---

Historically, the Treasury Department and the IRS have not required an Agency physically to inspect every low-income residential unit in a low-income project. Decades ago, the Treasury Department and the IRS determined that a sample was adequate if it included at least 20 percent of a project's low-income units, regardless of the total number of low-income units in the project. (T.D. 8430, 57 FR 40121, September 2, 1992).
 
The REAC protocol requires sample sizes that differ from those that the Treasury Department and the IRS had required. In developing that protocol, HUD sought to determine sample sizes that would yield equally reliable inferences regardless of the size of the number of residential units in a project. HUD's statistical analysis produced minimum sample sizes that are much lower than 20 percent of large projects' units but somewhat higher than 20 percent of total units for small projects. The implication of the HUD conclusions was that the tax regulations' 20 percent requirement for low-income housing credit inspections may have been unnecessarily burdensome for large projects and may have failed adequately to assess habitability in smaller ones.
 
In the temporary regulations, the Treasury Department and the IRS responded to that implication with a two-step process—minimum sample size was reduced for large projects, and taxpayers were asked whether analogous statistical considerations should be applied to increase minimum sample sizes for small ones.
 
One commenter responded that it was not concerned about ending the 20-percent rule for projects with a relatively small number of low-income units, because it is among those Agencies whose State or local rules require them to inspect a minimum number of units that exceeds the minimum numbers in Rev. Proc. 2016-15.
 
These final regulations remove the rule that allows minimum sample size to be the lesser of 20-percent of the total number of low-income units or the minimum unit sample size set forth in the Low-Income Housing Credit Minimum Unit Sample Size Reference Chart. Instead, under these final regulations, Agencies must inspect no fewer units than the number specified for projects of the relevant size as set forth in the Low-Income Housing Credit Minimum Unit Sample Size Reference Chart. The Treasury Department and the IRS have determined that the REAC numbers produce a statistically valid sampling of units, which establishes confidence in the compliance monitoring results for projects of varying size. The Treasury Department and the IRS have further determined that the REAC numbers reasonably balance burden on Agencies, tenants, and building owners with the need to adequately monitor habitability and compliance with the low-income housing credit income and gross-rent restrictions. Agencies, however, continue to have discretion to inspect and review more units as they see fit.

The temporary regulations require an Agency to select low-income units to inspect and low-income certifications to review in a manner that will not give advance notice that a particular low-income unit (or low-income certifications for a particular low-income unit) will or will not be inspected (or reviewed) for a particular year. The temporary regulations allow an Agency to give an owner reasonable notice that an inspection of the building and low-income units or review of low-income certifications will occur, whether or not an Agency is selecting the same units for inspection and for low-income certification review. The temporary regulations provide that reasonable notice is generally no more than 30 days, but they also provide a very limited extension for certain extraordinary circumstances beyond an Agency's control such as natural disasters and severe weather conditions.
 
The Treasury Department and the IRS requested comments on whether the same maximum amount of notice is reasonable for physical inspections as for low-income certification review. Additionally, the Treasury Department and the IRS requested comments on whether, for physical inspections, the reasonable-notice time frame should be shortened. For example, under the REAC protocol, an inspector provides a 15-day notice of an upcoming HUD inspection of a project but same-day identification of the units to be inspected. No comments were received.
 
These final regulations shorten the reasonable notice requirement to a 15-day notice that a project will experience an upcoming physical inspection or review of low-income certification. The Treasury Department and Internal Revenue Service believe that the 15-day notice period gives building owners reasonable notice that a review of low-income certifications will occur and gives building owners and tenants reasonable notice that a project will be inspected and that low-income units will be inspected if they are in the random sample that will later be selected.
 
The statistical validity of inspecting only a sample of the low-income units in a project depends on the sample being random and representative. Thus, the validity would be destroyed if a project owner had an opportunity to selectively prepare the units in the sample for inspection. Consistent with preserving the validity of the inspection process, an Agency must select the low-income units to inspect in a manner that will not give advance notice that a particular low-income unit will or will not be inspected. Accordingly, the final regulations clarify that an Agency may notify the owner of the particular low-income units for inspection only on the day of inspection. The Treasury Department and IRS note that, under the REAC protocol, HUD or HUD-Certified REAC inspectors randomly select low-income units for inspection on the day of inspection.