Around five months ago, in my 2017 Annual Report to Congress, I identified IRS audit rates and the distinction between “real” and “unreal” audits, as the fourth Most Serious Problem facing taxpayers. I had previously written about this topic in my 2011 and 2016 Annual Reports to Congress, and discussed it in a blog post six years ago.

So, what’s the deal with “real vs. “unreal” audits and why should you care?  I need to first give you a little background. Under section 7602 of the Internal Revenue Code (IRC), the IRS has the authority to examine any books, papers, records, or other data that may be relevant to ascertain the correctness of any return. I call these types of examinations, which can occur through correspondence, at the taxpayer’s home or business, or at an IRS office, “real” or traditional audits.

However, “real” audits don’t quite end the story. The IRS has several other types of compliance contacts with taxpayers that it does not consider to be “real” audits. These types of contacts, which I call “unreal” audits, include math error corrections, Automated Underreporter (AUR) (a document matching program), identity and wage verification, and Automated Substitute for Return (ASFR) (a non-filer program). Why are these types of contacts, which constitute the majority of IRS compliance contacts, important?  First of all, they require taxpayers to provide documentation or information to the IRS and may feel very much like a “real” examination to taxpayers. More importantly, “unreal” audits lack taxpayer protections typically found in “real” audits, such as the opportunity to generally seek an administrative review with the IRS Office of Appeals (Appeals) or the statutory prohibition against repeat examinations. And in case you are curious, the IRS is planning for the increased use of “unreal” audits through automated means with its “Future State” Initiative.

Let’s take a deeper dive into the distinction between “real” and “unreal” audits. In the Most Serious Problem, I raised the concern that this distinction results in the IRS publicly reporting incomplete and even misleading information, as the IRS only reports “real” audit statistics, which skews the coverage rate and understates the IRS’s actual level of compliance contacts with taxpayers. To give you a sense of the numbers, let’s take a look at the most recent year in which we examined both “real” and “unreal” audit figures. In fiscal year 2016 (October 1, 2015 through September 30, 2016), the IRS conducted slightly more than a million “real” audits, resulting in an audit rate of 0.7 percent. However, during the same timeframe, the IRS conducted approximately 8.5 million “unreal” audits. When adding these “unreal” audit numbers to the “real” ones, the IRS’s combined coverage rate jumps to over six percent. I find this eye opening. To give you a visual of these figures, see Figure 1 from the Most Serious Problem.

Figure 1, Real vs. Unreal Audits: FY 2016 Occurrences Compared to Returns Filed in Calendar Year 2015

And speaking of incomplete information, I also pointed out that the distinction between “real” and “unreal” audits causes the IRS to not completely and accurately report its return on investment for compliance activities. This is because the IRS does not include all “unreal” audit programs in its return on investment calculations.

Another issue, and a far more important one, is that the distinction between “real” and “unreal” audits limits a taxpayer’s ability to seek Appeals review. A taxpayer who disagrees with an “unreal” audit’s proposed assessment generally receives a statutory notice of deficiency but does not have the opportunity for independent Appeals review. I noted that this issue is even more pronounced in the “unreal” audit category of math error cases, where a taxpayer must respond to an IRS notice in a limited timeframe (60 days) or lose the opportunity to go to Tax Court — the only prepayment forum available. Contrast this with what takes place in “real” audits, where a taxpayer generally receives a 30-day letter offering an opportunity to go to Appeals prior to the IRS’s issuance of a statutory notice of deficiency and petitioning the Tax Court.

I find this lack of Appeals review in the “unreal” audit context to be quite troubling and one that directly and profoundly impacts fundamental taxpayer rights, including the right to challenge the IRS’s position and be heard, the right to appeal an IRS decision in an independent forum, the right to finality, and the right to a fair and just tax system. These taxpayer rights issues become even more glaring when you consider the “unreal” audit income demographics – “unreal” audits disproportionately impact low and middle-income taxpayers. These are the taxpayers who are least able to afford representation to challenge the IRS.

Finally, as I note above, I believe that the “real” versus “unreal” audit distinction circumvents statutory taxpayer protections from repeat examinations (see IRC section 7605(b)). Under the IRS’s current guidance, taxpayers subjected to an “unreal” audit may face additional “real” or “unreal” audits. While I believe that the IRS needs to conduct “unreal” audits for limited issues, its current position allows it to conduct an “unreal” audit and later conduct additional “real” or “unreal” audits. This is not fair to taxpayers. In the Most Serious Problem, I provide an example of this in the Affordable Care Act context, where the IRS requests substantially similar documentation in both an “unreal” and “real” audit, thereby circumventing statutory protections against repeat examinations.

What should the IRS do about this problem?  I believe it can do a few things to ease taxpayer burden. First, the IRS should work with my office and conduct a comprehensive review of its audit definition under current guidance to reflect IRS compliance activity today, as well as the application of the Taxpayer Bill of Rights. Second, the IRS should include “unreal” audits in its audit (or coverage) rate and return on investment calculations to properly reflect the actual compliance activity that it conducts. Third, the IRS should allow taxpayers the opportunity to seek Appeals review in certain “unreal” audit cases, such as in certain math error and AUR cases where Appeal rights do not already exist. Finally, where practicable, the IRS should address all issues in a “real” audit rather than conducting an “unreal” audit and then subsequently conducting a “real” audit.

You can read more about my concerns about IRS audit rates, “real” vs. “unreal” audits, and my recommendations to help address the issues in the Most Serious Problem: Audit Rates: The IRS Is Conducting Significant Types and Amounts of Compliance Activities That It Does Not Deem to Be Traditional Audits, Thereby Underreporting the Extent of Its Compliance Activity and Return on Investment, and Circumventing Taxpayer Protections.

A final word about this Most Serious Problem. In volume two of my recently released June Report to Congress, I published the IRS’s responses to my recommendations for this Most Serious Problem as well as my rebuttal to these responses. In short, the IRS did not adopt my four recommendations, and I reiterated the concerns that I expressed above. You can read more about this here.