Table of Contents
Where We Left Off
In the last grantor trust article, we explored some issues with the exercise of a power of substitution. Before developing this and other grantor trust powers however, it is important to call out the issue of spousal attribution.
What is spousal attribution? It is a rule found in IRC Section 672(e), which deems the grantor to hold any powers or interests held by their spouse. Spousal attribution relates back to the later of:
When the power or interest of the spouse is created, or
When spouses get married.
Does divorce later terminate grantor trust status? Not necessarily. This point has been debated by practitioners in the wake of the repeal of IRC Section 682 under the Tax Cuts and Jobs Act of 2017. While former IRC Section 682 didn’t necessarily state that spousal attribution terminates after divorce or legal separation, it did provide that income of a trust which a non-grantor (former) spouse is entitled to receive after divorce or legal separation will be taxed to such beneficiary and not to the grantor. This highlights a modern criticism of SLATs that continue to benefit a spouse post-divorce, as the income of the trust continues to be taxed to the grantor under the spousal attribution rules of 672(e).
It is important to note that IRC Section 672(e)(2) clarifies that an individual who is legally separated from their spouse under an applicable decree is not considered to be married. However, this rule is given only limited application to IRC Section 672(e)(1)(A), which considers whether spouses are married at the time a power or interest is created. This rule is also given only limited utility for a period where spouses are “legally separated.” While many interpret this to be a divorce rule, a plain reading indicates that this carve-out can only apply (presumably to powers created) during a period where spouses are legally separated, but not yet divorced.
This rule can seem punitive, as it means divorce does not terminate grantor trust status if grantor trust status is derived through this attribution. It is important to note, however, that some powers separately apply to the grantor and grantor’s spouse independent of attribution – but with testing for a particular tax year and not relating back to the inception of the interest. We will later discuss IRC Section 677, which examines the retention or use of income for the benefit of both the grantor and grantor’s spouse. The analysis of that Code Section necessarily takes place independent of the attribution rule of 672(e) by its wording, as it has a present focus and not a retroactive focus. Yet, the two can still work in tandem as we will explore further in this article.
There are, however, some tips and traps to be wary of when it comes to spousal attribution.
Attribution is One-Way
As discussed in this article on SLATs, 672(e) provides only for an attribution of a power or interest of a spouse who is not a grantor of a trust to the actual grantor. Recall that a grantor will be someone creating a trust, or making a gratuitous transfer to a trust.
Now as previously discussed, the scope of a grantor’s powers can technically extend to property they did not gratuitously transfer under Rev. Rul. 85-13. But where multiple gratuitous transferors are involved, they typically will each have the grantor trust rules apply independently to their respective gratuitous transfers. This means the analysis becomes a bit different when both spouses are “grantors” of the trust, as there can be cross-attribution each way only with respect to the powers or interests of the spouse not making a gratuitous transfer. Keeping this same line of reasoning, if only one spouse makes a gratuitous transfer, then only the powers of the non-contributing spouse would be attributed back to the grantor – the non-contributing spouse would not automatically become an income taxpayer with respect to the trust (short of filing a joint return).
Looking at the reverse however, a transfer – by gift or by sale – from the non-contributing spouse to the contributing spouse could fall under IRC Section 1041. This rule provides for non-recognition of gain on such transfers, but also imposes a carryover basis if there is a sale or exchange instead of a gift. This rule applies both to transfers during marriage, and after marriage between former spouses if the transfer is incident to divorce. This rule could create odd outcomes post-divorce if a former spouse continues to hold a power or interest subject to the attribution rule.
Of course, we could end up in a chicken-and-egg scenario here. Could 672(e) be superseded by 1041, or vice versa? Let’s say a non-contributing spouse sells assets to a SLAT, in order to get BDIT-type treatment. Arguably they have not made a gratuitous transfer so long as the sale is for fair market value, so spousal attribution couldn’t apply to them. This necessarily puts us in the territory of having to rely upon IRC Section 1041 to create a “deemed” sale to the grantor.
However, Rev. Rul. 85-13 only applies to the grantor. So, the non-contributing spouse (now making the sale) ironically could not be taxed on trust income – including for assets sold to the SLAT. In that situation, it is spousal attribution that applies to the grantor to further impute the spouse’s powers and interests over the sold assets and perhaps extend the penumbra of Rev. Rul. 85-13 over the sold assets. But if any portion of the non-grantor spouse’s contribution to the trust was deemed to be “gratuitous,” does this mean we now have cross-attribution with respect to all of the grantor’s prior transfers? In this situation, the transfer tax exemptions for the SLAT could be jeopardized with respect to the beneficiary spouse’s transfers only. Luckily, however, there is no direct spousal attribution rule for transfer tax purposes that could apply to the grantor – this is only an income tax rule.
Time of Marriage and Divorce
Keep in mind the rule of IRC Section 672(e)(1)(B), which creates attribution at the time of marriage. Under U.S. v. Windsor, married couples of any gender are treated as being married under federal law – including federal tax law. Note also that some states recognize common law marriage as well. Determining the time of marriage is important for determining the scope and effect of the spousal attribution rule, as it can create grantor trust treatment where such treatment was not anticipated. For example, if partners suddenly crossed the factual threshold for having a common law marriage under state law, this creates the moment of inception for spousal attribution in theory but it is also creates a conundrum of tracking the relevant facts and circumstances.
Similarly, termination of marriage must be carefully tracked. Arguably the death of a (non-grantor) spouse terminates spousal attribution, because the powers or interests of the deceased spouse are then terminated. However, termination during life must be carefully tracked and reconciled. Some states may provide that the interests of spouses are automatically terminated under trusts and other testamentary interests if and when a decree of divorce is issued by a court. The question becomes which state’s laws control. For example, if a grantor is domiciled in a state which applies this rule but the trust is administered under the laws of a different state, determining which laws control could be challenging. The trust may also expressly override these outcomes.
Likewise, let’s say there is an annulment. In that situation, there would have been attribution under 672(e) during the period of marriage. But annulment has a retroactive effect of treating spouses as if they were not married. Does this mean we get to look backwards and determine that there was no spousal attribution during this period?
Unfortunately, many of these questions are unsettled but nonetheless highlight areas that should be analyzed in divorce planning. Note however that attribution ends when powers end. Negotiation of this point during a divorce settlement is important. And while termination of powers may not be possible, could a pause – even for a second – be enough to break the attribution that is retroactive to the creation of the power or, if later, the marriage of the parties? For example, what if the trust provides for automatic termination of a spouse’s beneficial interest upon the entry of a decree but an immediate recommencement after the entry of a decree? An argument could be made that this is the creation of a new power that is not attributed after the marriage ends. Note also that the rules of IRC Section 677 only apply during marriage as well, so such a “pause” could end the application of this rule as well.
Of course, one must consider whether the transfers between spouses – along with pauses and recommencements of interests in a trust – could themselves be treated as transfers of property interests, perhaps under IRC Section 1001 or for gift tax purposes. Care should be taken to bring these powers within the purview of the aforementioned IRC Section 1041, and perhaps IRC Section 2516 (each of which impose certain requirements for transfers to a former spouse incident to divorce to avoid taxation).
Adverse Party Determinations
Some grantor trust powers distinguish between adverse and non-adverse parties. For example, IRC Section 672(c) carves out an exception whereby a trustee is not related or subordinate to a grantor if the trustee in question is an adverse party (such as by holding an interest in the trust). The spousal attribution rule of IRC Section 672(e), however, prevents a spouse’s powers or interests from being examined in isolation for purposes of determining whether or not they are an adverse party.
One of the points brought up by thought leaders who argue for the non-application of 672(e) post-decree is that the mere termination of marriage perhaps causes a former spouse to become an adverse party. This may be a stretch, but it is still a point to analyze if and when a marriage terminates.
Conclusion
Knowing how this spousal attribution rule interacts with, and even supersedes, certain grantor trust powers we consider in upcoming articles is paramount when applying the grantor trust rules. But, this attribution rule also highlights the importance of analyzing this rule in the wake of divorce or legal separation. While certain provisions of the Tax Cuts and Jobs Act are subject to sunset, note that the changes dealing with grantor trusts were made permanent and are not subject to sunset.