment.  Any project finance lender would want to see that a project’s developers, managers, and owners are willing to commit to exercise control and assume the responsibilities and liabilities that are needed to make the project successful, even though the loan is not recourse to any of them.  Maybe the federal participant can and will take an out-sized role relative to the others – but that’s a difference in degree, not principle.  Regardless of how excellent and beneficial the project is due to the federal participant’s capabilities and efforts, an independent decision must be made by the project’s non-federal beneficiaries to commit their non-federal resources to make long-term payments to the project.  Their decision, not those of the federal participant, is the basis of the project’s non-recourse debt repayment and therefore the source of external, non-federal discipline for the program loan’s FCRA treatment.  

• The ’former’ involvement of the federal government in aspects of the project (primarily land or existing large-scale federal infrastructure, I think) should be seen in the same context with respect to the specific FCRA non-federal issue.  Such historical facts may increase the federal participant’s role or responsibilities in a new project.  But no matter how large, these are in a sense ‘sunk costs’ – the marginal improvement attainable by the new project construction is what the project’s non-federal beneficiaries are willing to pay for, and thus the non-lender recourse debt’s source of repayment.  A concessionaire’s gift shop in a federal park is a tiny fraction of the park’s total operations, but that doesn’t make the shop’s loan from a local bank a ‘federal loan’.  The federal government is de facto committed to certain kinds of disaster relief that make it possible to obtain mortgages on structures built in flood plains where flood insurance is unavailable – are those mortgages ‘federal debt’? 

OMB staff reported that they developed these criteria based on the principles outlined in the 1967 Report of the President’s Commission on Budget Concepts.  OMB staff stated that OMB applies concepts from the report to determine if an activity is federal in nature, and thereby ineligible for special budgetary treatment under FCRA.  If an activity is determined to be federal in nature, then an agency must seek appropriations to pay for the full value of the activity and record its full liability for the activity consistent with the recording statute.

• This paragraph simply states what is already known, that OMB used the 1967 Report as a primary source, per the Congressional directive, and that FCRA treatment is not intended for federal borrowers.  But it’s quoted here in full because of the specific use of the word ‘activity’.  If the relevant activity is precisely defined as a ‘program loan that finances infrastructure construction for beneficiaries who will use their own resources for loan repayment’, then the FCRA non-federal issue is somewhat straightforward:  FCRA treatment is justified when the beneficiaries and their resources are demonstrably non-federal (thereby imposing external discipline to loan repayment), otherwise not. With that definition of ‘activity’, the paragraph makes sense with an additional clarification that the ‘agency’ is in fact the loan program itself.

• However, if the activity is defined as a ‘program loan to a project that might be considered ‘federal’ due to federal participation, in which case the loan is also necessarily ‘federal’, then I think the FCRA non-federal issue ‘could easily carry on into quite esoteric matters’, as the 1967 Report put it.  And those matters would seem to go well beyond program loans and FCRA.  If a complex, multi-party project financing with a federal participant, but without a program loan, is considered ‘federal’ per the OMB’s criteria, then should all the project’s non-recourse debt be considered ‘federal’, regardless of purpose or repayment source?  In this case, FCRA is not relevant because there’s no program loan, but shouldn’t the federal participant (as the ‘agency’ referred to in the paragraph) be recording the ‘full liability’ of the project anyway?  If not, why?  What if the non-recourse debt includes tax-exempt bonds — are these ‘federal’ tax-exempt bonds? And does scale matter?  What about a bank loan to a privately owned gift shop on the project’s grounds?  And so on.  Unfortunately, it seems that OMB was loosely using the latter definition of activity, or some variation thereof.   As a result, I think their criteria inevitably slid into a rabbit hole. The lack of a precise definition of ‘activity’ is a fundamental — but easily fixable — problem with the current criteria. More on this in future posts.

Other Items in the GAO Report

• GAO notes that OMB publishes government-wide criteria for various aspects of federal loan program implementation and budgeting (e.g., Circulars A-11 and A-129), but not yet for the specific FCRA non-federal issue.  This is apparently due to the infrequency of the issue arising, even at WIFIA.  But that may change now that the Corps’ part of WIFIA authorization, CWIFP, is being implemented.  Was the urgency of the Congressional directive related to the expectation of Corps’ projects applying for CWIFP loans?

• WIFIA managers noted that the Fargo-Moorhead program loan would not have passed the current criteria, but details weren’t provided.  The disqualification of this transaction under current criteria will be an important data and illustration point for future post in this series.

• GAO lists the five other infrastructure loan programs they investigated with respect to the FCRA non-federal issue – TIFIA, RRIF, and three USDA programs for water, energy, and telecommunications, respectively.  None except RRIF encountered the issue.  The way RRIF described its analysis of the FCRA issue for a loan to Amtrak, arguably a ‘federal’ operation per OMB’s criteria [3], appears to be simpler and more consistent with the 1967 Report’s external discipline principle, and notes same the policy objective of reducing federal cost as in WIFIA’s selection criteria (emphasis added):

Another way that case study programs screen for federal participation in projects applying for loans is by evaluating the overall sources of financing for a project.  Ensuring a borrower can finance a project with some non-federal sources helps reduce the amount of fiscal risk to the federal government.  For example, DOT officials told us that as part of their loan evaluation, they ensured the 2016 Railroad Rehabilitation and Improvement Financing loan provided to Amtrak was fully repayable by Northeast Corridor revenues.  DOT officials assessed Amtrak’s available revenues and did not consider any federal assistance as a repayment source for the loan.

• GAO’s recommendation that OMB should publish government-wide criteria (again, implicitly based on those published for WIFIA) is succinct and unqualified. It should be recalled that the Congressional directive explicitly excluded a public comment period for the development of WIFIA criteria. Would that also be the case for government-wide FCRA criteria?

_____________________________________________________________________________________________

Notes

[1] The second part of the prior post in this series sketches out a classic project financing.

[2] This in fact describes the vast majority of WIFIA loans. The need for loan assistance to highly rated public water agencies financing basic, essential water infrastructure projects is not clear, nor is the additionality. I don’t think OMB has ever questioned the program’s compliance with their A-129 criteria in their course of annual reviews.

[3] Amtrak is apparently considered a federal ‘independent agency’. If it were evaluated by WIFIA’s current criteria for the FCRA non-federal issue as a ‘project’, it likely would be considered a federal project due aspects of federal exercise of sovereign power, ownership, control and liability. Presumably, if OMB’s current WIFIA criteria were in place government wide when RRIF was considering the Amtrak loan in 2016, the program would have had to reject it, despite the persuasiveness of their alternative analysis based on a non-federal source of repayment.

In 2018, three water advocacy organizations opposed the SRF-WIN Act. The underlying dynamics were probably more complex, but they stated their nominal objections in a letter to the Senate committee holding hearings on the Act and other water legislation being considered at the time. One of the letter’s points is that loans with the sub-Treasury rates proposed for smaller SRFs in SRF-WIN would have very much lower credit subsidy leverage ratios (loan amount/cost) than the current WIFIA program — about 3.7 and 8.5 times for the Act’s 50% and 80% Treasury rates, respectively. This was contrasted (with the usual rhetoric) to WIFIA’s 100:1 ratio for its highly rated loans executed at rates reflecting full Treasury cost. (Well, executed at full hypothetical cost on the day but not actually funded until years later — you know where this is going, right?)

How’s that looking now? On the surface, WIFIA’s $16 billion portfolio at FYE 2022 did indeed only cost about $160 million of discretionary credit subsidy appropriations. And, if calculated at current rates, those SRF-WIN sub-Treasury would result in leverage ratios about what the statement predicted — 3.8x and 8.7x, respectively, assuming that the SRF loans were as highly rated as those of WIFIA, almost certainly the case.

But what’s missing is the elephant — the cost of WIFIA’s mandatory appropriations to cover Treasury’s funding losses due to interest rate re-estimates. The average interest rate of WIFIA’s undrawn loan commitments at FYE 2022 was about 2%. If those loans were drawn on 9/30/22, when the 20Y UST was about 4%, the cost would have been $3.3 billion. As the loans are actually drawn, the cost might be lower — or it might be higher. But it will certainly be far more than the discretionary appropriations allocated to the portfolio.

The chart above shows the economic cost and leverage ratio of WIFIA’s portfolio at FYE 2022 compared to those of the SRF-WIN alternatives on the same day. WIFIA’s mandatory appropriations are now included. It can be assumed that the SRF-WIN alternatives would have used only discretionary appropriations because the SRF loans would have been drawn shortly after execution, whether Treasury rates had risen or fallen. Far from being 100:1, WIFIA’s actual leverage ratio is about 4.8x — a little better than the Act’s 50% Treasury option but much worse than the 80% option, which was more likely to be utilized.

Ah, the elephant. It was perhaps predictable in 2018. But now it’s obvious in the actual results of WIFIA’s current $16 billion portfolio of sub-Treasury loan commitments. Once pointed out, it’s hard not to see it as a major policy error. Hard, but not impossible — just don’t talk about it, okay?

Here’s some slightly modified, cut-and-paste language from CIFIA law that would add the interest rate limited buydown feature to WIFIA’s statute. I think this would work as-is for CWIFP, maybe a further modified version would be needed for SWIFIA.

LB-amendment-draft-11152022-InRecap