NAELA Journal, Volume 10, No. 1
When Worlds Collide:
State Trust Law and Federal Welfare Programs
By Ron M. Landsman, Esq., CAP
About the Author
Ron M. Landsman, Esq., CAP, is principal of Ron M. Landsman, P.A., Rockville, Md. As Editor for Quality Control, he is a member of the NAELA Journal Editorial Board.
Mr. Landsman would like to thank the following individuals for their kind assistance in sharing their knowledge, experience, and insights with him in preparing this article: Tom Cholis, Managing Director, Chevy Chase Trust; Robert Fleming, CELA; David Lillesand, Esq.; William Meyer, Esq.; Mary O’Byrne, Esq.; Rene Reixach, Esq.; Prof. Mary Radford; Nell Graham Sale, CELA; Kevin Urbatsch, Esq.; and Neal Winston, CELA.
“Special needs trusts,” which enable people with assets to qualify for Supplemental Security Income (SSI) and Medicaid, are the intersection of two different worlds: poverty programs and the tools of wealth management. Introducing trusts into the world of public benefits has resulted in deep confusion for public benefits administrators.
Trusts traditionally involve wealth and its management. SSI and Medicaid are public benefit programs for poor people. Even if adjusted for inflation, the asset limits for SSI eligibility — the easiest door to Medicaid eligibility — or a year’s SSI income is of the same order of magnitude as just the fees for most private bank or trust firms. The one valuable asset permitted — a personal residence — is eschewed as an asset in wealth management trusts. SSI and Medicaid administrators and beneficiaries alike would have no reason to be familiar with trusts solely by involvement with SSI or Medicaid.
However, the federal agencies that administer SSI and Medicaid — the Social Security Administration (SSA) and the Centers for Medicare & Medicaid Services (CMS) — had no choice but to address how trusts would fit into SSI and Medicaid eligibility requirements given Congress’s authorization of special needs trusts. The confusion arising from the merger of trust law with public benefits is sharply drawn in the agencies’ attempts to define what it means for a trust to be for the sole benefit of the public benefits recipient. Public benefits administrators have focused on the distributions a trustee makes rather than the fiduciary standards that guide the trustee. The agencies have imposed detailed distribution rules that range from the picayune to the counterproductive and without regard, and sometimes contrary, to the best interests of the disabled beneficiary.
This article critiques the agencies’ treatment of sole benefit. It finds that the agencies, unfamiliar with trust law, overlooked state trust law as the source for understanding what it means for a trust to be for the sole benefit of a specific beneficiary. It is divided into five substantive parts:
• Part II reviews the history of Congress’s treatment of trusts regarding public benefits through its decision to bring trusts inside public benefits programs;
• Part III details CMS’s and SSA’s interpretations of the new public benefits trusts, especially their definitions of “sole benefit” and “solely for the benefit of”;
• Part IV analyzes sole benefit as a matter of state trust law, which Congress had chosen as the mechanism for implementing its policy;
• Part V reviews the reasons why CMS’s and SSA’s very different view of sole benefit would not be entitled to deference from the federal courts under the Skidmore doctrine; and
• Part VI outlines an approach CMS and SSA might take to achieve the goals they sought in monitoring the use of trusts.
II. Congress Brings Trusts Inside the Public Welfare System
Congress first legislated on trusts and Medicaid in 1986 to stop the use of trusts to avoid Medicaid’s resource limits by elders needing long-term nursing home care. It then enacted the “Medicaid qualifying trust” (MQT) provision under which a trustee of a discretionary first-party trust would be deemed to exercise discretion to make any payments permitted under the terms of the trust. All trust income and principal would be available if trustee discretion permitted payment for the benefit of the elder. The common view among Elder Law attorneys is that the statute was ineffective, although case law suggests otherwise.
Nonetheless, after something of an onslaught of publicity about Medicaid planning, Congress in 1993 enacted a new law, as part of the Omnibus Budget Reconciliation Act of 1993 (OBRA ’93), to address the use of trusts in Medicaid planning, replacing the 1986 act. Under the new rules, trust assets of the Medicaid-planning settlor or his or her spouse were deemed available if there were “any circumstances” under which payment could be made to or for the benefit of the settlor. This rule applied without regard to the purpose for which the trust was established or any limitations in its terms. At the same time, Congress exempted three specific trusts from the new trust rules: two to protect resources (and the income they generate) and one to solve an eligibility problem related to nontrust income. Congress also, for the first time, encouraged Medicaid planning by exempting from its antitransfer rules gifts to a trust for any disabled person under age 65.
The decision to exempt some trusts from the Medicaid trust rules and to actively encourage donors to qualify for Medicaid by funding trusts for children or grandchildren with disabilities reflected a sea change by Congress. All along, Congress’s goal had been to stamp out every vestige of Medicaid planning by individuals for themselves, though not for others. What Congress did by exempting special needs trusts from the new rules and providing an exemption from the antitransfer rules for funding trusts for others with disabilities was the opposite of its previous treatment of such trusts. Individuals with disabilities were now expressly allowed to keep assets in trust while they obtained or maintained Medicaid benefits. This was new.
These new rules exempted two kinds of resource trusts for people with disabilities, those with a single beneficiary and those operated by nonprofits, known as pooled trusts, with many disabled beneficiaries, each with his or her own account. The most significant requirement for both types of trusts is that upon the death of the beneficiary, Medicaid programs must be reimbursed for benefits they paid for the beneficiary — before the trustee could make any distributions to heirs or legatees. In both types of trusts, a parent, grandparent, legal guardian, or court may establish the trust or trust account. Aside from these elements, individual and pooled trusts have some differences provided by statute. Pooled trusts, but not individual trusts, also may be established by the beneficiaries themselves. There is no age limit for pooled trusts, but individual trusts must be established before the beneficiary turns age 65, and the nonprofit may retain funds rather than reimburse Medicaid. Other differences reflect the different nature of pooled trusts. These must be operated by a nonprofit entity, all participants must be disabled and each must have his or her own account.
Regarding transfers of assets, Congress added to the existing exemptions transfers “to a trust (including a trust described in subsection (d)(4)) established solely for the benefit of” the individual’s disabled child and “to a trust (including a trust described in subsection (d)(4)) established solely for the benefit of” any disabled individual under age 65. “Solely for the benefit of” was used three times, once to describe exempt pooled trusts and twice, in parallel provisions, to describe the trusts to which exempt transfers could be made. The term sole benefit also appears, twice, both times regarding transfers to or for “the sole benefit of the individual’s spouse.”
Congress extended the OBRA ’93 trust rules to SSI in 1999; it reintroduced an antitransfer rule for SSI, which had been dropped in 1993. The SSI provisions are similar, but not identical, to the prior Medicaid provision. Whether Congress intended different treatment is not clear, but that is beyond the scope of this article other than with respect to sole benefit.
III. CMS and SSI Implement the Trust Provisions
A. CMS: Transmittal 64
To implement the OBRA ’93 trust provisions, CMS amended its State Medicaid Manual, writing or adding Sections 3257–3259 in Transmittal 64. To a large extent, CMS only restated the statutory language. It also addressed the unique situation of 42 U.S.C. § 1396p(d)(4)(B) trusts, acknowledged the state law nature of trusts, and attempted to define sole benefit and “solely for the benefit of.”
In Transmittal 64 CMS said that a trust is “for the sole benefit” of a person “if the trust benefits no one but that individual, whether at the time the trust is established or at any time in the future.” Conversely, a trust or transfer is not for someone’s sole benefit if it “provides for funds or property to pass to a beneficiary who is not” in one of the three categories to whom exempt transfers can be made. However, a trust can still make a post-mortem disposition: “[T]he trust may provide for disbursal of funds to other beneficiaries, provided the trust does not permit such disbursals until the State’s [reimbursement] claim is satisfied.” CMS added that, to be “for the sole benefit of” an individual, a trust “must provide for the spending of the funds involved for the benefit of the individual on a basis that is actuarially sound based on the life expectancy of the individual involved.” Presumably, this provision made sure that no one else would receive any benefit from the trust. But this restriction does not apply to trusts that include a payback to the state Medicaid agency, i.e., if the state Medicaid agency is entitled to payback, sole benefit does not require the funds be dispersed to or for the benefit of the beneficiary during his or her anticipated life expectancy. If a parent or grandparent funding a trust for a disabled child or grandchild wants to enjoy the exclusions under (c)(2)(B)(iii) or (iv), respectively, in which payback is not otherwise required, he or she can satisfy sole benefit by adding payback in lieu of “actuarially sound” language.
B. SSI: Program Operations Manual System (POMS)
With Congress’s extension of the Medicaid trust rules to SSI, the SSI unit of SSA also had to address the meaning of sole benefit. Like CMS, SSA properly acknowledged that trusts are creatures of state law. Unlike CMS, SSA had no reason to defer to the states since SSI is exclusively a federal program. Also unlike CMS, SSA was not implementing its own statute; the exceptions to its antitrust rules are by statutory cross-reference to the same trusts exempted under the Medicaid statute. Again unlike CMS, SSA attempted to provide directions to staff about specific state trust law issues. Unlike CMS, SSA addressed not only sole benefit, which is in the federal statute, but also a range of state trust issues.
With respect to sole benefit, SSA was loyal to the principle of congressional intent, inferring different meanings where Congress used slightly different words, and distinguishing “for the benefit of” from “for the sole benefit of.” In language redolent of Transmittal 64, SSA explained that a trust is established for the benefit of an individual
if payments … from the … trust are paid to another person … so that the individual derives some benefit from the payment (emphasis added).
A trust is for the sole benefit of an individual
if the trust benefits no one but that individual … at any time for the remainder of the individual’s life (emphasis added).
After painstakingly making this distinction, SSA without explanation disregarded it and said that individual special needs trusts “for the benefit of” a disabled individual under (d)(4)(A) also must be for that person’s sole benefit. Sole benefit precluded a provision that would “provide benefits to other individuals or entities during the disabled individual’s lifetime,” but it then provided as an “[e]xception to the sole benefit rule for third party payments” any “[p]ayments to a third party that result in the receipt of goods or services by the trust beneficiary.”
In 2012, SSA added an example to POMS to show that reimbursing family members for travel expenses to visit the trust beneficiary was not for the sole benefit of the beneficiary. SSA reconsidered in the face of vigorous protests from charities and special needs trust attorneys first by removing the example but ultimately substituting rules that limit payment for third party travel to two narrow exceptions.
IV. The Proper Meaning of Sole Benefit Trust Comes From the State Law of Trusts
A. State Trust Law Provides a Clear, Comprehensive Meaning That Conforms to Congressional Intent
Congress chose to use trusts, creatures of state law, as the mechanism for coordinating private assets and means-tested public benefits and granting special treatment for certain federal programs for beneficiaries of trusts and for the donors of such trusts. These trusts all have to be for the “sole benefit” of the trust beneficiary. There is no general federal law of trusts, nor is sole benefit a term of art in state trust law, but familiar state trust law principles provide an interpretation that conforms precisely to Congress’s intent in providing special treatment for beneficiaries of trusts established “solely for” their benefit, and the donors to such trusts.
The notion of benefit or beneficial interest is fundamental in the law of trusts. Adding “sole” or “solely for” addresses another fundamental trustee duty, the duty to be impartial in managing the trust for the benefit of all beneficiaries. Together, these provide a definite and definitive meaning of what Congress sought to achieve in using trusts as part of the public benefits system: the needs and welfare of the person with a disability have absolute priority over those of any other beneficiary in all aspects of the trustee’s management of the trust.
The beneficiary is one of the three essential elements of a trust. Beneficiaries are the individuals, usually named in the trust document, to whom the trustee owes the duty of loyalty. “Who the beneficiaries of a trust are depends on the manifestations of the settlor’s intent.” It is the individual(s) selected and named — whether by name, or class, or some other characteristic — by the settlor as the one(s) he or she intends to benefit and to whom the trustee owes all of his or her duties.
A beneficiary gets the benefit of the trust through the trustee’s use of trust assets for his or her benefit. The trustee may give the beneficiary cash, may purchase goods or real property (e.g., an automobile, home, or computer) for his or her use, or may pay others to provide services for the beneficiary. The fact that the trustee pays someone else to do something for the beneficiary does not make the service provider a beneficiary of the trust, and that payment is not a benefit of the trust. Even naming a specific person to provide services to the trust, trustee, or beneficiary does not mean the settlor intended to give that person a beneficial interest in the trust; therefore, getting paid for providing those services is not considered getting a benefit from the trust. A fortiori, someone not named in the trust who is paid to provide services to the beneficiary is not, by the fact of payment alone, getting a benefit under the trust.
A special needs trust always has at least two beneficiaries: the current life beneficiary with a disability and whoever enjoys the benefit of the property, if any remains, after the death of the life beneficiary. The trustee is not to discriminate between them unless the settlor has directed the trustee to do so. Treating the beneficiaries with due regard for their respective interests requires nonfavoritism in procedure and results. Most of the cases on the duty of impartiality involve successive beneficiaries, in which one (or more) receives income for life and the other(s) later get the principal. The trustee’s duty is to “preserve a fair balance between them.” Where there are inherent differences in what each beneficiary is entitled to, as with income to one and remaining principal to the other, the trustee cannot treat them exactly the same, but the duty imposes on the trustee an obligation to consider the interest and needs of each in all of the decisions to be made. To say a trust is for the sole benefit of one individual or is established “solely for [his or her] benefit” means that the duty of impartiality has been suspended and that, between the life beneficiary and any remaindermen, the trustee is to consider only the interests and needs of the sole benefit beneficiary and is to give no weight to the interests or needs of the remaindermen.
Trustees may have broad discretion in meeting the duty of impartiality, but that hardly negates there is such a duty; eliminating the duty by making a trust for the sole benefit of one beneficiary reflects a real and substantial shift in the trustee’s obligations and in the resulting management and use of trust assets. The trustee should not, as would otherwise be the case, consider an investment strategy designed to ensure growth of principal so that there will be something to pass on to the remainderman, nor may the trustee stint on distributions to meet needs not met by other sources or resources to ensure that there is a remainder to be distributed. The remainderman is a beneficiary, one to whom the trustee may owe some duties. All of the beneficiaries have some interests in common — that the trustee be prudent, competent, loyal, and conscientious — and any of the beneficiaries may enforce those duties. The sole benefit requirement addresses only impartiality, not the other obligations a trustee has to all the beneficiaries.
This understanding of sole benefit squares perfectly with Congress’s intent in providing for individuals with disabilities who themselves seek to enjoy the benefit of earned, inherited, or otherwise acquired assets while using SSI and Medicaid. The same is true for those who qualify for Medicaid while committing their resources to the benefit of another person with a disability, as the antitransfer exceptions permit. Congress’s purpose was to make certain that the person with a disability received the beneficial enjoyment of the assets, and that is achieved by sole benefit, properly understood as a modification of the duty of impartiality.
B. Sole Benefit, Although It Derives Its Meaning from State Law, Is an Independent Federal Requirement that Preempts Contrary State Law
If state law provides a useful, functional definition of sole benefit that promotes federal policy, the next inquiry is how, if at all, that definition attaches to the federal statutory term. Federal law controls, but Congress did not oust state law entirely — rather the contrary, it adopted state law as the means for achieving its ends.
The Third Circuit in Lewis v. Alexander treated the interplay as a matter of preemption. The court analyzed which provisions of state law were pre-empted by the federal statute. This is not the only, or perhaps even the best, way to analyze the conflict between state and federal law, but it provides a framework for analyzing sole benefit. Here, as it has in many areas, Congress has provided a benefit within the context of state law. There is no general federal common law and no general federal law of property, trusts, or estate law. State law controls absent a specific federal statute, policy, or interest that requires otherwise. The Supreme Court noted recently, “In most fields of activity … this Court has refused to find federal preemption of state law in the absence of either a clear statutory prescription, or a direct conflict between federal and state law.” Where Congress operates through state law, “the basic assumption [is] that Congress did not intend to displace state law.” Application of preemption resulting in a state law being unenforceable is limited to specific points of conflict.
In Lewis v. Alexander, this question of how much state law was abrogated was central to resolving Pennsylvania’s attempt to regulate — a cynic might say strangle — pooled special needs trusts. The Third Circuit’s answer can be summarized as “not very much.”
[T]here is no reason to believe [Congress] abrogated States’ general laws of trusts or their inherent powers under those laws. … [W]e reject the conclusion that application of these traditional powers is contrary to the will of Congress. After all, Congress did not pass a federal body of trust law, estate law, or property law when enacting Medicaid. It relied and continues to rely on state laws governing such issues.
To the extent Congress specified what was permitted as a condition to Medicaid eligibility, the state was not free to add additional requirements. Reviewing the five specific provisions challenged by plaintiffs in the Third Circuit, the court found that the state could not limit pooled trust retention to less to 50 percent, special needs trust expenditures to needs related to the individual’s disability, participation in pooled trusts to those who could not meet their needs without the trust, or pooled trust participation to those under age 65. Congress addressed each of these and made no allowance for states to limit or modify what it, Congress, would allow.
Like the four provisions struck down, the sole benefit requirement was plainly delineated by Congress and is not one that the states can modify. It is an independent element of what Congress requires for those who want to use federally funded Medicaid benefits: Preemption precludes the states from adding more to the requirement. As a practical, functional matter, it means that the remainder beneficiaries have no standing to challenge disbursements for the life beneficiary as in derogation of their rights as remaindermen, and that the life beneficiary can require the trustee to disregard the remaindermen’s interests. The trustee can and should favor the beneficiary with a disability. The court in Lewis did not have occasion to address the meaning of sole benefit since the state had only added to its own statute a sole benefit provision similar to that in the federal statute, and the meaning of neither was at issue.
The court upheld the fifth provision challenged, subjecting pooled special needs trusts to petitions for termination by the attorney general. It said that the state may subject special needs trusts, like all trusts, to court jurisdiction. The threat of termination is just one of the arsenal of general trust law provisions “for protecting the trust and the interests of its beneficiaries.” Congress’s decision to draw on state trust law as the mechanism for implementing its Medicaid policy carries a certain tension, the court noted. If Congress has defined sole benefit trusts, the court asked, what about the other duties imposed on trustees?
There is necessarily some tension between this conclusion [that state trust law and the states’ powers under it are not abrogated] and the bar on states adding requirements. For example, even application of the trustee’s traditional duty of loyalty — to administer the trust solely in the interests of the beneficiaries — could be considered an extra requirement.
The court later said:
Pennsylvania’s general trust law contains numerous provisions for protecting the trust and the interests of its beneficiaries. For example, Pennsylvania law imposes duties of loyalty, impartiality, prudent administration, and prudent investment.
Sole benefit, understood as a modification of the duty of impartiality, would, under the Third Circuit’s preemption analysis, conflict with the duty of impartiality. Because the duty of impartiality is only a default rule, however, it presents a preemption conflict only in the limited sense of precluding an option trust settlors might otherwise prefer but are denied if they want to enjoy using both resources and Medicaid or SSI.
The difference between the federal agencies’ understanding of “for the sole benefit of” and that suggested by state law is vast. The question for a court asked to enforce the federal agencies’ interpretation is whether it should defer to the agencies or, instead, review the statute, legislative history, and other authorities to come to its own independent conclusion. The following section addresses the standards for undertaking that task.
V. On the Sliding Scale of Deference to Agency Determinations,
Neither CMS nor SSA has Earned the Right to Substantial Deference
for Its Definition of Sole Benefit
A. Absent Careful Use of Rulemaking Authority, Deference is a Function of an Agency’s Competence in the Specific Area
Judicial deference to action of an agency under a statute it administers ranges “from great respect at one end … to near indifference at the other.” At one end is almost total deference to agency exercise of legislative rulemaking authority given by Congress within the agency’s jurisdiction. It is just such authority that the Supreme Court typically has relied on in cases involving CMS and Medicaid and SSA and SSI. A court should defer to an agency that has engaged in legislative rulemaking under its statute if “the agency’s answer is based on a permissible construction of the statute.” The agency’s answer does not have to be compelling, or the only permissible reading, or even the one the court finds most reasonable, as long as it reflects “a reasonable accommodation of conflicting policies that were committed to the agency’s care by the statute.” Such regulations are “given controlling weight unless they are arbitrary, capricious, or manifestly contrary to the statute.” This degree of deference is warranted where the agency enjoyed, and exercised, specific rulemaking authority.
Here, CMS does not enjoy and thus could not exercise the kind of specific rulemaking authority it has in determining what constitutes income or resources. The Court in Blumer put substantial reliance on CMS’s authority under 42 U.S.C. § 1396a(a)(17), in which Congress required state plans to utilize:
reasonable standards … which … provide for taking into account only such income and resources as are, as determined in accordance with standards prescribed by the Secretary, available to the applicant or recipient … .
The mandate for the transfer and trust rules to be in state plans, in the next subsection, provides in its entirety that state plans shall:
comply with the provisions of section 1396p of this title with respect to liens, adjustments and recoveries of medical assistance correctly paid, transfers of assets, and treatment of certain trusts; …
42 U.S.C. § 1396p itself has nine specific grants of rulemaking authority to the secretary, but none concern what constitutes a trust “for the sole benefit” of a person or established “solely for [his or her] benefit.” The agency can define income and resources, and as CMS has noted, all of the normal rules may apply to income or assets going into or out of a trust, and as to that the agencies may well have rulemaking authority, but there is no similar clear and specific grant of such authority respecting trusts.
SSA would appear to stand on no better footing. Whatever rulemaking authority SSA has, it has no authority to make rules respecting the Medicaid program. When Congress added Medicaid-like trust rules into the SSI statute in 1999, it enacted those rules directly into Title XVI, but the special needs trust exclusions were only by cross-reference to the Medicaid statute. The secretary’s general rulemaking authority may be relevant, but it is not at all clear it extends to what the Medicaid statute means. In any event, the secretary has not utilized it. Like CMS, SSA says the usual asset and income rules apply to what goes in or comes out of a trust, but again, that is a distinct question.
Absent agency exercise of rule-making authority, judicial deference depends on the quality of the action taken by the agency, as set out in Skidmore:
The fair measure of deference to an agency administering its own statute has been understood to vary with circumstances, and courts have looked to the degree of the agency’s care, its consistency, formality, and relative expertness, and to the persuasiveness of the agency’s position. … The weight [accorded to an administrative] judgment in a particular case will depend upon the thoroughness evident in its consideration, the validity of its reasoning, its consistency with earlier and later pronouncements, and all those factors which give it power to persuade, lacking power to control.
This directive can be broken into perhaps five distinct areas of inquiry:
1. Care and thoroughness. Whether the agency action reflects consideration of all the factors that contribute to the meaning and the consequences of different choices.
2. Consistency. Whether the agency had come to similar decisions on similar facts, and different decisions on distinguishable facts.
3. Formality. Whether the agency procedure, even if short of full rulemaking, reflects a procedure that guarantees careful consideration.
4. Relative expertness. Whether the subject of the decision, even if within the agency’s nominal jurisdiction, is one to which the agency brings greater expertise than courts or others.
5. Validity of reasoning and persuasiveness. The ultimate basis for deference is the power of the reasoning.
B. Application of the Skidmore Factors Suggests Deference Is Not Appropriate
Three of the five factors warrant extended consideration.
This is the first of the three areas in which a Skidmore factor suggests that a court should not defer to CMS or SSA. Neither agency has utilized its rulemaking authority. Transmittal 64 and POMS both reflect a lower degree of formality than notice-and-comment rulemaking. POMS has been cited by the Supreme Court to support its statutory interpretations and is frequently cited by the federal appellate courts, but mostly it appears to confirm the courts’ conclusions based on other sources rather than to persuade the contrary to other authorities.
2. Relative Expertise in State and Federal Law
Here, too, a Skidmore factor suggests, now much more strongly, that a court should not defer to either agency. Federal welfare officials have no professional expertise in state trust law. State courts have declined to defer to their state Medicaid agencies when they go outside of their specific area of competence to deal with substantive state law of property or other matters. The forays of SSA, in particular, into trust law reveal such serious miscomprehension that a court should be chary of deferring to its views. In using state trust law to achieve its policy, Congress has introduced a notion foreign to state trust law, something not understood by the agencies exactly because of their lack of expertise, so that even the federal nature of the issue does not translate into agency expertise.
a. State Trust Law
SSA uses words and phrases from trust law but ascribes its own, different meaning to them. If this only concerned nomenclature, it might be a minor problem solved by keeping clear the different meanings and their uses. But it goes beyond that: SSA attributes meaning to trust terms used in the federal statute that results in misconstruing Congress’ intent and frustrating Congress’ purpose.
i. Benefit and Beneficial Interest
SSA’s most serious deviation from accepted trust law is the meaning it gives to “benefit” from a trust. SSA says that the beneficiary must derive some benefit under its definition of “for the benefit of” and defines “solely for the benefit of” to mean that no one else can receive any benefit from the trust. It then provides an exception to this definition by allowing payments to a third party for goods or services for the benefit of the beneficiary. State trust law, by contrast, defines benefit in functional terms: it is the beneficial enjoyment of trust assets or income, for which one is not required to perform a service or deliver a good. The beneficiary is the one to whom the trustee owes its duties in the use of trust assets and income.
SSA’s definitions only make sense if what SSA means is that any payment by a trustee is a benefit of the trust to the person who receives the payment. This looks like nothing so much as SSA’s old “name on the instrument” rule. From that perspective, a trustee’s $20 payment to a cabdriver to take the beneficiary to a doctor’s office for an exam gives the cabdriver a trust benefit. Such a payment is saved from violating “sole benefit” by the POMS exception allowing payments to third parties for the purchase of goods or services for the beneficiary. A definition that accommodates the central purpose of the thing it defines only as an exception to its general rule does not have a well-grounded general rule.
Like the Ptolemaic system, SSA’s definitions are a jerry-built approximation of the real thing and must in time lead to inaccurate, not to say bizarre, results. This happened for example in a state Medicaid agency’s attempt to deal with a trustee’s payment of compensation to a mother to remain home to care for her seriously disabled son. Following SSA’s theory, payments to the mother were treated as giving her a benefit of the trust, in violation of the sole benefit rule, and since she was doing what mothers (or at least parents) are expected to do, it was not shoehorned into the permissible exception. A traditional analysis under state trust law of sole benefit, as a guide to trustees, might permit payments to parents, but it would do so only when to do so were in the best interests of the beneficiary.
SSA has applied state trust law inconsistently and often incorrectly in determining when trusts are irrevocable. It asserts as a “general principle of trust law that if a grantor is also the sole beneficiary of a trust, the trust is revocable regardless of language in the trust to the contrary,” and a trust that does not name specific individuals as beneficiaries is always revocable. The modern rule, found in Restatement of the Law (Third), Trusts, is to the contrary: A trust is not for a sole beneficiary when the settlor “names heirs, next of kin, or similar groups to receive the remaining assets [after the beneficiary’s] death”; therefore, the trust is not revocable per se. But SSA requires state-by-state analysis and finds that the rule survives absent an express judicial decision to the contrary, often overlooking statutory reversal, resulting in “unrelenting and incorrect application of the [Doctrine of Worthier Title].”
This error, combined with its failure to appreciate the significance of third-party creators, can lead to anomalous results that violate long-standing state law. Parents who sought to avoid SSA’s finding of revocability by having the trust name themselves as contingent beneficiaries could run afoul of the deeply established rule that no one can make a will for another person. Such a plan of distribution could easily deviate from state intestacy law depending on who survived, if anyone remarried and had later-born children, and the state’s anti-lapse statute.
“Establish” is not a term of art in the law of trusts. Historically, a trust came into existence when the owner of property conveyed the beneficial interest to another. The question often requiring resolution was when, if ever, those duties arose, viz., whether the conveyance itself defined the duties, whether a mere declaration by the settlor/trustee could cause enforceable duties to come into existence, or whether later acquired property could be subject to duties previously declared. A person could not establish a trust of someone else’s property — that is, convey a beneficial interest to another — any more than he or she could sell it to another and convey good title. Because the trust arose from the transfer of one’s own property, the person who transferred the property was the settlor who “established” the trust, and perforce that was the person who determined the terms of the trust. Similarly, because a trust involves obligations respecting property, a trust could not come into existence until a trustee was in possession of property.
In authorizing parents and grandparents to “establish” a trust, respectively, of their child’s or grandchild’s assets, Congress introduced a new notion into trust law – authorizing a person to establish a trust for management of another’s property.
SSA has tried to find the meaning of “establishment” in state trust law without recognizing that Congress has created an authority that has no basis there since Congress had introduced a practice not found in state trust law. Where a person establishes a trust with his or her own assets, he or she is the grantor, as SSA says; the trust comes into existence when the trustee receives property subject to the obligations imposed by the grantor and spelled out in a trust document. SSA, focusing on what trust law accepted as a result of its premises, treats the person whose assets fund the trust as the grantor, and treats the trust as established only when assets are delivered to the trustee. And since it was the act of funding the trust that brought it into existence, the disabled beneficiary whose assets fund a special needs is the grantor, SSA reasons, so the trust does not meet all of the requirements of Section 1396p(d)(4)(A).
This misapplies state trust law. State trust law focuses on funding by the owner because that is the only person who can establish the rules of the trust governing his or her property. In the situation created by Congress, where parents are to establish a trust to hold their child’s assets, the obligation that arises when the trust is funded is not determined by the funding but by the trust document executed earlier by the parents as settlors and accepted by the trustee. In most states, the terms of the document control. Thus, it is the parents who create the trust document who are the settlors, not the child whose assets later fund the trust.
SSA gets out of this conundrum by another jerry-built solution: looking for authority in state law for the creation of a trust without property, what it calls a “dry” or “empty” trust, by which it means an unfunded inter vivos trust. Absent express authorization of “dry” trusts, SSA treats the parent-created trust as inadequate under 42 U.S.C. § 1396p(d)(4)(A). SSA has found that some states do and some states do not, citing and relying on indistinguishable state court decisions that all require a trustee have title to property.
These misunderstandings of state trust law are not minor deviations from an otherwise sound approach. They concern essential features of trusts — creation, beneficial interests, and revocability. Having gotten them so wrong, SSA frankly cannot be trusted to construct a system that remains true to what Congress intended in adopting trusts to promote public benefit programs. Moreover, the fundamental incomprehension of what a trust is carries over to the problem of finding meaning in a trust for the sole benefit of a person as a matter of federal law.
b. Sole Benefit as a Feature of Federal Law
Even if the meaning of “sole benefit” is a question of federal law, the federal agencies can still claim no special expertise. Special needs trusts are a recent exception in the public benefits arena, and understanding sole benefit does not draw on any existing agency experience. The meaning of sole benefit does not involve a close reading of a highly complex statute with hundreds of moving parts, each of which affects dozens of others. Both agencies have the authority and the competence to address what kind of state property entities are available as resources or income and, for example, what kind of entities constitute income and what kind constitute resources, how they are counted or excluded, how they interplay with one another, and deeming from one to another; these are the kinds of statutory questions that CMS and SSA are uniquely qualified to answer. Both agencies correctly claim to continue to have primary authority to define and address the status of income and resources as they go into or come out of trusts, but that is a different matter from interpreting the law of trusts or addressing how trusts should operate, which is what sole benefit properly concerns.
Nor does interpretation involve a close reading of specific legislative history and Congress-agency interchange to interpret an otherwise opaque statute. CMS first addressed the issue in the wake of Congress’s enactment of OBRA ’93, and it has hardly visited the issue since. Special needs trusts were introduced into SSI with no history of agency action or reaction. The strongest argument in favor of the CMS/SSA definition is that Congress carried over much of the Medicaid statue to SSI after CMS had promulgated its informal State Medicaid Manual provision.
3. Validity of Reasoning and Persuasiveness
To address the validity of the agencies’ reasoning requires, first, considering what question they were trying to answer. Congress’s primary goal, inferred from the statutory structure, was to ensure that the person with a disability received the benefit of the resources set aside in trust. Much of what has been said above reflects the failure of CMS and SSA to understand how this operates in the context of a trust relationship. Three additional points merit discussion.
First, the definitions of sole benefit and “solely for the benefit of” add little to the statutory terms. They are the regulatory equivalent of saying the same thing, only louder. Where the statute says “solely for the benefit of,” both CMS and SSA say “the trust benefits no one but that individual … .” It is difficult to see what the definition adds; it provides no operational direction, no indication of what the trust document should say that the statute does not already address, and no guidance to trustees on what they can or cannot do.
Second, CMS provided the one real elaboration on the basic definition of sole benefit, its dog-in-the-manger requirement for actuarially sound distributions. This addition is neither well reasoned nor persuasive. In requiring minimum required distribution–type distributions, CMS is telling trustees to make distributions even when doing so is not required to pay for any goods or services the beneficiary needs or wants. This can, if it has any effect at all, only result in reducing trust resources prematurely and unnecessarily, wasting resources and potentially leaving the beneficiary with inadequate resources later in life.
CMS’s offer of payback as an alternative to actuarially soundness suggests that one of its reasons for requiring actuarial soundness was to protect beneficiaries from trustees who are also remainder beneficiaries. Payback would reduce if not eliminate an interested trustee’s incentive to preserve funds. Perhaps unsophisticated family member trustees are swayed by payback that forecloses effective self-interest, but imposing the requirement on disinterested professional trustees suggests SSA does not fully understand or appreciate what trustees do.
Third, the agencies’ definition of sole benefit fails to address trustee duties. A trust that does not relieve the trustee of his or her obligations to remainder beneficiaries under the duty of impartiality can only result in underserving the beneficiary with a disability. Without an express waiver in the trust document, a trustee must take into account both the lifetime and remainder beneficiaries’ interests. The conscientious trustee might well decline to spend trust funds when it threatens the remainder beneficiaries’ ultimate share. Both CMS and SSA, unsophisticated as they are in their understanding of trusts, are silent on this crucial aspect of trust administration. In their discussion of sole benefit, they have failed to understand the nub of the problem (i.e., the need to relieve a trustee of a duty of impartiality between the life and remainder beneficiaries) and instead focused their efforts on a proxy, requiring distributions for the life beneficiary.
The elements for discarding Skidmore deference are present. These issues are outside the normal agency purview, the agencies’ analyses have pervasive errors with respect to the state trust law they attempt to analyze, and their own definitions fail to fully serve Congress’s purpose and are not supported by solid reasoning.
VI. A Practical and Effective Approach to
Special Needs Trust Regulation That Utilizes Trustees as Partners
The task for CMS and SSA is to use their authority to develop standards and guidelines that utilize, rather than thwart, competent, responsible, properly trained trustees as their partners in making special needs trusts an effective tool in serving the needs of people with disabilities. If this were done properly, capable trustees would be the allies of the federal and state agencies in the efficient use of limited private resources. Beneficiaries would live better, more rewarding lives to the extent that resources can make a difference, at lower cost to Medicaid, with a greater possibility of more funds recovered through payback.
As a threshold matter, the agencies would appear to have somewhat different authority but common areas of concern. Even before getting to the substance of supervision of special needs trusts and trustees, CMS and SSA should resolve between themselves whether, as the court held in Lewis, there is a single, nationwide rule for special needs trusts rooted in federal law except to the extent it relies on state law operation. Related, it would be a rare but impressive moment if the two sister agencies could coordinate or even combine their efforts to develop a single, uniform approach for special needs trusts.
As part of that process, the agencies ought to come to recognize the limit of their ability to manage special needs trusts. Even aside from their many errors, the agencies went fundamentally astray in focusing on benefit as a measure of each transaction rather than as a guide to trustee conduct. By focusing on who received payment, a familiar cash benefits idea, they came up with a rule that is both too narrow and too broad.
Consider whether a trustee should pay a parent to provide necessary daily care for a severely disabled child, as in Hobbs v. Zenderman. Who benefits when a trust pays a mother a little more than minimum wage to stay home and care for her seriously disabled 7-year-old son? Under the CMS/SSA standard, as applied by a state Medicaid agency, this benefitted the mother. This was the result even though the independent corporate trustee with probate court approval concluded that paying the mother was preferable to paying substantially more while the mother worked outside the home for less. The trustee was denied the opportunity to exercise its discretion and purchase the care for the child the least expensive way possible. Plainly, the definition can be too narrow when it asks, “Who gets paid for caring for a disabled 7-year-old?” and does not permit payment on facts such as these.
But by the same token, it is too broad when it asks only, “Who gets paid for caring for a disabled 7-year-old?” Who benefits when a trust pays for certified nursing aides for 24-hour care for the seriously disabled 7-year-old oldest daughter of a family of five, with a stay-at-home mother and a highly compensated father? The CMS/SSA standard would see this as the child’s “sole benefit,” even though the parents and the other two children plainly benefit as well. The parents are relieved of a huge burden, while the younger children might feel like they get their mother back. The trustee might well conclude that the disabled child’s funds should be preserved while the family provides some of the care. A definition that always permits such payments is plainly too broad.
The proper question is not, “Who was paid to do the necessary work?” but “Is this the most cost-efficient way to take care of the child’s needs, taking all of the facts and circumstances into account?” This is how state trust law typically handles these issues, under the best interests of the beneficiary standard. A trustee might answer the question differently depending on the situation of the family. Is the mother home with a housekeeper because the family does not need her income and she elected to be an at-home mother? Does the trust have sufficient assets and scheduled income from a structure to pay for full-time aides now and still have sufficient funds to meet needs potentially 80 years in the future? These are decisions best left to a trustee to make on a case-by-case basis.
SSA’s recent effort to address using special needs trusts funds for travel to allow beneficiaries to visit with relatives, discussed above, illustrates the limit on what an agency can do through general rules. Consider the following scenario. The Washington, D.C., parents of a severely autistic child, living in a specialized Florida facility, would like to visit him, and the facility’s staff agrees that maintaining family ties is extremely beneficial for the boy. The options are to pay the cost for a round trip for one of the parents to visit the boy in Florida, perhaps including other travel costs such as car rental, hotel, and meals, or to pay the cost for two round trips for the boy and an aide to go to Washington, along with wages and other expenses for the aide.
For the trustee, the issue is the child’s best interests, including both current and future needs. Assuming the cost of the parents visiting is less, the trustee can consider whether trust funds are necessary. If the family is relatively affluent and could or would go even without a trust contribution to the costs, the trustee could reasonably decline to pay the costs, based in part on the need to preserve assets for the child’s long-term needs. If the family could not afford such a trip without financial support from the trust, the trustee must weigh the long-term benefits of providing family contact against the child’s future needs. But the standard would not be the formalistic “Who used the service?” but the real and practical one of “What is best for the child?”
Trustees of special needs trusts are subject to state trust laws that provide numerous protections for the beneficiary and for compliance with the terms of the trust. State probate court judges and state legislatures have established well-developed protections for beneficiaries of trusts, including accountings, removal, surcharge, penalties, fines, and in some circumstances even penal sanctions. This is where the supervision of trusts should be managed, not with public benefits eligibility workers with little or no experience in how trusts are managed or with sweeping policies in federal policy manuals that cannot take into account all of the facts of individual situations.
Sole benefit refers to the standard that guides the trustee, not to specific transactions. That is the level at which CMS and SSA can and should operate. Aside from clearing up the errors arising from failing to recognize the centrality of the role of the trustee and the proper use of state trust law, the agencies should address trustee standards and duties, perhaps where the problems are the most complicated. Few problems are more vexing than providing a home and caregiving for a disabled child whose family is poor and cannot afford appropriate accommodations and care. These and other complicated problems would benefit from a process in which the agencies get the benefit of trustees’ experience in the real world of special needs trusts.