A Closer Look At IRS Tax Audit Selection Bias

Retired IRS economist Kim Bloomquist discusses his findings of geographic — and perhaps race and class — bias in the tax agency’s selection of audits.

This post has been edited for length and clarity.

William Hoffman: Welcome to Tax Notes Talk, Kim Bloomquist.

Kim Bloomquist: Thank you for inviting me.

William Hoffman: You worked for the IRS for 20 years and then a few years at the Taxpayer Advocate Service. Correct?

Kim Bloomquist: That’s correct. I worked about 20 years for the Office of Research, the first four of those in the Chicago office when they had a research office there. I then worked the other 16 or 17 so years in Washington, D.C. 

William Hoffman: Give us some highlights from that CV. What have you been doing since?

Kim Bloomquist: I retired in 2018 after my final three years with the National Taxpayer Advocate’s office. I retired in 2018 and moved out here to Washington state where it’s very rainy and nice most of the year.

William Hoffman: In a 2019 piece published in Tax Notes, you documented regional and racial bias in the IRS’s audit selection. This report has been cited by many since its publication, including in the New York Times. Would you summarize for us just briefly what you found?

Kim Bloomquist: I showed that many of those who claimed the earned income tax credit live in certain parts of the country. They’re not all scattered randomly around the country.

Historically I don’t think IRS has been that interested in looking at the geographic impacts of their enforcement policies for various reasons, including because it’s politically sensitive to highlight that. By looking at the geographic location of EITC claimants, one gets an idea also of the impact of the IRS audit policy.

In recent years, IRS audits have dropped off dramatically because they’ve been affected by budget cuts imposed on them by Congress. But the one area where audits have not fallen as fast is with EITC filers, mainly because these audits are fairly low cost to conduct. They can be done by low-graded personnel.

If you want to look at it from a business perspective, they’re fairly lucrative because they get a high probability of a tax change from these audits. The IRS, I think, is somewhat reluctant to keep cutting EITC audits when they have other budget problems.

William Hoffman: Could your 2019 analysis on regional bias also be interpreted or reveal a class bias on low-wage taxpayers who happen to reside disproportionately in those counties and an implicit bias in favor of high-income taxpayers who can afford to challenge the audit?

Kim Bloomquist: Yeah, I think one could infer that. I don’t think it’s intentional on the IRS’s part — at least I don’t think they’re trying to intentionally hold bias toward one ethnic group or class group based on income. From a business point of view, it makes sense for them to continue auditing low-income taxpayers because it’s reasonably lucrative.

But yes, I think that there is an overlap between the types of taxpayers that we’re targeting for this, which is mainly low-wage, working taxpayers. A lot of them are African American or Hispanic. I also found that in parts of the country where there’s high Native American populations, there’s high audit rates and EITC audit rates in those counties as well.

Any place where there is reasonably low-wage earners that have high EITC claim rates, the audit rate is also high. That is not only a concern from a fairness perspective. There was also a lot of interest on the part of certain members of Congress to find out that many taxpayers in their districts, their states, were being targeted at much higher rates than other parts of the country.

William Hoffman: You also wrote in March 2019 that the purpose of your report is to “show that willfully ignoring geographic location in audit case selection does not ensure balanced regional coverage.” What is the goal of balanced regional coverage?

Kim Bloomquist: Well, just a little bit of context: The IRS says that they’re not biased in the way they select audits because they don’t take a geographic location into account. They don’t take other racial or ethnic characteristics into account. They somewhat exonerate themselves by saying, “We’re not being biased because we don’t do these things.”

Unfortunately, that doesn’t really address the issue. Clearly there is a regional bias in the way they selected audits here; it just doesn’t show up in their audit policies. There’s somewhat of an implicit bias based in the way that their system selects cases for audit.

I think that this is something that the IRS needs to address in the future, or at least they need to be aware of it. I think members of Congress need to be aware of it and need to periodically remind the IRS that they shouldn’t be focusing all of their enforcement efforts in certain parts of the country to the exclusion, to a large extent, to other parts of the country.

William Hoffman: How does balanced regional coverage relate to audit selection fairness? Does it promote equitable coverage across both regions and income classes and levels?

Kim Bloomquist: The IRS is a national agency, and therefore they should apply their enforcement actions more or less equally across the country. They don’t have to be exactly equally, of course. There are certain parts of the country that specialize in certain things like agriculture or finance, and perhaps they run certain special programs that focus on those industries. There needs to be some flexibility in the way audits and the kinds of programs are conducted.

Nevertheless, for the broad number of taxpayers out there, the IRS should try to spread their audit activities more or less evenly across the country when they can. I think they need to focus on that as one of their fairness goals. 

William Hoffman: You also noticed in your research that the 2016 U.S. Government Accountability Office report criticized the IRS for not clearly defining the measures for audit selection fairness. Recognizing that fairness is a subjective standard, how would you define audit selection fairness?

Kim Bloomquist: What you need to do is to at least monitor the impacts of your audit policies. If you’re not doing that, and the IRS to my knowledge doesn’t, then there’s no way you can get a clear indication of who you’re impacting, where these people live, and whether or not you should be doing something different.

I think it’s important for an agency like the IRS to be perceived as being fair. When so many of their audits are focused on relatively low-income taxpayers concentrated in certain areas of the country, the IRS needs to be aware of it and to take into account their perception that they’re not being fair, however that’s defined.

In tax year 2015, about 40 percent of all audits were conducted on EITC claimants. That didn’t make up much less than that, of course, of all filers. They’re probably closer to 15 percent of all filers. They’re being selected at three or four times more than their representation in the population. That could be perceived as unfair by many people.

William Hoffman: What elements precisely would a fairness measure include? Balanced regional coverage and what else? What measures of fairness would be both realistic to implement and helpful to taxpayers?

Kim Bloomquist: I would say that, again, a relatively even distribution of audits among income classes regionally ought to be a consideration. Not necessarily that it has to be exactly the same across the country, because certain types of activities are specialized in certain regions of the country. There should be a regional assessment of audit impacts. There should be an assessment of audits by category of adjusted gross income. Those are probably the two main areas.

Again, I would say because the IRS has conducted tax gap studies over the years, they’re aware of where the majority of unpaid tax originates from. That’s primarily in the top 10 percent to 20 percent of the income distribution and primarily from business taxpayers within that category as well.

I would think that if the IRS was focused on the tax gap, which has always been their agency’s mission, then they should be focusing on more business oriented taxpayers, which I know are more expensive to audit. But the IRS as well as Congress should be taking this into account in terms of advising an audit policy that actually addresses the problem of underpaid tax.

William Hoffman: You touched on my next question which is that the EITC seems to be at the heart of your analysis. You advise the IRS to institute broader audit selection criteria and maybe a bigger enforcement budget from Congress. But isn’t a big part of the problem with the EITC, the law itself?

Kim Bloomquist: I wouldn’t dispute that at all. I think it’s a very complex law for both taxpayers to follow and for the IRS to enforce. One of my first projects when I started with the Office of Research in 1999 was one of the initial EITC compliance studies. I’d have to go back and look at the numbers, but roughly a third or so, maybe 30 percent of overclaims of the EITC were happening in the mid-1990s. A decade later, the IRS did another study at the EITC noncompliance and it was basically the same.

Over a decade of auditing and other enforcement activities had little impact on EITC compliance, largely because it’s complex. The IRS does not have access to the data right away that it needs. For many years they did not have access to the data on eligible children, for example. That’s changed recently.

It’s also difficult for the IRS to ascertain business income sort of in real time. I would say in nearly two decades, the IRS has really had made little progress in terms of compliance among EITC taxpayers and to a large extent, because it’s a very complex and difficult area of the tax code to enforce.

There’s been some changes recently with the Protecting Americans From Tax Hikes Act of 2015 that requires refunds that are related to the EITC and the additional child tax credit to be held until mid-February before they’re paid out. That gives the IRS a greater opportunity to match reported incomes to any third-party reporting documents that they would receive. We won’t know the results of this for several more years until the IRS does another tax gap study specific to tax year 2016 and later.

It could be several more years before we hear about that. But because of two decades of poor progress that the IRS has made, not for lack of trying, it’s just really impossible for the IRS to do much about enforcement in this area. It might be time for the IRS to even consider or for Congress to consider removing programs like the EITC from the tax code and giving it over to other agencies like Social Security, who’s in charge of administering social support programs for decades and could probably do a good job with this.

We’d have to require an increase in budget for Social Security to administer this and we’d have to think through how to handle it, but I think it would be probably something that would be good for both the IRS and for the taxpayers involved because you wouldn’t be requiring audits of these folks anymore. People could go into their local Social Security office and file for the EITC, just like they do for Social Security.

William Hoffman: It’s an interesting idea, especially with the emerging social justice movement and the Democrats now looking at a Congress and a White House of their own. Do you anticipate any movement by policymakers to address the issues that you have identified with the EITC?

Kim Bloomquist: I certainly hope so. I know that following some of these congressional hearings on this issue back in 2019 that IRS Commissioner Charles Rettig came out with a very abbreviated audit plan, or he said it was best to increase audits of high-income, high-wealth taxpayers. But it’s only a fairly minuscule effort there because of budget constraints are still more or less imposed.

Perhaps with the new Congress and the new administration that will change and they will be able to make much greater progress. Maybe the IRS can get back to where it was over a decade ago, where it was auditing about 1 percent of all tax returns as opposed to like less than 1 percent. Now I think that will be good progress, particularly if they can target most of any increase in budget or again, the main source of the tax cap, which is the high-income and business tax filers.

I really think Congress needs to, because Congress has been actually talking about expanding the EITC and other programs like this. They need to be aware that doing so under the current approach is only going to exacerbate, I think, the number of filers who are caught up in this compliance web. And I think they really need to consider that if you’re going to greatly expand the number of EITC claimants out there, you’re also going to greatly expand the number of — or likely IRS will be tempted to expand the number of EITC audits.

Congress needs to be aware of that and they need to probably think about telling the IRS that this is something they don’t want to do or restructure the EITC in such a way to remove it from the tax code entirely and try to administer it in some other way,

William Hoffman: Does the incoming Biden administration need a new IRS commissioner to accomplish any of these goals?

Kim Bloomquist: I would say I leave that up to the new Biden administration. I suspect they probably will be looking for somebody to replace the current commissioner. It may not be high on their list of things to do as they seem to have lots of other priorities at the moment as well. It may be less important.

The commissioner role may be less important than increasing the budget for the IRS and then hiring people that they need to do a better job of assisting taxpayers and modernizing the IRS.

I know when I worked there we had always issues trying to modernize the computer systems and so forth. That’s certainly where their focus needs to be. I think the commissioner probably won’t be concerned or as much a concern in the short run as the budget situation will be.

William Hoffman: Have you been in contact with anyone from the incoming Biden administration?

Kim Bloomquist: No, I can’t say that I have.

William Hoffman: OK. What’s the first advice you’d offer about the IRS and tax administration generally, if they did contact you?

Kim Bloomquist: I would just reiterate what I’ve more or less told you, which is that if we’re going to focus on the EITC, then they need to consider what the enforcement implications will be by expanding it.

It may be a good idea from a social perspective, because with the growing inequality in this country, much more needs to be done to help out low-income taxpayers and low-income households from further sliding down into poverty. This is something we need to address dramatically in this country.

The EITC might be part of that. But I think there needs to be some further thought as to what the implications are from an enforcement perspective.

I would recommend — if I had the opportunity to say something to the incoming Biden administration — to consider restructuring the EITC and related credits so that the people who we’re trying to help out with these things are not disadvantaged by the IRS in terms of way they feel compelled to enforce the tax code.

William Hoffman: Would you be available to serve in the Biden administration if asked?

Kim Bloomquist: No, I don’t think so. Currently, my health would preclude that. I’m permanently retired.

William Hoffman: Fair enough. Good for you. Thank you very much for taking time for us.

Kim Bloomquist: Well, thank you again very much for inviting me. My pleasure.