Last year I blogged about one unhappy feature of the Affordable Care Act (ACA): §36B has no wiggle room in it to deal with reasonable taxpayer errors in calculating their Premium Tax Credit (PTC). As a result, taxpayers sometimes get hit with really big deficiencies, even when the error is someone else’s fault.
Section 36B(f)(2)(B) tries to limit the harshness. For families whose income is less than 400% of the relevant poverty line, that provision puts a cap on how much an error will cost them. But that is no help to families who are just outside of the 400% threshold. Nor does the law permit the Tax Court or the IRS to distinguish between good faith errors and taxpayer negligence. I like how Christine Speidel put it in this post: “The problem for taxpayers hoping to avoid strict reconciliation is that section 36B simply does not have a mechanism to consider equity in the reconciliation of APTC.”
The good folks at Procedurally Taxing have really been on top of this §36B problem. Here’s a list of their blog posts on the subject. In particular, Christine Speidel has a good summary of the law here, and Samantha Galvin illustrates two recent §36B cases here.
Two recent Tax Court decisions reveal a new dimension to §36B harshness: it’s interplay with the lump-sum SSI election in §86(e). In Charles W. Monroe and Rebecca A. Monro v. Commissioner, T.C. Memo. 2019-41 (Apr. 24, 2019) (Judge Ruwe), and in Levon Johnson v. Commissioner, 152 T.C. No. 6 (Mar. 11, 2019) (Judge Gerber), the question was whether an SSI lump sum catch-up payment had to be counted in calculating PTC eligibility when the taxpayers had made an election under §86(e) to reallocate the lump-sum payment from current year into prior years. The Court looked at the text of §36B and said “yep, gotta include the lump sum.”
I doubt the Court enjoyed coming to its conclusion because the harsh result turns the tax law into a “gotcha” game that even reasonable taxpayers will lose. But if you learn the lesson these cases teach, you can at least help your clients who are in similar situations.
There was an alternative interpretation here, however, that practitioners might press upon the Tax Court in future cases. A closer look at the §36B language suggests it may well have room for a different "plain language" interpretation consistent with the §86(e) policy. While the Court's holding was not unreasonable, neither was was it desirable, and it was sure as heck not inevitable. Details below the fold.
The Law: §36B and the ACA
The ACA requires all individuals to purchase health insurance, requires all health insurance plans to contain certain provisions, and subsidizes the purchase of health insurance via a tax credit mechanism. It’s the tax credit I want to focus on. You find it in §36B, titled Premium Assistance Credit. It’s more commonly called the Premium Tax Credit (abbreviated PTC).
The tax credit creates a timing problem. Taxpayers are eligible for the subsidy based on their yearly income. They need the help during the year but won’t know their yearly income when the year ends. To deal with this, Congress permits taxpayers to get the credit paid in advance through what the Advance Premium Tax Credit (APTC). The APTC provisions are not codified but can be found in §1412 of the ACA.
The APTC is based on the taxpayer’s guesstimate of what their income will be. When the year ends, the taxpayer trues-up that guesstimate with the actual income. If the taxpayer overestimated income, the taxpayer may get additional tax credits. If the taxpayer underestimated income, the government recaptures the excess tax credit by increasing the tax liability for the year dollar-for-dollar. While the APTC is paid directly to the insurance provider and not to the individual taxpayers, it still needs to be recaptured if the true-up process shows it was excessive.
This is, at best, an awkward system. The true-up process occurs during each filing season on the Form 8962. The IRS must coordinate here with HHS because HHS regulations determine whether a taxpayer is eligible for the advance payments and IRS regulations govern the true-up process. The Tax Court gives a very nice detailed explanation in McGuire v. Commissioner, 149 T.C. 254 (2017).
Key to the true-up process is figuring out whether something called “household income” ends up having been more than 400% of the federal poverty rate for households of the taxpayer’s size. If a taxpayers household income is less than that, the taxpayer qualifies for some amount of PTC and, importantly, §36B(f)(2)(B) limits the recapture of any excess APTC. But if a taxpayer’s household income is greater than 400% then there is no limit on the recapture amount.
So what is household income? The statute defines it as the sum of the taxpayer’s Modified Adjusted Gross Income MAGI) and the MAGI of certain family members. §36B(d)(2).
So what’s MAGI? That’s the question for today’s lesson. Both the statute and regulations define MAGI as “adjusted gross income (within the meaning of section 62) increased by...Social security benefits (within the meaning of section 86(d))...not included in gross income under section 86 for the taxable year.”
That seems straightforward enough...until you look at §86 and the policies behind it. So let’s go there.
The Law: §86 and Social Security
Social Security is not only an insurance program but also an anti-poverty program. True, SSA benefits are not fully means-tested (yet), but the tax provisions, coupled with how benefit calculations are tilted towards lower-income recipients, represents a modest move in that direction. Starting under President Regan in 1983, Congress has chosen to tax an increasing percentage of a taxpayer’s Social Security payments, depending on how much other income the taxpayer has. As income rises, taxpayers must include a larger and larger percentage of their SSA benefits (up to 85%) as gross income.
That is the point of §86. It gives the rules for how SSA benefits are taxed. Section 86 allows taxpayers to exclude some SSA benefits from income while paying tax on the rest. Call that the yearly benefit exclusion. It works like this: once taxpayer’s total income (including half of Social Security benefits) exceeds a threshold amount (currently $32,000 for a married couple), benefits must be included in gross income, up to a maximum of 85% of the benefit payments. What percentage of benefits must be included in gross income depends on how much other income the taxpayer has.
In other words, §86's yearly benefit exclusion is linked to ability to pay, to the anti-poverty purpose of Social Security. It’s a statutory formula that applied regardless of taxpayer’s desires.
In addition to the yearly benefit exclusion, §86 also contains a special allocation provision. Congress recognized that it may take over a year for SSA to approve applications for benefits, particularly disability benefit applications. Practitioners tell me those routinely take two years or more, what with the various rounds of appeals and the SSA problem with staffing, similar to the IRS problem.
When that happens, the SSA pays out a lump-sum representing the benefits accrued during the application process. That lump-sum helps taxpayers because it is often how they are able to pay the lawyers or other representative! But it hurts taxpayers because it distorts income in the year of the lump-sum. It bunches income to that year, raising their gross income and AGI, thus possibly throwing the taxpayer out of otherwise applicable benefits that are based on the AGI.
Section 86(e) takes care of this distortion problem by allowing taxpayers to elect to have the lump-sum excluded from gross income in the year received. Critically, however, §86(e) is not really an exclusion. It does not allow a taxpayer to avoid paying taxes on the SSA lump-sum benefits. It is instead a re-allocation. Taxpayers who make the §86(e) election must re-allocate benefits to the prior year or years in which they should have been paid. Taxpayers will still take a tax hit to the extent that the added amounts would have been included in gross income in the prior years. The tax hit is taken on the current-year return but represents a tax payment now to reflect the amount the taxpayer would have paid if the benefit had arrived on time. There won’t be any tax hit, however, if half the amount allocated added to the taxpayer’s adjusted gross income for the prior year is less that the base amount for including benefits as gross income. Thus, while it is true that the election affects the taxes due in the current year, that is not because of how §86 operates on the current year; it is because of the §86(e) exclusion calculation for the prior year.
Thus, I think it best to view §86(e) as a timing provision, addressing the question of in what tax year a taxpayer must account for SSA benefits. That is different that the yearly exclusion formula, which address the question of whether the taxpayer must include SSA benefits in gross income in the current year. Section 86(e) recognizes that SSDI benefits accrue during the time period the bureaucracy is processing an ultimately successful application and so allows an allocation of those benefits to the time period where they more properly belong.
The effect of §86(e) is to allow cash method taxpayers to elect an accrual method of accounting for the lump-sum to avoid distorting income for the time period the payments are actually received. This is not as weird as you might think. Individual income items of a cash basis taxpayer can sometimes be accounted for by accrual principles. That’s the entire idea of the constructive receipt doctrine, which forces cash basis taxpayer to take an item into income for a tax year if the items was available to the taxpayer and it was only the taxpayer’s own volition that postponed the actual receipt into the next year. See Treas. Reg. 1.451-2.
It is this reallocation policy that is in tension with the plain language of §36B. Let’s see how it plays out in the two cases.
In 2014, Mr. Johnson received about $4,500 of APTC, because he estimated his income would be less than the 400% threshold for PTC. However, by the end of 2014 Mr. Johnson had received a total of $26,000 in SSA payments, of which $12,000 were a lump-sum distribution attributable to 2013 and the other $14,000 were attributable to 2014. Including the $14,000 as gross income in 2014 left Mr. Johnson eligible for some PTC that year but less than he had actually received. So he had to recapture that excess and add it to his 2014 tax liability. Importantly, however, if he only had to include the $14,000, his liability to repay the excess APTC in 2014 would be capped at $1,250 because of §26B(f)(2)(B). But adding in the $12,000 attributable to 2013 pushed him over the 400% threshold, meaning he had to repay all $4,500 of the APTC he had received. And that yearly salary income of about $24,000 and yearly SSA benefits of about $14,000.
So the question was whether the $12,000 for which he made the election to allocate to 2013 had to be counted as MAGI for Mr. Johnson’s 2014 PTC eligibility.
In 2015, Mr. and Mrs. Monroe received about $8,500 in APTC because they estimated their household income would be less than the 400% threshold. However, by the end of 2015 they had received just over $35,000 in SSA benefits, of which about $24,000 was a lump sum representing amounts attributable to four prior years (yep, THAT approval process took a loooooong time). Of the remaining, $5,300 were benefits for 2015 and $6,000 was a payment for attorneys fees. Mr. and Mrs. Monroe made the §86(e) election to have the lump sum allocated to the prior years: $350 to 2014, $10,250 to 2013, $11,000 to 2012, and 2,600 to 2011.
Like Mr. Johnson---and countless other taxpayers each and every year---if the $24,000 in lump-sum payments had to be counted as MAGI for 2015, it would push the Monroes above the 400% threshold.
One infers from the four years of allocation and the payment of attorneys fees that the Monroes had fought quite the battle for their benefits. While they won against the SSA, they lost to that more fearsome three-letter agency, the IRS, just like Mr. Johnson.
Lesson: Plain Language Trumps Policy
Let’s look at the Tax Court’s rationale, a rationale that was fully reviewed in Johnson and then simply applied in Monroe, as it will likely be simply applied in future cases unless the Court decides to revisit Johnson or some appellate court forces it to do so.
Both Mr. Johnson and the Monroes were represented by counsel. Mr. Johnson was represented by Professor Ted Afield, the Director of the Georgia State University School of Law Low Income Taxpayer Clinic, and the Monroes were represented by EA E. Robert Clifton. In Monroe, Judge Ruwe simply relied on Johnson to hold that the Monroes must include all $24,000 in lump sum payments in MAGI for 2015. So we must examine Johnson to discover the Tax Court’s rationale.
The Johnson opinion focused its attention on the statutory language in §36B that tells taxpayer they must include in MAGA “an amount equal to the portion of the taxpayer’s social security benefits (as defined in section 86(d)) which is not included in gross income under section 86 for the taxable year.” §36B9d)(2)(B)(iii).
The Johnson opinion starts its analysis with the proposition that the “year of receipt, as opposed to the year to which the Social Security benefits are attributable, is the significant definitive factor.” The Court observes that “a cash method individual generally reports income in the year it is received, even if the benefits are attributable to a prior year.” It then cites to the timing statutes and regulations in §451.
From the premise that the lump-sum benefits were actually reportable in the year of receipt, the Tax Court then looks at the phrase “under section 86” and decides that “under” means all the subsections of the section, including the election in subsection (e).
Thus, the basic rationale in Johnson is that the taxpayer had to include the lump-sum payments in calculating MAGI because (1) cash method taxpayers had to include lump-sum SSA payments in gross income “for the taxpayer year” they were actually received, and (2) once those payments were included in gross income they were only then “not included in gross income under section 86” because of the subsection (e) election.
Comment: The Language is Not So Plain
I think the Tax Court may have stopped its plain language analysis too soon.
Words are rarely straight-jackets, admitting of only one interpretation. Words are flexible and often admit of multiple interpretations. For a tedious demonstration of that idea, you are welcome to dive into my article “Dual Construction of RICO: The Road Not Taken in Reves,” 51 Wash. & Lee L. Rev. 61 (Winter 1994). There I showed how even the same word in a statute might take on different meanings in different types of cases. In that article I recover a doctrine of statutory interpretation that I call “Dual Construction.” The doctrine of dual construction applies whenever a statute has dual purposes, one of which requires a broad interpretation and one of which requires a narrow interpretation. The article gives several examples of statutes where courts---both state and federal---apply that approach to statutory construction. In our own Tax Code there is the example of "willfully." Courts construe the term “willfully” very broadly in in §6672 but very narrowly in §7202. “Oh,” you say, “but that is because one statute is civil and the other is criminal.” Well...yes. That. Is. The. Point. Different statutory purposes cause courts to construe even the same word differently. It happens in the civil context as well, as I show in the article.
I think the doctrine can apply here to support a different interpretation of the words in §36B in the context of the fact pattern in Monroe and Johnson. Look at the word “under” more closely. Section 36B says that MAGI includes amounts “not included gross income under section 86.” The Tax Court said: hey, §86(e) is "under" §86. The Court did not explain why. If the word “under” were a straightjacket, the Tax Court would not need to explain because it would not have had any choice in how to interpret the statutory text.
But the word “under” has a bunch of different meanings as you can quickly see in the Oxford online entry. The Tax Court had a choice. It was not limited to just one interpretation of the word. One meaning is “at a lower level than” or “used to express a grouping.” That is the meaning the Tax Court gave the word. Subsection (e) is certainly at a lower level than §86. It is under the umbrella of §86. The Tax Court's reasonable interpretation of the word does not lead to any absurd results, just really sad and harsh results. If the word were a straight-jacket, we would all shake our heads and move on.
But. If a different interpretation of “under” would help effectuate the purpose of §86(e), without undermining the purpose of §36B, then one would think that interpretation is preferable to an interpretation which hurts even wary taxpayers based on events they cannot control (the timing of the SSA approval that results in the lump-sum payment). Wow. That’s a long sentence. Sorry.
And guess what? Another interpretation of “under” is available that does indeed permit a more salubrious interplay of §36B and §86(e). The word “under” also means “governed by,” or “controlled by,” or “pursuant to....” And here we go back again to the two different exclusions contained in §86. The first one---the yearly exclusion---is governed by, controlled by, and pursuant to §86 in the sense that the statutory formula in §86(a)(1) tells us how much of the SSA benefits are included and how much is excluded each year. The statutory formula controls the exclusion. Moreover, regardless of whether you agree or disagree with the policy choices, including those excluded SSA amounts in MAGI is consistent policy in both statutes. Both policies are choices about ability to pay. The amounts are excluded under §86 based on a concern about ability to pay, but they are included in MAGI for the same reason: Congress decided that all SSA benefits should be counted as available to pay premiums regardless of whether the benefits were taxable or not. Personally, I think that is a sucky policy choice, but I’m just a blogger. And, as the Tax Court properly states in Johnson: “we cannot ignore the law to achieve an equitable end.”
But the §85(e) election is different, as I discussed above. It’s a choice about timing and reflects a Congressional concern about the fairness of timing. Once you see subsection (e) as a timing rule, reflecting a Congressional policy about clear reflection of income---you can see how it is more accurate to say that the amounts are “reallocated” rather than “not included.” And the reallocation is governed by, controlled by, and pursuant to the taxpayer’s decision, not pursuant to a statutory command. The allocation of lump-sum benefits to prior years is not controlled by the statute itself but is done pursuant to an election authorized by the statute. The choice is given to the taxpayer but is not controlled or forced or governed by §86.
I think the Tax Court had an interpretive choice available here. It would be well within its powers to interpret §36B(d)(2) as requiring taxpayers to count as MAGI those amounts truly excluded from income pursuant to §86’s command, but not those amounts that are simply reallocated and not excluded pursuant to the election that §86(e) offers but does not command. The Tax Court starts down that road when it notes that cash basis taxpayers generally report as income amounts received during the year. But just as constructive receipt is an exception to that rule for certain income items, so I think §86(e) can be reasonably viewed as an exception to that rule for the SSA lump-sum catch-up payments.
I don’t offer my interpretation as the “only” interpretation. That is part of my point. A statute’s words are not solid welded steel so much as lego blocks thrown together perhaps too hastily by a distracted writer. As with many statutes, the Tax Court has a range of legitimate interpretive choices to make. The interpretation made in Johnson (and Monroe) was plausible, but leads to an unreasonable (not absurd!) result. The Court had available an alternative interpretation, one that was perhaps not as obvious, but was legitimate and one that would lead to a much more reasonable result.
Bryan Camp is the George H. Mahon Professor of Law at Texas Tech University School of Law.