Guest Blogger: Carlton Smith
Since 1989, when the section 6662 accuracy-related penalty was created and the 6663 civil fraud penalty was rewritten, the IRS has been required to impose them on the “underpayment”, as defined in section 6664(a). Thousands of times a year, the IRS has issued notices of deficiency that not only disallow a refundable tax credit (such as the earned income tax credit, the additional child tax credit, or the first-time homebuyers credit), but has treated the disallowed credit as part of the accuracy-related or fraud penalty underpayment. In a stunning court-reviewed opinion issued yesterday by the Tax Court in Rand v. Commissioner, only two judges agreed that the IRS has been appropriately imposing these penalties in the case of disallowed refundable credits. Thirteen judges told the IRS that refundable credit disallowances, either partly or fully, are not part of the “underpayment” on which the penalties can be imposed. I was invited to do a blog post on the Rand case since I filed an amicus brief in Rand on behalf of the Cardozo Tax Clinic, and my arguments were adopted by the 10-judge majority position.
Before getting into the facts of Rand, it is important to note how many notices of deficiency involve this issue. So, practitioners – particularly those who represent low-income taxpayers who claim refundable credits – need to go back and look at notices of deficiency issued in recent years that may present the issue. I know that, when I ran the Cardozo Tax Clinic, I found that there were always one or two cases in the clinic that presented the Rand issue. That’s why I filed an amicus brief. I didn’t want a poorly-reasoned opinion to affect my clients. While the Taxpayer Advocate Service does not regularly publish statistics, National Taxpayer Advocate Nina Olson noted in 2001 that in tax year 2000 alone, the IRS issued 17,300 notices of deficiency disallowing EITCs that also asserted either the accuracy-related penalty or the fraud penalty on those disallowances. TAS had been monitoring the Rand case, and already, earlier this year, it had been trying to get a handle on how many notices of deficiency currently would be affected if the taxpayer won Rand.
The facts of Rand are fairly simple: For the 2008 year, the Rands reported almost $20,000 of compensation income (mostly wages). Against this income, they took dependency exemptions for a number of children. The exemptions were sufficient to bring taxable income down to zero. So, on their return, the Rands reported their section 1 income tax as $0 and their self-employment tax as $144. On their return, they also claimed EITCs of $4,824, additional child tax credits of $1,447, and a section 6428 recovery rebate credit of $1,200. All three credits were reported in the “payments” section on the return, since they were refundable credits. Under section 6401(b)(1), the excess of refundable credits over the sections 1 and 1401 taxes are considered an “overpayment” that can be credited or refunded under section 6402(a). The three refundable credits totaled $7,471. After subtracting the $144 of self-employment taxes under section 1401, the Rands sought — and were paid — a refund of $7,327.
When the IRS audited the Rands, it disallowed all of the credits on the grounds that none of the reported compensation income was actually compensation income – i.e., earned income. Thus, it issued a notice of deficiency stating that there was a deficiency of $7,471. The three credits only were available if the Rands had earned income. The IRS also sought an accuracy-related penalty of 20% on the $7,471 of disallowed credits — taking the view that the “underpayment” under section 6664(a) on which the penalty was imposed was calculated the same way as the “deficiency” was calculated under section 6211.
In the Tax Court, the Rands conceded that they had no earned income, that the deficiency was thus correct, and that they did not have reasonable cause and good faith to avoid any penalty. However, they asked the Tax Court to hold that there was no accuracy-related penalty because, unlike section 6211, which had specific provisions at section 6211(b)(4) that addressed refundable credit disallowances, section 6664(a)’s definition of “underpayment” did not include a provision dealing with refundable credits. They argued that section 6664(a), thus, did not include any refundable credit disallowances in the “underpayment” on which the penalty could be imposed.
In contrast, the IRS argued that, while section 6664(a) did not literally have a provision dealing with refundable credits, it should be read as if it did, since Congress in 1989 intended the definition of “underpayment” to largely track that of “deficiency”. The IRS also asked the Tax Court to give deference under Auer v. Robbins, 519 U.S. 452, 461 (1997), to the IRS’ interpretation of its regulation under section 6664 – even though that regulation did not specifically discuss refundable credits, either. Auer deference is given to an argument an agency sets out — even for the first time in its briefs — if the issue is the agency’s interpretation of its own, ambiguous regulations.
As an amicus on behalf of the Cardozo Tax Clinic, I made a third argument in the case: Both section 6211 and section 6664(a) define the “deficiency” and “underpayment” as, essentially, the excess of the correct “tax” over the “tax shown on the return”. In this case, the parties agreed that the correct tax was $144 because no credits were actually allowable. Where I parted company with the parties was in the definition of the “tax shown on the return”. I argued that section 6211 does include in the calculation of deficiency most credits (including refundable credits like the EITC, ACTC, and FTHBC) that bring the “tax” down to zero. However, Congress assumed that, absent a provision like section 6211(b)(4), the “tax shown on the return” could not be negative. In 1988, Congress amended section 6211 to add subsection (b)(4) that essentially provides that the excess of refundable credits over the tax (computed without such credits) should be added into the deficiency calculation as a negative amount of tax. The way I saw it, without counting the rule of section 6211(b)(4), the “deficiency” and “underpayment” would be the excess of the correct tax of $144 over the tax shown on the return of zero – zero being the amount shown on the return after refundable credits reduced the tax to zero by being counted as part of the “tax”. Under my view, the rule of section 6211(b)(4) merely added $7,327 (the amount refunded) to the deficiency as a negative amount of tax. In that way, adding $144 plus $7,327 created the “deficiency” of $7,471 – not coincidentally the sum of all refundable credits. I argued that since section 6664(a)’s definition of “underpayment” lacked a provision like section 6211(b)(4) providing for adding in negative amounts of tax, the “underpayment” in Rand’s case could only be $144. Thus, the correct penalty was 20% of $144, or $29. The Rands actually agreed with my position as their backup position, though they spent little time in their briefs explaining it.
Suffice it to say that the Rand majority in the Tax Court agreed with my statutory construction analysis – holding that Congress in 6211 did not think the “tax shown on the return” could be negative; otherwise, section 6211(b)(4) would be surplussage. And section 6664(a)’s definition of “underpayment” should likewise be read as not letting the “tax shown on the return” go below zero – absent some statutory or regulatory provision like section 6211(b)(4). The majority said that there was no provision like section 6211(b)(4) in section 6664 and that the regulation under section 6664(a) was silent as to modifying the definition of “underpayment” to deal with refundable credit disallowances. The majority refused to give Auer deference to the IRS’ interpretation of its regulation, finding that nothing in the regulation dealt with refundable credits, so there was nothing to interpret. As a caveat, the majority warned that it was not holding that the statute’s definition of “tax shown on the return” was “ambiguous” — as that term is used under Chevron analysis. Under Chevron, in step one, a court determines whether a statute is ambiguous. Only if it is ambiguous, the court than defers to any reasonable agency interpretation of the statute set out in regulations. In effect, the Tax Court warned the IRS that simply promulgating a regulation to overrule the holding in Rand might not work.
Judge Gustafson and two other judges dissented – agreeing with the taxpayers’ primary argument that no refundable credit can be used in computing the “underpayment” because refundable credits are treated like payments, so are not part of the calculation of the excess of the correct “tax” over the “tax” shown on the return.
Judge Morrison was the original judge to whom the case was submitted fully stipulated. He wrote a dissent joined by one other judge, agreeing with the IRS.
I had written an article in Tax Notes this past February, “IRS Wrongly Ignores the 20 Percent Excessive Refund Penalty” (subscription required), 2013 TNT 39-10. In the Tax Court, part of the IRS’ argument was that reading the statute as either I or the taxpayers did would deprive the IRS of civil penalties to sanction widespread inappropriate claiming of refundable tax credits. The IRS argued that this made no policy sense. In the briefs in the Tax Court, neither the taxpayers nor I responded directly to this IRS argument – other than to say that a penalty statute must be enforced exactly as written and construed in favor of taxpayers under the rule of lenity. We did not want to point out to the IRS that it had other penalties it could use to sanction the Rands. However, in my article, I responded to the IRS policy argument – noting that in section 32(k) one can find the civil sanction for improper claiming of the EITC. Section 32(k) provides a disallowance of taking the credit in the following two years if an EITC was erroneously claimed as a result of recklessness or intentional disregard of rules and regulations (10 years, in the case of fraud). And in 2007, Congress adopted section 6676, which provides a 20% penalty for excessive refund claims that are not subject to section 6662 penalty and do not arise from the EITC. So, the IRS can assess the 6676 20% penalty against people who claim excessive ACTCs and FTHBCs. It appears that Judge Buch, writing for the majority, may have read my article, since he made these very same points about the alternative penalties near the end of the court’s opinion – that the IRS was wrong that by ruling against the IRS the world would end.
It is not yet clear what the IRS’ response to the Tax Court’s Rand opinion will be. Rand is appealable to the 7th Circuit. I hope the IRS just acquiesces – seeing that it could only get two Tax Court judges to favor its position. But, I fear the IRS may declare war on all Rand-type cases coming out of the Tax Court – like the IRS did in the innocent spouse Lantz cases and the section 6501(e) Home Concrete cases. It would be a shame if pro bono attorneys and tax clinics are forced to defend the Tax Court in every Circuit while the IRS looks for a Circuit split to take to the Supreme Court.
Speaking of tax clinics and pro bono attorneys, kudos need to go out to Andy Roberson and his crew at Latham & Watkins, who first took on the Rand case pro bono and knew enough to make this argument. Andy moved mid-case to become a partner at McDermott Will in Chicago. Once again, like in Lantz, pro bono attorneys and clinicians have stepped in to litigate issues that pro se taxpayers would never have known to have raised.
One final matter: I wrote in my article that I think the current penalty regime, as written, is in need of some statutory adjustment and coordination. It is particularly odd that the IRS should be relegated to asserting the 6676 penalty against the Rand taxpayers, since that is an assessable penalty that is not the subject of notices of deficiency. As Judge Morrison pointed out in his Rand opinion, the IRS could not currently raise the 6676 penalty as an alternative to the 6662 penalty in the Tax Court deficiency proceeding. Since the penalties alone under 6676 are usually going to be pretty small, this may discourage the IRS from ever asserting them – too much work for too little money. That’s also bad policy. I recommended that penalties that relate to deficiency-procedure adjustment be able to be included in notices of deficiency. I hope Nina Olson’s office pushes next year for some reworking of these penalties to align them more sensibly procedurally.