Turn(er)ing Back to Basics With FLPs

With all of the political tumult, the all-time high federal exemption rate, the talk of potential new taxes on the wealthy and all of the navel gazing about what the future might bring, it’s all too easy to lose focus on the issues within the control of the estate planner. Just before Thanksgiving in 2018, those seeking a useful reminder of the importance of basics had something to be thankful for—the third installment in the Turner line of cases.¹ Each of the cases in Turner addressed a distinct issue that can often come up in tax-efficient estate planning: the application of Internal Revenue Code Section 2036 to family limited partnerships (FLPs) in Turner I, the potential risk of the marital deduction mismatch in Turner II and now the application of estate tax apportionment and federal reimbursement clauses in Turner III. Let’s review each of the cases with a special emphasis on Turner III and the need to be proactive about estate tax apportionment even with the high exemption amounts that are currently in effect. Importantly, as the landscape for FLPs continues to change and evolve with respect to the potential application of IRC Section 2036, this case serves as a reminder to estate planners to review the tax apportionment and reimbursement provisions contained in the boilerplate when a client has an FLP, as a few words can make a huge difference in the tax consequences on an estate if Section 2036 applies and a marital deduction is anticipated.

FLPs and Section 2036

FLPs are commonplace in affluent families’ overall structures. While they frequently receive great scrutiny and skepticism by the Internal Revenue Service, these entities are often necessary in structures for non-tax reasons, such as proper governance and succession planning. Whenever the ownership of the FLP becomes divided between the generation that established the entity and younger family members, there’s a risk that the IRS will assert that underlying assets attributable to the partnership interests transferred to other family members get pulled back into the senior generation’s gross estate by virtue of Section 2036. This IRC provision can cause estate tax inclusion of an entity if the transferor retained the:
(1) possession or enjoyment of the property (in the case of an FLP, the assets contributed by the decedent into the FLP) or its income (Section 2036(a)(1)), (2) right to designate the persons who’ll receive the benefit of the property (Section 2036(a)(2)), and/or (3) right to vote shares in a controlled corporation (Section 2036(b)). Most of the IRS scrutiny has been on the two inclusion provisions of Section 2036(a). The IRS and the estate argue over whether the facts and circumstances in a case prove that the decedent retained a beneficial interest in or control over the contributed property. Turner I served as a reminder to planners of the factors that will be looked at in determining whether the decedent retained a beneficial interest in the transferred interest of the FLP.

In this case, the FLP in question was created less than two years before the decedent’s passing. Clyde W. Turner, Sr. and his wife, Jewell, contributed approximately $8.67 million of various liquid assets, such as cash, stocks and certificates of deposits, into the FLP. In return, each took back a 0.5 percent general partnership (GP) interest and a 49.5 percent limited partnership (LP) interest. Inthe month the transfer of assets to the FLP was complete and only 13 months before Clyde’s death, Clyde and Jewell each gifted approximately a 21.7 percent LP interest to their children and grandchildren. The values of the LP interest transferred were reported on gift tax returns, which applied various valuation discounts as determined by professional appraisers that were in excess of 40 percent when compared to the underlying value of the assets held by the FLP. Similar discounts were applied to the FLP interests Clyde retained when they were reported on his estate tax return. 

The IRS claimed that the value of the underlying assets of the LP interests transferred were includible in the decedent’s estate under Section 2036(a)(1) as well as 2036(a)(2). The Tax Court agreed with the IRS position, focusing primarily on the Section 2036(a)(1) argument. In finding that Clyde retained “the possession or enjoyment of, or the right to the income from, the property…”² the court held that a variety of factors indicated that he retained a benefit such as the decedent: (1) receiving a partnership management fee to do exactly what he did prior to the formation of the partnership, (2) receiving disproportionate distributions from the partnership, (3) using personal funds to invest on behalf of the partnership without a repayment plan, (4) commingling personal and partnership assets,
(5) using partnership assets to pay the legal fees for his own estate planning, and (6) using partnership assets to pay insurance premiums on policies in his insurance trust. As previously mentioned, whether the decedent retained a benefit is a facts and circumstances test. The fact that Section 2036(a)(1) applied came as no surprise in this case given that many of the above factors clearly demonstrated the existence of a retained interest and/or prior cases held that such factors were indicative of a retained interest.³Turner I also held that Section 2036(a)(2) applied given the powers retained by Clyde as one of the general partners.⁴

The decision also addressed whether an exception applied to the transfers made by Clyde and Jewell. Section 2036(a) doesn’t apply to transfers that were “a bona fide sale for an adequate and full consideration in money or money’s worth…” The court applied the traditional two-prong test used to analyze whether the exception applies: (1) Did the transferor receive full and adequate consideration, and (2) Was it a bona fide sale?⁵ For the exception to apply, both prongs must be met. The parties stipulated that there was full and adequate consideration. The parties disagreed on whether it was a bona fide sale, and the court sided with the IRS that it wasn’t. The determination of whether a transfer is a bona fide sale is a question of motive—specifically, was it an arm’s length transaction in which there were significant non-tax reasons for forming the FLP? Factors from both the formation and operation of the FLP led the court to find there were no significant non-tax reasons for the transaction, and it wasn’t at arm’s length. The court noted that the assets contributed didn’t suggest there was a non-tax reason, as they were passive investments as opposed to assets that require active management such as a closely held business, an operating company or real estate activity. Further, there were no changes as to how the assets were managed after the formation of the partnership. Finally, Clyde was on both sides of the transaction, as the donees had no meaningful participation in the transfers, which suggested this wasn’t at arm’s length.  

While the decision in Turner I properly applied a straightforward application of Section 2036 to an FLP, the first installment of this trilogy reminds planners of the fundamentals of Section 2036(a)(1). Both factors at the formation and during the existence of the FLP will be used in determining whether an interest was retained and if the bona fide sale exception applies. At the formation, planners must establish significant non-tax purposes for the formation of the FLP, as well as ensure that assets are properly titled (that is, not commingled), and governing documents don’t grant certain powers to the founder of the FLP. More importantly, the case is a reminder that unlike the Ronco Oven—FLPs aren’t “set it and forget it” propositions.⁶ 

Turner II—Two Layers of Mismatch

In Turner II, the Tax Court was asked to address the impact of the estate tax marital deduction on the inclusion of the assets. Planners learned that Section 2036 inclusion with an FLP doesn’t just strip away any benefit of the FLP planning, but also can cause more harm than doing nothing. In Turner II, the estate asserted that the decedent’s estate plan provided a formula clause that Jewell was to receive sufficient assets to reduce the estate tax liability to zero; therefore, the marital deduction should be increased to reduce the estate tax liability to zero. Such provisions are commonplace in estate plans. The Tax Court rightly rejected this argument as it was conflating two different concepts. The decedent’s estate plan required that sufficient probate assets be transferred to increase the marital deduction to make the taxable estate non-taxable (taking into account credits available). The decedent’s estate plan didn’t and couldn’t make the marital deduction itself equal the correct number to ensure the estate had no tax liability. Under Section IRC 2056, the marital deduction only applies to assets that “pass” to the spouse.⁷ Tax-sensitive estate plans that rely on formula provisions seek to pass the correct amount of assets to achieve the desired deduction and tax result—importantly, however, they can’t create a phantom deduction for assets that don’t pass to the spouse. This result is necessary for policy reasons because there’s no provision in the IRC that would subject assets that don’t pass to the spouse to transfer taxes on her death. 

In addressing the issue of the marital deduction, the court addressed the “marital deduction mismatch” issue. In our view, there are two different layers of the marital deduction mismatch. It’s quite possible either or both layers can apply in the context of when Section 2036 applies to an FLP, and they can cause unexpected and potentially draconian results. The first layer is that the inclusion could create a higher estate tax liability than had other or no planning been done because of the inclusion of the phantom assets. The second layer is a marital deduction that doesn’t even cover the value of the FLP that passes to the spouse because of differing values for the inclusion of assets and the value of the assets for which a marital deduction is allowed.  

The first layer of mismatch can occur when the decedent had a plan with standard language to fund a marital bequest (outright or qualified terminable interest property) to lower the taxable estate to the available credit amount, but FLP interests are given during life, and later, the contributed assets are included in the gross estate by virtue of Section 2036. This is exactly what was presented in Turner. The potential result can be that a first death estate tax is owed. If Section 2036(a) applies to pull contributed FLP assets back into the decedent’s gross estate at full, undiscounted value, that’s a phantom asset that will reduce the exemption amount. Meanwhile, the estate can only obtain a marital deduction for those assets that pass to or for the surviving spouse. If the donee of the Section 2036 assets isn’t the spouse or a trust that would qualify for the marital deduction, then no marital deduction is available for the Section 2036 assets. If there isn’t sufficient credit to cover the value of all of the phantom assets included in the gross estate, the tax clause in the estate plan won’t be able to achieve its purposes of preventing a first spouse death estate tax. 

The second layer is a question as to whether the estate is entitled to a marital deduction for FLP interests that do pass to a spouse that’s equal to the full undiscounted value of the underlying FLP assets included in the gross estate or whether the marital deduction will be limited to the value of the LP interest that the decedent owned, which would be lower due to the application of valuation discounts. The issues arise because the estate will only be entitled to a deduction for the interest that will “pass” to the spouse. If the correct inclusion value is the higher value (undiscounted underlying FLP assets) but the marital deduction is for the value of the entity interest that will be lower, then there will be a mismatch between the value of the inclusion of the asset and the marital deduction allowed. 

The IRS has argued that the marital deduction doesn’t apply to this difference because the spouse doesn’t actually receive the phantom inclusion value under Section 2036, but, rather, receives the LP interest, which may have a lower fair market value due to valuation discounts. That is, the IRS has attempted to affirmatively use the application of valuation discounts to the disadvantage of the decedent’s estate by arguing that the marital deduction should be less due to valuation discounts. Turner wasn’t the first case to raise this issue. The IRS raised this “marital deduction mismatch” issue in the past in the context of FLPs in the Black and Shurtz cases.⁸ Luckily for the estates, in both cases, the marital deduction mismatch issue was ultimately determined to be largely academic, due to the Tax Court’s determination that the bona fide sale exception to Section 2036 was satisfied. While that finding saved the day for those two estates, the risk posed by this argument is one of which planners should take note for an estate that’s not able to prevail on a Section 2036(a)(1) challenge.⁹ Ultimately, in Turner II, even though the bona fide sale exception didn’t apply, the Tax Court indicated that because the IRS didn’t raise this layer of the mismatch argument, the court would “leave this mismatch problem for another day.”¹⁰ That day never came in the Turner case as this issue also became moot given the holding in Turner III that a party other than the spouse was ultimately responsible for the estate tax liability. 

Unfortunately, these mismatch issues are often ignored or not appreciated by estate-planning practitioners when creating vehicles such as FLPs, but the risk is one that has some sharp tax teeth. Section 2036 is an artificial or “phantom” inclusion statute that includes assets in a decedent’s estate for estate tax purposes; however, this phantom inclusion can and often does occur in the context of an FLP when the LP interests have already been given away for property law purposes, and are, therefore, no longer included in the decedent’s probate estate. Because qualification for the estate tax marital deduction depends on an asset actually “passing” to the surviving spouse or in a marital trust for his lifetime benefit, such a disconnect or mismatch between phantom estate inclusion versus exclusion of previously transferred assets from the probate estate could result in estate taxation of assets with no ability to obtain an offsetting estate tax marital deduction.¹¹ Unless the Section 2036 potential exposure, availability of the marital deduction and estate tax apportionment and right of reimbursement are implemented, monitored and maintained properly, the tax result could be far worse for having set up the FLP than if the taxpayer had just held the assets outright at death with an estate plan that deferred estate taxes until the second spouse’s death. 

Turning to Tax Apportionment 

The focus of Turner III was the computation of the estate tax liability. The issue in Turner III was how to compute the estate tax liability given the holdings in the prior cases. There were two issues the parties disagreed on when computing the liability. One involved the estate’s argument that the estate tax marital deduction should be increased by the amount of post-death income that was distributed to the spouse. Given the clear statutory language of Section 2056(a), the Tax Court found that as the value of the income wasn’t included as part of the gross estate, it couldn’t be taken as a deduction because the deduction is only permitted “to the extent that such interest is included in determining the value of the gross estate.”¹² 

The other and more interesting issue is an important one that has a bearing on all taxable estates—the impact of estate tax apportionment and federal reimbursement rights. In calculating the estate tax liability, the IRS argued that the marital deduction should be reduced by the amount of estate taxes paid because the only property the estate had readily available to pay the liability was property that would pass to the surviving spouse and otherwise would qualify for the marital deduction. If the amount of assets passing to the spouse was the source of estate tax payments, the IRS is correct that this would indeed reduce the marital deduction and create an even greater estate tax liability given the circular “tax on tax” computation that would apply.¹³ Given the inclusive nature of the estate tax, whereby estate taxes are imposed on the value of all property included in the gross estate including the dollars used to pay the estate tax liability, the consequence of the lost deduction is that the estate tax will increase exponentially given a circular computation. For every dollar of deduction lost to pay estate taxes, additional estate taxes are owed on the amount used to pay estate taxes, which requires additional funds that would have otherwise qualified for the marital deduction, which then reduces the marital deduction, which results in additional estate taxes being owed, and so the circular estate tax on tax wheel spins.

The estate successfully countered the IRS’ argument by asserting that the estate was entitled to reimbursement under IRC Section 2207B from the recipients of the property included by virtue of Section 2036. As the estate would be reimbursed for the estate taxes it paid, the value of the marital deduction shouldn’t be reduced because ultimately those funds would pass to the spouse. 

In reviewing this issue, the Tax Court methodically worked through the tax apportionment issue. In general, both state and federal law provide that decedents may determine the apportionment of estate taxes in their estate-planning documents. In the absence of directions by the testator, state law will provide default provisions for the apportionment of estate taxes. In addition to those provisions, there are also federal provisions that could effectively act as apportionment provisions for certain non-probate transfers. The court found that while “[o]ften the decedent’s will provides guidance to the executor regarding how the executor should allocate the tax burden,” the parties agreed that Clyde’s will had “no express provision” regarding the apportionment of estate taxes. As Clyde died domiciled in Georgia, the apportionment rules of that state applied. Under Georgia law, estate taxes are to be apportioned to the residue of the probate estate in the absence of a contrary direction in the estate-planning documents.¹⁴ Thus, an application of the default provisions of state law would have resulted in the argument that the IRS asserted in that assets in the probate estate that would otherwise pass to the surviving spouse would instead need to be diverted to pay estate taxes, which would further increase the tax liability because those assets themselves would no longer pass to the surviving spouse and thus wouldn’t qualify for the marital deduction. 

However, the court further noted that state law isn’t the only provision that has a bearing on estate taxes. Federal law provides four scenarios in which the estate is entitled to reimbursement for federal estate taxes paid on certain non-probate assets that are included in the gross estate. These provisions are: IRC Section 2206 (for life insurance proceeds); IRC Section 2207 (for property included because of a general power of appointment); IRC Section 2207A (for property included in the gross estate because of IRC Section 2044); and Section 2207B (for property included in the gross estate because of Section 2036). The specifics of the operation of each of these provisions are slightly different, and an in-depth dive into each of them is outside the scope of this article. The important
commonality of these statutes is that, in the absence of a specific waiver of such right(s), they provide an estate the right to be reimbursed for federal estate taxes paid (based on a computation specific to each statute). These provisions apply in limited circumstances as they don’t apply to probate property and other non-probate transfers. Further, they don’t provide a right to reimbursement for state estate taxes, only federal taxes. Although these four federal statutes may conflict with the terms of the state apportionment statutes, the U.S. Supreme Court has held that they’re valid and supplemental to state provisions and, in this case, Georgia’s apportionment statute specifically states that federal reimbursement rights aren’t impacted by state law.¹⁵

In Turner III, the estate contended that Section 2207B applied, and therefore, the estate had a right to reimbursement for estate taxes paid on the property included in the gross estate under Section 2036 as a result of the holding in Turner I. The resolution on this issue depended on the outcome of three questions: (1) Did Section 2207B apply to this transfer? (2) Did the executor have a duty to enforce the right? and (3) Does the reimbursement amount allow for the marital deduction to be preserved? With respect to the first question, Section 2207B provides a right to reimbursement when property is included in the gross estate by virtue of Section 2036 unless the decedent waived the right of recovery in the manner prescribed by the federal statute. The Tax Court’s holding in Turner I
clearly held that the underlying assets of the FLP interests that were transferred by the decedent were included in his gross estate because of Section 2036. Further, Section 2207B(a)(2) provides that the right of the estate to be reimbursed may be waived only if “the decedent in his will (or a revocable trust) specifically indicates an intent to waive any right” to be reimbursed. The current version of Section 2207B doesn’t require a reference to the IRC section itself in the decedent’s estate plan, though that would be the clearest way to draft such a waiver; but a general provision that an estate pay all expenses from the residue wouldn’t be specific enough to waive this provision.¹⁶ Given that the parties agreed the estate plan was silent on the apportionment issue, there was no specific waiver of the provision, and therefore, the Tax Court determined that the reimbursement right provided under Section 2207B(a)(2) applied.

The next question is whether the estate would enforce the reimbursement right. While federal law clearly provided a right to reimbursement, what actions a fiduciary must take to exercise such right and what duties apply are a questions of state law. The decision in Turner III didn’t analyze Georgia law or discuss fiduciary duties. Instead, the court focused on the decedent’s intent and how the executor should act in light of that intent. The court found that “[i]t is reasonable to assume that [the decedent] would want his executor to take all steps necessary to ensure that the property passing to his surviving spouse and qualifying for the marital deduction not be impaired.” Given this intent, the court found “that the executor must exercise the right of recovery under [Section] 2207B” (emphasis added). For reasons discussed later in this article, the fact the court had to address these two questions should be both concerning and instructive to estate planners. Proactively addressing the apportionment issues in the estate-planning documents would have kept the court from having to go through the analysis of what the decedent intended. The court’s application of Section 2207B may have been an attempt to arrive at an answer that felt right and didn’t have a draconian result by arriving at this conclusion based on a determination of what the decedent would have wanted. However, one shouldn’t presume that this will always be the outcome when the application of the statute isn’t entirely clear on its face. The important takeaway for the planner on this issue is to be proactive in deciding whether to waive the right to reimbursement, as the implications can be dramatically different.

The final issue is whether the right to reimbursement would preserve the estate tax marital deduction. The IRS argued that because the executor needed to pay the estate taxes from property that would have otherwise passed to the surviving spouse and then seek reimbursement, the marital deduction shouldn’t be available for the property that was used to pay the estate tax by the executor. The court agreed with the IRS in the operation of how the reimbursement right works, but it also disagreed that this should impact the marital deduction. Under IRC Section 2002, it’s the executor who’s responsible for paying the estate tax. However, as the court noted, “an executor must find a way to pay the tax liabilities created by section 2036 assets while dealing with the reality created by section 2036: The actual section 2036 assets are not in the possession or control of the executor and cannot be used by the executor to pay the tax liabilities they create.” Section 2207B provides a right of reimbursement, meaning that by the statute’s own terms, the right of the estate to get a contribution from Section 2036 recipients is only for tax that “has been paid…” The court found that as the reimbursement right existed and had to be enforced by the executor for the reasons discussed previously, the surviving spouse’s share would be replenished by the reimbursement amount. The court determined that the amount of the reimbursement under Section 2207B would fully cover the liability created by the inclusion of the value of the underlying assets of the FLP. Given the dollar-for-dollar replenishment, it wouldn’t be appropriate to reduce the marital deduction by the funds advanced by the estate to pay the estate taxes. This part of the decision is welcome news for estate planners as its analysis could apply not only in this scenario but also to all of the reimbursement provisions for both the marital and presumably charitable estate tax deduction.¹⁷ 

Apportionment Provisions

While the subject of transfer tax apportionment may not be considered “sexy” (by estate-planning standards, that is), these provisions shouldn’t be left to chance in the boilerplate as the results may not be benign; indeed, the unanticipated application of these provisions can have potentially draconian effects. There are four overarching problems that may occur if special care isn’t taken with apportionment provisions:

1. The decedent’s dispositive goals aren’t carried out because of the apportionment/reimbursement provisions inadvertently shifting the beneficial interests;

2. An increased estate tax, and a resulting circular tax on tax, due to the reduction of the marital and/or charitable deduction as taxes are apportioned against property that was intended for a charity or surviving spouse;

3. Differing directions applying for federal and state transfer taxes; and

4. The potential for a deemed gift.

To explain each of these potential pitfalls in more detail, let’s start with the one that’s the easiest to conceptualize. It’s self-evident that if the amount a decedent wanted to pass to a beneficiary is reduced, then the intended beneficiary won’t receive the full amount, and the decedent’s wishes aren’t carried out by the drafting. For example, say an individual wishes to leave $1 million to his brother. A bequest simply stating: “I leave $1 million to be distributed outright to my brother,” or a beneficiary designation on a life insurance policy includible in the gross estate, or any other non-probate machination may (emphasis on the word “may”) get the job done. However, if estate taxes are owed by the estate, it’s possible that the $1 million will be reduced by a reduction of estate taxes, and given the wishes of the decedent in this example, a reduction of even $1, let alone $100 or $100,000 or more, means that the drafting didn’t achieve the desired result. As such, when a client has specific bequests or non-probate transfers, planners must always give special care to the apportionment and reimbursement language instead of relying on boilerplate apportionment provisions.

The second danger of not carefully crafting the apportionment and reimbursement language is the inadvertent reduction of the marital or charitable deduction. The dispositive provisions of estate-planning documents often use formula provisions in transferring property to a surviving spouse or beneficiaries that would qualify for the estate tax marital or charitable deduction. Inadvertent apportionment could prevent the intended benefits of the formula allocations to be reduced. The IRS’ desired outcome in Turner III is an illustration of that result. As discussed above, if the transfer taxes attributable to the property included by virtue of Section 2036 were payable out of property that would otherwise have passed to the surviving spouse in a manner that would otherwise qualify for the estate tax marital deduction, then the property used to pay the estate taxes would no longer qualify for the marital deduction. This would lead to the circular estate tax on tax computation that would substantially increase the amount of estate taxes owed. This isn’t at all an academic issue, especially when put into context with the understanding of clients who’ll likely recall the discussion when setting up an estate plan when the planner explains the formula clause as resulting in no estate taxes until both spouses die.

Careful drafting is essential to avoid this harsh result. Boilerplate savings language may not be enough to save the marital or charitable deduction from state estate tax apportionment provisions. Such savings clauses often provide broad language that states in general terms that, notwithstanding any provisions to the contrary, certain transfers were intended to qualify for a marital and/or charitable deduction, and provisions that would disqualify such a deduction aren’t to apply to save the deduction. However, whether such savings language is effective is a question of intent. Jurisprudence and administrative guidance demonstrate that while these provisions can save deductions, there have also been times when they haven’t.¹⁸ In determining the intent of the decedent, the estate is left having to prove the decedent’s intent to taxing authorities and possibly courts when some concise and well drafted apportionment language could have avoided such an emotionally charged and costly ordeal. Further, and importantly, boilerplate savings clauses don’t address the federal reimbursement provisions. As discussed, Sections 2206, 2207, 2207A and 2207B require a reference (in varying levels of specificity) to the provision for it to be effectively waived and not apply. As demonstrated in Turner III, a standard marital bequest funding clause wouldn’t meet the level of specificity required for any of those provisions.

The third complication that can arise is the differences between the state and federal estate tax apportionment and reimbursement statutes. In general, state law dictates the default provisions and what estate-planning documents must say to override the default provisions for all state estate taxes and all federal estate taxes except those to which the federal reimbursement provisions apply. Given this rule, in most cases, the apportionment of federal and state estate taxes will be applied in a similar manner. However, for property subject to the federal reimbursement provisions, there’s the potential that the state and federal taxes are apportioned to different sources of payment. For example, consider a case like Turner, in which the underlying assets of an FLP were included in the decedent’s gross estate by virtue of Section 2036. In such a scenario, the decedent died domiciled in a state that imposed a state estate tax. The decedent’s testamentary instrument had an apportionment clause that stated: “all taxes arising as a result of my death, whether attributable to assets passing under this will or otherwise, be paid out of the residue of my probate estate.”¹⁹ If, however, the jurisdiction where the decedent died had adopted the estate tax apportionment provisions of the Uniform Probate Code or the Uniform Estate Tax Apportionment Act, the comments to those uniform acts make clear that the probate estate would indeed be responsible for the estate taxes, and it wouldn’t be apportioned against the recipient of the Section 2036 property. However, because the example apportionment language doesn’t waive the federal reimbursement right of Section 2207B, a share of the federal taxes would be paid by the recipient of the
Section 2036 property while the state estate taxes on that very same property would be paid by the probate estate; thus, creating a further mismatch.

Even though this bizarre and likely unintended result would only occur when state estate taxes would be imposed, this discrepancy between what’s required to waive the default provisions under state and federal law applies to all estates that owe a federal estate tax. Using the example from the above, with the modification that the decedent isn’t subject to state estate taxes, the explicit apportionment provision in the will is rendered ineffective despite the clear intent because of the specific waiver requirement of Section 2207B. The drafting to avoid the result in this specific instance wouldn’t have been unduly burdensome. If the testator wished to apportion the transfer taxes to the probate estate, the document simply could have stated, “all taxes arising as a result of my death, whether attributable to assets passing under this will or otherwise, be paid out of the residue of my probate estate. I hereby waive the right to reimbursement under Internal Revenue Code
Section 2207B.” Thus, addressing apportionment issues doesn’t call for lengthy or unwieldy drafting, but does require the estate planner to consider all scenarios and capture the client’s wishes taking into account the requirements of federal and state law. 

The fourth problem that can occur if the federal reimbursement provisions aren’t proactively planned for and properly applied is the risk of a deemed gift. Even if the beneficiaries to an estate agree or the fiduciary decides that the apportionment of estate taxes should be implemented in a manner other than what’s required by the terms of the documents to capture the decedent’s intent—such a decision doesn’t come without fiduciary duty issues or tax consequences. The fiduciary duties of an executor are defined by state law—not federal. As a result, there’s a lack of uniformity and some uncertainty among the various jurisdictions as to what an executor should do regarding a reimbursement right. As previously discussed, the Tax Court in Turner III found that a duty did exist to enforce it without reference to Georgia’s law, but reached such a conclusion based on the intent of the decedent, which reflects that the court was attempting to reach a “fair result.” A look across jurisdictions would show various approaches on whether there’s a duty to enforce the federal reimbursement provisions. For example, while some cases have found there’s a duty to enforce a reimbursement right,²⁰ some cases have held that duty only comes into play if the estate can’t pay the estate’s creditors,²¹ while many jurisdictions have no clear precedent as to what duty exists. Given that one party or another will be impacted by the exercise or non-exercise of a right to reimbursement, it puts the fiduciary in an uncomfortable position that will likely incur the expense of court proceedings to confirm the proper approach to protect the fiduciary from liability. Thus, much uncertainty and expense can be saved with clear apportionment language and direction as to reimbursement rights. 

In addition to the fiduciary duty issues (in fact, compounding them), there are potential transfer tax consequences if a reimbursement right isn’t exercised. If a right to reimbursement isn’t exercised when it existed, the economic result is that the party who should have paid a reimbursement for estate taxes should have received less, and the party who paid the estate taxes and didn’t receive the reimbursement it was entitled to should have received more. The question arises as to whether this shift in the value of property received is considered a gift for transfer tax purposes. Out of the four federal reimbursement statutes, only Section 2207A has explicit guidance on this point. The accompanying Treasury regulations for Section 2207A state that a failure to exercise the right of recovery is considered a gift from the party disadvantaged by the failure to exercise the right of reimbursement to the party that benefited from the non-exercise.²² 

Although Sections 2206, 2207 and 2207B are silent on whether there would be a deemed gift between parties, applying the general gift tax principles of Section 2511 the way they were in
Section 2207A would lead to a similar rule. The position taken in Section 2207A does in many ways appear to be the natural interpretation of Section 2511. Consistent with the Treasury regulation, the IRS has interpreted the right to reimbursement as an interest in property.²³ Donative intent isn’t a requirement for there to be a taxable gift, so it’s not relevant whether the party disadvantaged by the failure to seek the reimbursement right ended up in this result knowingly.²⁴ Property also needn’t be transferred for there to be a taxable gift under Section 2511—the forgiveness of an amount owed is a gift.²⁵ Finally, even though the reimbursement right doesn’t run directly between the party advantaged and disadvantaged (as the estate would be in the middle), indirect gifts are nonetheless considered gifts for federal tax purposes.²⁶ There’s one issue, though, that makes the deemed gift result seem unjust. Under the federal reimbursement provisions, it’s the “executor” or the “decedent’s estate” that has the right to recovery—not the party disadvantaged.²⁷ Effectively, although the executor is the party who can enforce the right, the executor’s failure to do so could lead to a deemed gift from the party disadvantaged by the estate not recouping the reimbursement amount—a true insult to injury. This puts a beneficiary of a decedent’s estate in a position in which the beneficiary must be aware of these reimbursement rules and ensure the fiduciary enforces them. In addition, the issue can be further complicated in the not uncommon situation in which the executor may also be a beneficiary of the estate, perhaps along with other non-fiduciary beneficiaries.

This deemed gift risk again demonstrates the potential harm that can occur without thoughtful, client-specific tax apportionment advice. It should be noted that, arguably, the same deemed gift risk could apply if the failure to ensure apportionment provisions under state law or the governing document are properly enforced. However, given the differences among various jurisdictions and the fact that the issue is sufficiently illustrated through the federal reimbursement provisions discussed above, nothing more will be said here. In a situation in which the parties can agree on the proper apportionment
despite unclear or inaccurate drafting, it’s possible to resolve the issue in a transfer tax-neutral manner. As the reimbursement right is an interest in property, it may be disclaimed by using a qualified disclaimer under IRC Section 2518.²⁸ However, the usual impediments of qualified disclaimers apply. First, the disclaimer must be completed within nine months of the creation of the reimbursement right, which in the case of federal reimbursement rights is when the estate makes the estate tax payment. Second, the party disclaiming can’t direct where the property disclaimed passes, so the other terms and governing law of the decedent’s estate must result in the amount passing to the desired party. Qualified disclaimers can also be used to address state estate tax apportionment issues, subject to the same limitations. However, there’s also slightly more flexibility when it comes to estate tax apportionment issues when there’s uncertainty and disagreement as to the meaning and application of the provisions in estate-planning documents. Assuming there’s a bona fide dispute among the parties, a settlement or mutual distribution agreement shouldn’t trigger gift taxes.²⁹ While some comfort may be found in palliative and transfer tax-free solutions, the point of this whole discussion is to show how efficient and effective it is to clearly address the apportionment in the estate-planning documents.

Turn(er)ing the Page 

In the ongoing developments with respect to FLPs over the past years, Section 2036 continues to be an issue of concern for families that have these entities in place, and the apparent expansion of its application is one that practitioners need to be aware. Turner III and its predecessor cases serve as an important reminder that the failure of an FLP isn’t a “nothing to lose” situation but rather one that has the potential for a severe result with the loss of the estate tax marital deduction and, potentially, a first death estate tax and circular tax on tax. In Turner III, the court concluded that at least the estate tax on tax wouldn’t result, but that was due to a taxpayer friendly interpretation of a right to reimbursement under Section 2207B, based on the intention of the decedent. This case should serve as a very important prompt to planners who have clients with FLPs to carefully examine their reimbursement provisions and determine whether a waiver of such rights should be in place, or importantly, not be in place, as the different consequences can be dramatic. This is particularly important for those families who have an increased risk of Section 2036 exposure due to the recent developments in the case law.

—The views expressed in this article are those of the authors and do not necessarily reflect the views of Ernst & Young LLP or Withers Bergman LLP. 

Endnotes

1. Estate of Turner v. Commissioner, 151 T.C. 10 (2018) (Turner III); Estate of Turner v. Comm’r, 138 T.C. 306 (2012) (Turner II); Estate of Turner v. Comm’r, T.C. Memo. 2011-209 (Turner I).

2. Internal Revenue Code Section 2036(a)(1). 

3. See, e.g., Estate of Schauerhamer v. Comm’r, T.C. Memo. 1997-242; Estate of Reichardt v. Comm’r, 114 T.C. 144 (2000); Estate of Harper v. Comm’r, T.C. Memo. 2002-121. 

4. In more recent years, the Internal Revenue Service has been asserting and succeeding in IRC Section 2036(a)(2) arguments. A detailed discussion of Section 2036(a)(2) inclusion is outside the scope of this article. For a discussion of the application of Section 2036(a)(2) to family limited partnerships (FLPs), see N. Todd Angkatavanich, James I. Dougherty and Eric Fischer, “Estate of Powell: Stranger Than Strangi and Partially Fiction,” Trusts & Estates (September 2017); N. Todd Angkatavanich and Eric Fischer, “Family Co-Investments in the Wake of Powell and Cahill: Time to Kick the Tires on Old Vehicles,” Tax Management Estates, Gifts and Trusts Journal (January 2019). 

5. Some cases have approached the two prongs as distinct tests, while others have applied them as interrelated tests. See Estate of Bigelow v. Comm’r, 503 F.3d 955 (9th Cir. 2007) (holding that the tests are interrelated); Estate of Strangi v. Comm’r, T.C. Memo. 2003-145, aff’d 417 F.3d 468 (5th Cir. 2005) (holding that the two prongs were distinct tests). 

6. N. Todd Angkatavanich and Stephanie Loomis-Price, “Set it, But Don’t Forget It,” Trusts & Estates (September 2011).  

7. For purposes of this article, the term “pass” refers to transfers that fall under the definition of IRC Section 2056(c). 

8. Estate of Black v. Comm’r, 133 T.C. 340 (2009); Shurtz v. Comm’r, T.C. Memo. 2010-21.

9. See generally N. Todd Angkatavanich, “Black Shirts (Black, Shurtz) and the Marital Deduction Mismatch,” Trusts & Estates (June 2010).

10. For a further discussion of the marital deduction mismatch issue, see Stephanie Loomis-Price and N. Todd Angkatavanich, “Turn(er)ing the Tables on Taxpayers,” Trusts & Estates (July 2012). 

11. While this mismatch issue is relatively new in the context of FLPs, the concept isn’t a new one. In Private Letter Ruling 9050004 (Aug. 31, 1990), the decedent owned 100 percent of a corporation of which 49 percent passed on death to a marital trust. The IRS determined that the estate was entitled to a reduced deduction based on a discounted minority interest, despite the fact that the 100 percent stock was valued without application of a discount for gross estate purposes. In PLR 9403005 (Oct. 14, 1993), the decedent’s gross estate included a controlling block consisting of common and preferred shares. The IRS ruled that the decedent’s estate was only eligible for the marital deduction based on the discounted value of preferred shares, which constituted a minority interest. While the gross estate was valued to reflect the decedent’s controlling block consisting of both preferred and common shares, such control didn’t exist with respect to the stock passing to the marital deduction trust so that only a lesser marital deduction was available. Similarly, in Disanto v. Comm’r, T.C. Memo. 1999-421, the Tax Court held that a valuation discount applied for marital deduction purposes to minority interests passing to the surviving spouse as a consequence of her disclaiming some portion of the controlling block of shares owned by the decedent. 

12. See Section 2056(a). For those interested in what the estate was trying to argue despite the clear statutory language, the estate was claiming that post death income would be included in the marital share under Treasury Regulation Section 20.2056(b)-4(d)(1)(iii). This regulation, which the court accurately describes as “confusing,” only applies for purposes of computing the impact of estate administration expenses on the marital deduction.

13. Section 2056(b)(4).

14. Georgia Code. Ann. Section 53-4-63(a).

15. SeeRiggs v. Del Drago, 317 U.S. 95 (1942); Georgia Code. Ann. Section 53-4-63(e).

16. See HR Rep. No. 148, 105th Cong., 1st Sess. 1, 614 (1997). By borrowing the IRS’ interpretation of what would be sufficient under IRC Section 2207A, which has similar language as to what’s required for a waiver, reference to IRC Section 2207B, Section 2036 or using terms from Section 2036 should be specific enough. See PLR 200452010 (Sept. 13, 2004) for a discussion of what’s sufficient for Section 2207A.   

17. One item in which there could be a difference between the IRC sections is the interest and penalties owed on the estate tax liability. Section 2007B(c) provides that the right of reimbursement includes not only the estate tax liability, but also interest and penalties. Section 2207A(d) provides the same right of reimbursement against qualified terminable interest property assets. However,
Sections 2206 and 2207 have a more limited reimbursement right that only covers the tax liability—not the interest and penalties. 

18. See Revenue Ruling 75-440; Technical Advice Memorandum 199932001 (April 29, 1999) and Estate of Todd v. Comm., 57 T.C. 288 (1971) in which savings clauses were upheld. But compare Estate of Walsh v. Comm’r, 110 T.C. 393 (1998), in which a savings clause didn’t work.

19. Clause is directly from comments of the Uniform Probate Code (1969 as amended in 2010) (comment to Section 3-9A-103) and the Uniform Estate Tax Apportionment Act (2003) (comment to Section 3). 

20. See Pearcy v. Citizens Bank & Trust Co., 121 Ind. App. 136 (1951).

21. Jeromer v. United States, 155 F. Supp. 851, 854 (S.D.N.Y. 1957).

22. Treas. Regs. Section 20.2207A-1(a)(2). 

23. PLR 200127007 (March 30, 2001).

24. Treas. Regs. Section 25.2511-1(g)(1).

25. Treas. Regs. Section 25.2511-1(a). 

26. Treas. Regs. Section 25.2511-1(c).

27. This is assuming that the party disadvantaged isn’t serving as the court appointed executor of the estate or, if there’s no court appointed executor, then as a statutory executor under IRC Section 2203. While IRC Section 2205 does state that any party other than the estate that paid estate taxes “shall be entitled” to reimbursement from the estate that hadn’t yet been distributed by the executor—this provision shouldn’t be viewed as creating a right under federal law for the party that paid part of the estate tax liability it did now owe under the federal reimbursement provisions. In fact, the regulation interpreting this IRC Section (Treas. Regs. Section 20.2205-1) only goes as far as stating such party “may be entitled to reimbursement,” and jurisprudence has held that federal law creates no such right as it’s a matter for state law. See Riggs, supra note 15. 

28. Estate of Boyd v. Comm’r, 819 F.2d 170 (7th Cir. 1987); see supra note 23.

29. Redstone v. Comm’r, 145 T.C. 259 (2015).