The "Average Income Test" or "income averaging" added a new technique to qualify a building as low income under the Low-income Housing Tax Credit (LIHTC) rules. Prior to the change in Section 42, a building could qualify for the LIHTC if 20% or more of its units are at 50% or less of area median or 40% or more of its units are at 60% or less of area median. The new Tax Code provision provides that a building can now have 40% or more of its units meet their designations, provided the designations average out to 60% of area median. Of course, once a building meets whichever of these minimum requirements it elects, it then computes its actual credits based on actual occupancy.
After income averaging became part of the statutory law, the IRS issued regulations on how it should be done, and there was a fair amount of criticism of the IRS's interpretation. It's all in the mathematics, and I'll give some illustrations below.
As I noted above, the Tax Code provision refers to whether or not you designate units that average out to 60% of area income in order for the building to qualify as low income. The occupancy requirement only arises when computing the 40% minimum and the maximum amount of credits.
On the other hand, the IRS's proposed regulations incorporate a "designated and occupy " test even for the purpose of doing the 60% computation.
To understand this distinction, consider a building with ten units, five are designated 50% units and five are designated 70% units. Suppose further that three of the 70% units are occupied by 70% tenants, and two of the 50% units are occupied by 50% tenants, with the rest occupied by over income tenants.
Under the first analysis, advanced by most in the tax community, we do these three tests—first, we make sure that the designations meet the 60% requirement. Here, the average of the designations is 60% (five times 50 plus five times 70 divided by ten), so the designations pass. Second, the building meets the minimum 40% requirement, because five of the ten units (i.e., 50%, which is more than 40%) are occupied by people who meet the designations for the units they live in. Third, the maximum credit is 50% of the building's basis, using the same 50% computation.
Under the IRS's proposed regulations, the analysis is not so positive—the regulations don't do the first two tests of the preceding example. Instead, they do the following computation—the designated and occupied units average out as follows: we have three 70% occupied units times 70 plus two 50% occupied units times 50 divided by a total of five units occupied in compliance with their designations, which is 62%. Since this is over 60%, the IRS view is that the building fails the minimum set-aside requirement.
To address the 62% problem illustrated above, the proposed regulations include a "two-for-one" cure, in which any time you lose a below 60% unit, you can just remove an offsetting above 60% unit from the computation and still be okay.
For example, in my preceding illustration, to make up for not having enough 50% units, we could take a 70% unit out of the computation. Now, our average is two times 70 plus two times 50, which is 60%. And the building is still low income because we have four out of the ten units counted and in compliance, which meets the 40% minimum set aside requirement. Of course, this is still not a good answer—the maximum computation of credits under the IRS's rule is now only 40% because we no longer count one of the 70% units. Compare my earlier computation, which indicates that the Code would seem to set the maximum credit at 50% on the same facts.
And don't forget that the 40% requirement is a cliff; the building must have at least 40% of its units qualify under the average income test or it can lose all of the credits. Here's where the IRS rule can be even more problematic. For example, suppose our project had three 70% units and one 50% unit all properly occupied, and the rest are occupied by over income tenants. Our first interpretation of the rules would call this building 40% low income and, therefore, qualified, while the IRS two-for-one interpretation would require us to back out two of the 70% units, leaving us with only one 70% unit and one 50% unit (so as to average out at 60%). Of course, with just two units out of ten in compliance, the building would fail the minimum 40% set aside and lose all of its LIHTCs.
To date, the IRS has stuck to this position (as expressed in the proposed regulations) and hasn't issued newer guidance. Nonetheless, we are seeing investors and developers getting comfortable with reducing the risk of the proposed regulations by setting the target percentage for designated and occupied units at something that averages out to less than 60%, say 55%." As a result, even if the project loses a couple of below 60% units, the average across all its remaining units will likely only creep up to 57% or 58%, and the building can still be okay without having to apply the two-for-one rule or risking a significant reduction in maximum credit percentage. Obviously, this technique requires a bit of sensitivity analysis to assure that the risk is as small as I am suggesting. This strategy is most likely to be successful with larger projects. As the number of units gets smaller, it becomes harder for the numbers to work out, as you saw with the ten-unit project of my illustrations.
There are additional issues in the proposed regulations besides the mathematical computation problems:
A few months back, I sent the IRS and Treasury about a dozen examples like the mathematics in this post to illustrate the perceived flaw in the proposed regulations. Of course, many others also voiced their concerns with the mathematics and the other problems at the IRS's hearing and in other submissions.
We're waiting to see if these comments have any effect and will share any guidance that the IRS provides on these issues and their impact on the affordable housing industry.