Approach to LIHTC Regs Should Be Reconsidered, Individuals Say

December 28, 2020

Dillon J. Taylor
Attorney
Department of the Treasury
Internal Revenue Service
1111 Constitution Avenue NW., Room 5107,
Washington, DC 20224
dillon.j.taylor@irscounsel.treas.gov

Re: Comments on Notice of Proposed Rule Making, Section 42 Low-Income Housing Credit Average Income Test Regulations
REG-119890-18, RIN: 1545-BO92

Dear Mr. Taylor:

Thank you for promulgating the proposed regulations addressing the average income test (the “Average Income Test”) for low-income housing tax credit (“LIHTC”) projects provided in Section 42(g)(1)(C). In response to the Internal Revenue Service (the “IRS” or “Service”) and the Department of Treasury (“Treasury”) request for comments on the proposed regulations, we submit this letter. The persons signing below are the primary authors of this letter. We would also like to acknowledge the assistance of Judy Crosby of Kutak Rock LLP, Nicholas Anderson of Lathrop GPM LLP and Angela Christy of Faegre Drinker Biddle & Reath LLP. While we are all active members of the Tax Credit and Equity Financing Committee (the “Committee”) of the American Bar Association's Forum on Affordable Housing and Community Development Law (the “Forum”), and while we have consulted with other members of the Committee and the Forum, this request is not made on behalf of the Forum and has not been approved by the House of Delegates or the Board of Governors of the American Bar Association. Accordingly, it should not be construed as representing the position of the Association.

Brad Tomtishen
Tomtishen Aoun PLLC
Co-Chair of the Tax Credit and Equity Financing Committee

B. Susan Wilson
Enterprise Housing Credit Investments LLC
Co-Chair of the Tax Credit and Equity Financing Committee

---

OVERVIEW

We wish to thank the Department of Treasury and the Internal Revenue Service for promulgating the proposed regulations addressing the average income minimum set-aside for the low-income housing tax credit. This provision was added in 2018 and guidance will help remove uncertainty with respect to how the provision should be used and applied. As discussed below, we think that there are a number of alternative approaches and considerations that should be reviewed before the proposed regulations are finalized.

PART I — COMMENTS REGARDING THE GENERAL APPROACH OF THE PROPOSED REGULATIONS

We include comments and suggestions on specific provisions of the proposed regulations in Part II of this letter. Prior to discussing those specific points, however, we respectfully request that the Service reconsider two aspects of the approach taken in the proposed regulations. These two aspects are at the heart of two principal concerns that we have about the proposed regulations: (i) the proposed regulations create the possibility of a complete loss of credits that was never intended, and (ii) the proposed regulations do not provide the flexibility necessary to practically administer a low-income housing project for which an average income election has been made. We believe that reconsidering the general approach of the regulations will allow the reconciliation of the Service's tax policy concerns with the need for certainty and flexibility in the administration of the LIHTC.

A. CALCULATION OF THE AVERAGE OF THE IMPUTED INCOME LIMITATIONS

The proposed regulations require that the average of the imputed income limitations of the low-income units in the project not exceed 60% of area median gross income (“AMGI”). We respectfully submit that this is contrary to the statute, which requires that the average of the imputed income limitations designated shall not exceed 60% of AMGI. IRC 42(g)(1)(C)(ii)(II). That statute further states that “[t]he taxpayer shall designate the imputed income limitation of each unit taken into account under such clause.” See IRC 42(g)(1)(C)(ii)(I).

The difference between the proposed regulations and the statute is significant. First, the proposed regulations approach is based on averaging actual rentals of units. While this may be a worthy intention, the language of the statute simply refers to the units designated by the taxpayer. Second, the interpretation in the proposed regulations potentially results in a project losing all of its credits as a result of the failure of a single unit to qualify as a low-income unit.

Example 1 The taxpayer owns a 100-unit project and designates imputed income limitations as follows: 90 units at 60% of AMGI, 5 units at 40% of AMGI and 5 units at 80% of AMGI. All units are rented in compliance with their imputed income limitations (and rent restrictions) except that one of the units designated at 40% of AMGI is leased to a tenant with an income equal to 45% of AMGI. In accordance with the statute, imputed income limitations have been designated for 100 units and the average of the imputed income limitations designated is 60%. 99 units satisfy the limitation applicable to that unit and the minimum set-aside is met. Under the proposed regulations, there are only 99 low-income units and the average of the imputed income limitations of the low-income units is 60.2% and the requirement of the average income set-aside is not met and the project does not qualify for LIHTC on any unit.

As the preamble recognizes, this result cannot possibly be intended by the statute. Accordingly, the proposed regulations propose various mitigation actions. The preamble discusses, and the proposed regulations give an example of, a situation in which a unit is out of service or uninhabitable. If that is the sole concern of the Service, we believe it can be addressed in an alternate manner, as discussed below. But by describing the test as requiring the average of the low-income units to be 60%, the proposed regulations encompass other failures to qualify as low-income units for which the mitigation provisions are inadequate, including, most importantly, the failure to satisfy income limitations and rent restrictions.

The minimum set-aside election contained in Section 42(g) establishes the income limitations for determining if a unit is a low-income unit. The income limitation applicable to a unit is then used to determine the rent restriction. A residential unit can qualify as a low-income unit only if it meets the applicable income limitation and rent restriction. Thus, the determination of whether a unit is a low-income unit can only be made once the applicable income limitation is known.

Compare how the other set-aside tests work. Can a unit leased to a tenant with income at 55% of AMGI qualify as a low-income unit? No, if the 20-50 set-aside has been elected, but, yes, if the 40-60 set-aside has been elected. In other words, the income limitations exist before, and apart from, a determination of whether the limitations are satisfied. A unit that does not satisfy its limitation is subject to a limitation; it is circular to argue otherwise. In the example above, 100 units, not 99, have imputed income limitations designated and the average of those designated limitations is 60%.

The possibility of a total failure of credits is extremely unattractive to investors and will put projects using the average income election at a disadvantage to projects using the 40-60 set-aside election. In practice, we expect that investors will demand a “buffer” so that the average of the designated limitations will be below 60%. In a situation similar to Example 1, investors will likely insist that 3 or 4 units be designated at 80% of AMGI, rather than 5 units. As a result, rather than allowing an average income project to be on equal footing with a project making the 40-60 set-aside election, the average income project will have aggregate rent restrictions that are more onerous than a 40-60 project. See Example 13, on page 18 of this letter for an illustration of how a buffer might be expected to work.

As noted above, the income limitations determine the rent restriction. Given the manner in which income restrictions are determined (see Revenue Ruling 2020-4), the maximum LIHTC rent that may be collected on 2 units subject to a 60% AMGI limit is the same as the rent that may be collected on 2 units with designated imputed income limitations that average to 60%. Requiring a buffer so as to assure that the failure of a small number of units will not result in the average AMGI rising above 60%, yielding a catastrophic loss of credits, will have the ironic effect of reducing the total rental income of those projects for which the flexibility offered by the average income election is most critical. By reducing the rental income of such projects, they can support less commercial debt and thus will be less feasible.

We think that by focusing on the language chosen by Congress, the above problems can be avoided. Section 42(g)(1)(C)(ii)(I)-(II) provide the following:

42(g)(1)(C)

(ii) Special rules relating to income limitation

For purposes of clause (i) — 

(I) Designation

The taxpayer shall designate the imputed income limitation of each unit taken into account under such clause.

(II) Average test

The average of the imputed income limitations designated under subclause (I) shall not exceed 60 percent of area median gross income.

Based on the plain language of the Section 42(g)(1)(C)(ii)(I) and the above discussion, we believe that the better reading of this provision is that the 60% average is computed based on the designations made by the taxpayer rather than having the actual rental of a unit possibly result in that unit not being included in the computation.

To be clear: units not in compliance would not generate current tax credits and would be subject to recapture. However, solely for purposes of assuring that the 60% average test had been passed, the plain wording of Section 42 providing that the taxpayer's designations should be used.

It may be that certain theoretical policy issues regarding the application of the average income set-aside may concern the Internal Revenue Service and the Department of Treasury, but we believe that such issues can be adequately addressed by application of the statute as written. We discuss these below.

First, there may be a concern that the average income set-aside could be subject to abuse. A taxpayer could ignore restrictions on units designated with income limitations below 60%, charging market rents and foregoing credits on those units while claiming credits on units designated with income limitations above 60%. This situation could be addressed with a general anti-abuse provision allowing the Service to disregard designations made in bad faith and/or requiring an annual certification made to the Service or the State Housing Credit Agency (“State Credit Agency”) that the taxpayer believes, in good faith, that all designated units are in compliance with their applicable income limitations. See the discussion below for further discussion of a such an approach. A determination that designations would not be respected absent good faith attempts at compliance would provide a powerful deterrent to abusive behavior. We also note that there are enforcement mechanisms contemplated by the statute other than the loss of credits, including the restrictions in the extended use agreement that may be enforced by the State Credit Agency or any tenant. And, we note that the market provides an alternate enforcement mechanism through the tension between interests of developers and investors in a LIHTC project — tax credits accrue to the benefit of investors while cash flow from higher rents accrues primarily to the benefit of developers.

Second, as described in the preamble and the proposed regulations, there are situations in which units could be taken out of service as a result of casualty or otherwise. Once it is determined that a unit is no longer in service or otherwise not habitable, it should be appropriate to allow re-designation of units as well as mitigation of the type described in the proposed regulations. If the regulations require an annual certificate, the instructions could provide that it would be bad faith to certify that designated units are in compliance if those units are not in service unless appropriate re-designations or mitigation has been put in place. As described below, we think it is appropriate to provide mitigation similar to the next available unit rule for casualty losses and similar events that are outside the control of the taxpayer. Another alternative discussed below would be to allow the income designations to be used for units suffering a casualty solely for the purpose of determining the average of such designations. Such units would not count for meeting the 40% minimum set-aside and standard rules relating to casualty losses would apply for the purpose of determining current tax credits or the applicability of recapture.,

Third, there appears to be an assumption in the proposed regulations that the quid pro quo for taking LIHTC on a unit with a limitation above 60% AMGI is the actual compliance of a unit that has a limitation below 60% of AMGI, so that if a unit with a limitation below 60% AMGI fails to qualify as a low-income unit, a corresponding unit (or units) with limitations above 60% AMGI must also fail to qualify for credits. We do not see any indication in the statute that is the case and we respectfully submit that the policy arguments that we provide here outweigh the policy arguments that appear to have motivated the proposed regulations. It seems unlikely that Congress intended that the consequence of noncompliance on one unit was the loss of all credits. Instead, we respectfully submit that the legislature more likely expected that regulations would be drafted to avoid total credit loss and instead substitute credit loss on one or two additional units.

Consider a simple example in which all units have designated imputed income limitations of 60% except for one unit designated at 40% AMGI and one unit designated at 80% AMGI. We contend that the quid pro quo for taking LIHTC on the 80% unit is agreeing to the lower imputed income limitation, and the corresponding rent restriction, on the 40% unit.

In summary, we do not believe that the statute supports the interpretation of the minimum set-aside set forth in the proposed regulations. Concerns about abuse of the provision can be addressed directly with anti-abuse or good faith provisions such as the good faith test we describe in the following section. Changes in facts, such as destruction of units can be addressed by changes in designations which, as described below, we believe should be allowed in many circumstances. In almost all cases, violations of the applicable income limitations, and the associated rent restrictions, will be inadvertent and made in good faith and will be subject to penalty through the loss of credits on those units that do not qualify as low-income units. We do not believe the statute intended, or tax policy requires, that failure to qualify one unit as a low-income unit results in the draconian remedy of loss of all credits, or even the loss of credits on one or more additional units that have qualified as low-income units.

B. GOOD FAITH RELIANCE ON TAXPAYER'S DESIGNATIONS ALLOWED AND CAN AVOID NEED TO MITIGATE

As described above, we believe that Section 42(g)(1)(C) provides that, for purposes of testing compliance with the 60% average requirement of Section 42(g)(1)(C)(ii)(II), the income designations used would be those made by the taxpayer. To address policy concerns, including potential abuse of the provision, we propose that the designations made by the taxpayer would not be respected unless the taxpayer certifies and establishes that it made a good faith effort to comply with the designations. We would propose a rule as follows:

If, during a taxable year, there is a determination that the rental of a unit does not comply with the income designation assigned to such a unit or the unit did not otherwise meet habitability or other requirements to be eligible for tax credits, for purposes of Section 42(g)(1)(C)(ii)(II) such determination shall not apply to any period before such year and Section 42(g)(1)(C)(ii)(II) shall be applied without regard to such determination if (1) the taxpayer certifies and establishes a good faith effort to comply with Section 42(g)(1)(C) and other requirements of Section 42, and (2) the failure is corrected within 1 year from the date of the determination.¹ The good faith of a taxpayer may be established by showing that the Project complies with the Available Unit Rule of Regulation 1.42-5(c)(1)(ix) and the General Public Use Rule of Regulation 1.42-9 or other criteria in the judgment of the State Housing Credit Agency. A taxpayer attempting to rely on a designation for which it purposefully did not comply would be deemed abusive and such designation would then be disregarded for purposes of determining if the project complies with the 60% average requirement of Section 42(g)(1)(C)(ii)(II).

The foregoing rule would provide for the certainty as to designations by taxpayers as prescribed in Section 42(g)(1)(C)(ii)(I), prevent taxpayers from abusing such designation and also allow for a reasonable time where such designations could be relied on while errors are corrected. Importantly, a unit not following the designation will not generate tax credits and can cause recapture even though a good faith effort was made to comply with Section 42 requirements, but the satisfaction of the minimum set-aside will not be unknowingly threatened.

Examples of the foregoing rule are below:

C. FLOATING UNITS

The proposed regulations require that unit designations be made by no later than the end of the first year of credit period, and once made, no changes to such designations are permitted. Prop. Reg. 1.42-19(a)(3). Permanently fixing unit designations is inconsistent with other provisions of Section 42, policies and procedures of State Credit Agencies, and statutory tenant protections in a low-income housing project. For the reasons discussed below, we believe the better approach would be to allow State Credit Agencies to set policies and procedures providing for how unit designations can be changed.

We do not believe there is any reason not to allow the withdrawal of temporary designations, or modifications to designations as long as the owner follows State Credit Agency procedures as to how this is to be done. We believe that such an ability is an important component of the long established ability of low income projects to float which units are low income and which are not.

We recommend a reasonable middle ground approach, consistent with other Section 42 guidance. Changes to unit designations should be permitted in all events under the tenant protection examples provided above. Outside that scope, changes to unit designations should be permitted where the taxpayer demonstrates to the State Credit Agency reasonable grounds for the requested change pursuant to procedures established by the State Credit Agency. By analogy, Section 42 guidance on other matters including the next available unit rule and recent 8609 guidance/discussions with the IRS indicate that such an approach is supported. Extreme cases of inappropriate use of unit designation changes could be identified by the state State Credit Agency and such requests would be rejected. The anti-abuse rules of the Code can also address “bad-actor” activity. As we have described, there are many other instances in which a change to the unit designations is warranted and if not made, could cause unintended consequences for the owner.

In summary, our recommendations are as follows:

PART II — DETERMINING WHEN MITIGATION, TIMING OF MITIGATION AND TYPES OF MITIGATION

A. TIMING OF MITIGATION

The proposed regulations allow a taxpayer to elect to take mitigating actions in order to avoid failing the Average Income Test due to a non-compliance event. However, these mitigation actions need to be taken within 60 days after the end of the year when the non-compliance event occurred. This deadline is problematic and will often render such actions ineffective.

Often, a violation is inadvertent and is not detected until after the end of the year. The discrepancy could be discovered during the annual audit, which can frequently take longer than 60 days following year end, or upon the compliance review conducted by the State Credit Agency, which have varying due dates. In either case, discovery of the non-compliance issue is likely to come more than 60 days following the end of the year the noncompliance event occurred.

We believe that the mitigation period should begin upon the discovery of the noncompliance event. The taxpayer could be required to notify the State Credit Agency of the discovery of an instance of noncompliance and the chosen course of mitigation. The notification would then be attached to the taxpayer's subsequent tax return with the average income calculation and tax credits computed reflecting the mitigation action. In addition, we believe that the mitigation period should be longer in order to properly assess and remedy the situation. We suggest that it continue until at least the earlier of the due date of the tax return with extensions or one year from the date the noncompliance event is discovered.

A mitigation period of one year from the date of the discovery is consistent with the cure period provided by Internal Revenue Code Section 42(h)(6)(J) that allows a taxpayer to correct a noncompliance event within one year of the determination of the noncompliance. We think this one-year period could be treated as grant of relief under Treas. Reg. Section 301.9100, and the taxpayer could file for the relief pursuant to and in accordance with Treas. Reg. section 301.9100-2. Such regulation already provides automatic relief for a number of different regulatory elections, and taxpayers would benefit from the implementation under a commonly understood and straightforward process. We recommend a period of one year and not a lesser period such as 90 days allowed in Regulation 1.42-5(e)(4). A period as short as 90 days may be insufficient to allow the market rate conversion mitigation to be useful as an owner may need to wait for a market rate tenant's lease to expire to allow it to then rent that unit to a low-income tenant. Therefore, we believe that the one-year period in Section 42(h)(6)(J) is more appropriate.

In addition, we believe there should be a default action for mitigation if none is elected by the end of the mitigation period. This would prevent the potentially catastrophic loss of all credits due to one unit falling out of compliance without discovery and corrective action being taken during the mitigation period. Our recommendation for the default mitigation action would be to remove sufficient units from the tax credit project in order to bring the average income of the low-income units to or below 60% AMGI. Of course, this automatic mitigation would not be necessary if our request to allow mitigation to be done up until one year after a determination that a unit is out of compliance.

If the proposals described in other portions of this letter are rejected, then we recommend that the proposed regulations be revised to by default have a unit removed to reduce the number of units eligible for tax credits to result in a percentage of 60% or less. By doing so, the tax credit project would then consist of 8 units with an average income of 60%, in compliance with the Average Income Test, and no longer putting the taxpayer at risk for ineligibility for the tax credits. This default option should only apply if the owner did not elect to take other mitigating actions by the end of the mitigation period.

B. TYPES OF MITIGATION

Mitigation Strategies in the Proposed Regulations. If one of the designated units ceases qualifying as a low-income unit — for example, because it becomes uninhabitable — then the proposed regulations provide mitigation remedies that can be employed to avoid failing the minimum set-aside, and which applies depends upon the specific facts.

As we discussed above, and setting aside the question of whether this remedy goes beyond what is called for in the Code, the removed unit remedy may not be such a useful solution. Eliminating units which are otherwise complying with the original land use restriction agreement and the requirements of the State Credit Agency will reduce the building's applicable basis, thereby reducing annual credits. We acknowledge that it is appropriate to eliminate the nonqualifying unit but adding a removed unit to the computation has at least twice the impact. Further, we agree with the favorable treatment accorded to the units if and when they are reinstated. However, and perhaps most important, if the nonqualified and removed unit (if applicable) cause the building to fail the minimum set-aside, then we can have the catastrophic result of the building no longer qualifying for any credits.

C. OTHER DESIGNATION ISSUES

Ability to change unit designations. As noted above, the proposed regulations contemplate designating a formerly fair market value unit as a low-income unit if that can solve a non-qualified unit problem. But they otherwise do not allow a change in designations. These mitigation strategies do not address other potential issues. What about tenants who are forced to relocate, or if another government program requires the designations to be changed? Or suppose a situation in which permitting an over 60% unit to be designated as a lower percentage unit (e.g., 40%) would have a far less adverse effect than removing a companion over-60% unit). Perhaps the proposed regulations did not contemplate “floating units” because the IRS thought they could be difficult to monitor and enforce. However, we observe that State Credit Agencies have been charged with assuring, and successfully addressed, compliance with the set-aside rules throughout the life of the LIHTC incentive, and we see no reason to not trust them with at role here as well.

Accordingly, we recommend that:

PART III — OTHER ISSUES

A. IRS PROPOSAL FOR NEXT AVAILABLE UNIT TYPE RELIEF

We agree with the concept of being able to correct an inadvertent violation by using a remedy that would require renting the next available unit to a tenant that would bring the average into compliance.

B. REMOVED UNITS NOT LOW INCOME UNITS AFTER THE CREDIT PERIOD

Proposed Regulation 1.42-19(f)(4) states the following:

(4) Long-term commitment. For purposes of applying section 42(h)(6)(B)(i) to any taxable year after the credit period, removed units are not taken into account as low-income units.

We request clarification on this point, especially how the computation of the applicable fraction would be computed in order to comply with the requirement under IRC 42(h)(6)(B)(i) that the applicable fraction continue to be maintained.

CONCLUSION

We sincerely appreciate the significant effort of the Department of Treasury and the IRS in writing proposed regulations in this complex area. However, we respectfully believe that modifications are called for. In particular, the proposed regulations create significant potential for projects to inadvertently fail the minimum set-aside, a risk that is much higher than with the other minimum set-asides under Section 42. The proposed regulations also significantly restrict the flexibility owners traditionally have (often with the participation of their State Credit Agency) and will likely create conflicts with other laws. Finally, the proposed regulations negatively impact the economics and financial feasibility of projects, which will cause investors to avoid projects that use income averaging.

We believe that the above issues can be avoided by (i) focusing on the statute's reference to the taxpayer making unit designations and permitting flexibility in such designations, while providing good faith and anti-abuse rules, as well as oversight by the State Credit Agency, (ii) providing for longer mitigation periods which would address real world timing issues, and (iii) continuing the long-standing tradition of allowing which units are low-income units to change or “float” from year to year by allowing the designations of units to change, generally with notice to the State Credit Agency.

Thanks very much for your consideration of these recommendations.