delivered the judgment of the court, with opinion.
Chief Justice Fitzgerald and Justices Freeman, Thomas, Kilbride, Karmeier, and Burke concurred in the judgment and opinion.
OPINION
This case involves a dispute among the surviving children and grandchildren of Max and Erla Feinberg regarding the validity of a trust provision. The circuit court of Cook County found the trust provision unenforceable on the basis that it is contrary to the public policy of the state of Illinois. The appellate court affirmed. 383 Ill. App. 3d 992. Michael Feinberg, the Feinbergs’ son and coexecutor of their estates, filed a petition for leave to appeal pursuant to Supreme Court Rule 315 (210 Ill. 2d R. 315), which we allowed. We also allowed Agudath Israel of America, the National Council of Young Israel, and the Union of Orthodox Jewish Congregations of America to file a brief amici curiae pursuant to Supreme Court Rule 345 (210 Ill. 2d R. 345).
For the reasons that follow, we reverse.
BACKGROUND
Max Feinberg died in 1986. He was survived by his wife, Erla, their adult children, Michael and Leila, and five grandchildren.
Prior to his death, Max executed a will and created a trust. Max’s will provided that upon his death, all of his assets were to “pour over” into the trust, which was to be further divided for tax reasons into two trusts, “Trust A” and “Trust B.” If she survived him, Erla was to be the lifetime beneficiary of both trusts, first receiving income from Trust A, with a limited right to withdraw principal. If Trust A were exhausted, Erla would then receive income from Trust B, again with a limited right to withdraw principal.
Upon Erla’s death, any assets remaining in Trust A after the payment of estate taxes were to be combined with the assets of Trust B. The assets of Trust B were then to be distributed to Max’s descendants in accordance with a provision we shall call the “beneficiary restriction clause.” This clause directed that 50% of the assets be held in trust for the benefit of the then-living descendants of Michael and Leila during their lifetimes. The division was to be on a per stirpes basis, with Michael’s two children as lifetime beneficiaries of one quarter of the trust and Leila’s three children as lifetime beneficiaries of the other one quarter of the trust. However, any such descendant who married outside the Jewish faith or whose non-Jewish spouse did not convert to Judaism within one year of marriage would be “deemed deceased for all purposes of this instrument as of the date of such marriage” and that descendant’s share of the trust would revert to Michael or Leila.
In addition, the trust instrument gave Erla a limited testamentary power of appointment over the distribution of the assets of both trusts and a limited lifetime power of appointment over the assets of Trust B. Under the limiting provision, Erla was allowed to exercise her power of appointment only in favor of Max’s descendants. Thus, she could not name as remaindermen individuals who were not Max’s descendants or appoint to a charity. The parties dispute whether Erla’s power of appointment was limited to those descendants not deemed deceased under the beneficiary restriction clause. The trial court did not make a finding on this question and the appellate court did not discuss it.
Erla exercised her lifetime power of appointment over Trust B in 1997, directing that, upon her death, each of her two children and any of her grandchildren who were not deemed deceased under Max’s beneficiary restriction clause receive $250,000. In keeping with Max’s original plan, if any grandchild was deemed deceased under the beneficiary restriction clause, Erla directed that his or her share be paid to Michael or Leila.
By exercising her power of appointment in this manner, Erla revoked the original distribution provision and replaced it with a plan that differs from Max’s plan in two significant respects. First, Erla altered the distribution scheme from per stirpes to per capita, permitting each of the grandchildren to take an equal share, rather than favoring Michael’s two children over Leila’s three children. Second, Erla designated a fixed sum to be distributed to each eligible descendant at the time of her death, replacing Max’s plan for a lifetime trust for such descendants. The record suggests that Erla’s gifts will deplete the corpus of the trust, leaving no trust assets subject to distribution under Max’s original plan. Thus, while Erla retained Max’s beneficiary restriction clause, his distribution provision never became operative.
All five grandchildren married between 1990 and 2001. By the time of Erla’s death in 2003, all five grandchildren had been married for more than one year. Only Leila’s son, Jon, met the conditions of the beneficiary restriction clause and was entitled to receive $250,000 of the trust assets as directed by Erla.
This litigation followed, pitting Michael’s daughter, Michele, against Michael, coexecutor of the estates of both Max and Erla.
The trial court invalidated the beneficiary restriction clause on public policy grounds. A divided appellate court affirmed, holding that “under Illinois law and under the Restatement (Third) of Trusts, the provision in the case before us is invalid because it seriously interferes with and limits the right of individuals to marry a person of their own choosing.” 383 Ill. App. 3d at 997. In reaching this conclusion, the appellate court relied on decisions of this court dating back as far as 1898 and, as noted, on the Restatement (Third) of Trusts.
ISSUE PRESENTED
As a threshold matter, we must clarify the issue presented. We need not consider whether Max’s original testamentary scheme is void as a matter of public policy because Erla altered his scheme in 1997. Indeed, she could have done so again at any time before her death in 2003, exercising her lifetime or testamentary powers of appointment in any number of ways. For example, she could have named her grandson, Jon, as the sole beneficiary of the entire trust, or excluded the grandchildren entirely, appointing the entire corpus of the trust to Michael and Leila.
Indeed, counsel for Michele acknowledged at oral argument that Max and Erla could have accomplished the goal of benefitting only those grandchildren who married within their religious tradition by individually naming those grandchildren as beneficiaries of the will or the trust, without implicating public policy. Counsel argued that the violation of public policy occurred when Max used a religious description to define a class or category of descendants he wished to benefit, rather than mention them by name.
Of course, at the time Max prepared his estate plan, his grandchildren were too young to marry and it was possible that more grandchildren might have been born before the trust provisions took effect. As a result, Max could not have accomplished his purpose in the manner suggested by Michele. Even by the time Erla exercised her power of appointment, not all of the grandchildren had married.
Thus, the question we must answer is whether the holder of a power of appointment over the assets of a trust may, without violating the public policy of the state of Illinois, direct that the assets be distributed at the time of her death to then-living descendants of the settlor, deeming deceased any descendant who has married outside the settlor’s religious tradition. In effect, we are not called upon to consider the validity of Max’s estate plan as a whole, which would have continued to hold the assets in trust for the benefit of the grandchildren only so long as they complied with the restriction. Rather, we must assess Max’s beneficiary restriction clause in conjunction with Erla’s directions for distribution.
When the issue is clarified in this way, it becomes apparent that many of the arguments raised by Michele are not relevant. For example, under Max’s plan, an unmarried grandson would have begun to receive distributions from Trust B upon Erla’s death, only to forfeit further such payments if he were to marry a non-Jewish woman who did not convert to Judaism within one year. We need not decide if such a provision would violate public policy because no such provision is implicated in the present case.
Michele also suggests that a granddaughter who was married to a non-Jewish man at the time of Erla’s death might subsequently divorce and remarry, this time to a Jewish spouse, and make a claim upon the trust. This circumstance would raise the issue of whether such a descendant, previously deemed deceased, would be “resurrected.” Such an occurrence would require construction of the language of the trust document. Under Erla’s plan, however, this circumstance cannot arise because a fixed amount became distributable upon her death only to those grandchildren who then met the requirements previously set by Max.
Similarly, Michele’s argument that the beneficiary restriction clause is invalid because a court might be called upon to determine whether the spouse of a particular descendant is or is not Jewish is not well taken. It is undisputed in the present case that only one of the five grandchildren meets the requirements established by Max.
STANDARD OF REVIEW
This court has not had occasion to identify the applicable standard of review on the question of whether a provision in a trust document or will is void as a matter of public policy. It is clear, however, that such findings are subject to de novo review, because public policy is necessarily a question of law. Holstein v. Grossman, 246 Ill. App. 3d 719, 726 (1993), citing Zeigler v. Illinois Trust & Savings Bank, 245 Ill. 180 (1910). This conclusion is consistent with the well-established principle that whether a provision in a contract, insurance policy, or other agreement is invalid because it violates public policy is a question of law, which we review de novo. O’Hara v. Ahlgren, Blumenfeld & Kempster, 127 Ill. 2d 333 (1989) (affirming grant of summary judgment on basis that fee-sharing agreement offended public policy).
ANALYSIS
Michael argues before this court that the beneficiary restriction clause in his father’s trust was intended “to encourage and support Judaism and preservation of Jewish culture in his own family,” and that it was not binding upon Erla, who exercised her power of appointment consistently with the provision because it expressed her intent as well as Max’s. Michael argues, further, that even if Max’s beneficiary restriction was not revocable by Erla, the provision does not violate the public policy of this state when it is given effect via his mother’s distribution scheme. He asserts that the distribution scheme is a valid partial restraint on marriage of a type that has long been enforced in Illinois and elsewhere. According to Michael, the beneficiary restriction clause has no prospective effect that might subsequently influence a descendant’s decisions regarding marriage or divorce because, upon Erla’s death, no contingencies remained. He distinguishes the cases relied upon by the appellate court and urges this court to reject the cited Restatement provision as not accurately stating Illinois law.
Michele defends the Restatement provision and argues that this case comes within a line of cases dating back to 1898 in which this court invalidated testamentary provisions that operated to discourage the subsequent lawful marriage by a legatee or to encourage a legatee to obtain a divorce. Specifically, she argues that under Ransdell v. Boston, 172 Ill. 439 (1898), testamentary restrictions on marriage are valid only if they operate to benefit the intended beneficiary. Further, she argues that enforcement of the clause would violate both state and federal constitutions and that it violates public policy by offering a financial inducement to embrace a particular religion.
We note that this case involves more than a grandfather’s desire that his descendants continue to follow his religious tradition after he is gone. This case reveals a broader tension between the competing values of freedom of testation on one hand and resistance to “dead hand” control on the other. This tension is clearly demonstrated by the three opinions of the appellate court. The authoring justice rejected the argument that the distribution scheme is enforceable because it operated at the time of Erla’s death and could not affect future behavior, stating that its “clear intent was to influence the marriage decisions of Max’s grandchildren based on a religions criterion.” 383 Ill. App. 3d at 997. The concurring justice opined that while such restrictions might once have been considered reasonable, they are no longer reasonable. 383 Ill. App. 3d at 1000 (Quinn, P.J., specially concurring). The dissenting justice noted that under the facts of this case, grandchildren who had complied with the restrictions would “immediately receive their legacy” upon Erla’s death (383 Ill. App. 3d at 1000 (Greiman, J., dissenting)), and that the weight of authority is that a testator has a right to make the distribution of his bounty conditional on the beneficiary’s adherence to a particular religious faith (383 Ill. App. 3d at 1002).
We, therefore, begin our analysis with the public policy surrounding testamentary freedom and then consider public policy pertaining to testamentary or trust provisions concerning marriage.
When we determine that our answer to a question of law must be based on public policy, it is not our role to make such policy. Rather, we must discern the public policy of the state of Illinois as expressed in the constitution, statutes, and long-standing case law. O’Hara, 127 Ill. 2d at 341. We will find a contract provision against public policy only “ ‘ “if it is injurious to the interests of the public, contravenes some established interest of society, violates some public statute, is against good morals, tends to interfere with the public welfare or safety, or is at war with the interests of society or is in conflict with the morals of the time.” ’ ” Vine Street Clinic v. HealthLink, Inc., 222 Ill. 2d 276, 296 (2006), quoting E&B Marketing Enterprises, Inc. v. Ryan, 209 Ill. App. 3d 626, 630 (1991), quoting Marvin N. Benn & Associates, Ltd. v. Nelson Steel & Wire, Inc., 107 Ill. App. 3d 442, 446 (1982). Thus,
“In deciding whether an agreement violates public policy, courts determine whether the agreement is so capable of producing harm that its enforcement would be contrary to the public interest. [Citation.] The courts apply a strict test in determining when an agreement violates public policy. [Citation.] The power to invalidate part or all of an agreement on the basis of public policy is used sparingly because private parties should not be needlessly hampered in their freedom to contract between themselves. [Citation.] Whether an agreement is contrary to public policy depends on the particular facts and circumstances of the case.” Kleinwort Benson North America, Inc. v. Quantum Financial Services, Inc., 181 Ill. 2d 214, 226 (1998).
Because, as will be discussed below, the public policy of this state values freedom of testation as well as freedom of contract, these same principles guide our analysis in the present case.
Public Policy Regarding Freedom of Testation
Neither the Constitution of the United States nor the Constitution of the State of Illinois speaks to the question of testamentary freedom. However, our statutes clearly reveal a public policy in support of testamentary freedom.
The Probate Act places only two limits on the ability of a testator to choose the objects of his bounty. First, the Act permits a spouse to renounce a testator’s will, “whether or not the will contains any provision for the benefit of the surviving spouse.” 755 ILCS 5/M28 (West 2008). Thus, absent a valid prenuptial or postnuptial agreement (see, e.g., Golden v. Golden, 393 Ill. 536 (1946) (wife can effectively bind herself to accept provisions of husband’s will, thereby estopping her from renouncing the will after his death)), the wishes of a surviving spouse can trump a testator’s intentions. Second, a child born to a testator after the making of a will is “entitled to receive the portion of the estate to which he would be entitled if the testator died intestate,” unless provision is made in the will for the child or the will reveals the testator’s intent to disinherit the child. 755 ILCS 5/4 — 10 (West 2008).
The public policy of the state of Illinois as expressed in the Probate Act is, thus, one of broad testamentary freedom, constrained only by the rights granted to a surviving spouse and the need to expressly disinherit a child born after execution of the will if that is the testator’s desire.
Under the Probate Act, Max and Erla had no obligation to make any provision at all for their grandchildren. Indeed, if Max had died intestate, Erla, Michael, and Leila would have shared his estate (755 ILCS 5/2 — 1(a) (West 2008)), and if Erla had died intestate, only Michael and Leila would have taken (755 ILCS 5/2 — 1(b) (West 2008)). Surely, the grandchildren have no greater claim on their grandparents’ testate estates than they would have had on intestate estates.
Similarly, under the Trusts and Trustees Act, “[a] person establishing a trust may specify in the instrument the rights, powers, duties, limitations and immunities applicable to the trustee, beneficiary and others and those provisions where not otherwise contrary to law shall control, notwithstanding this Act.” 760 ILCS 5/3 (West 2008). Thus, the legislature intended that the settlor of a trust have the freedom to direct his bounty as he sees fit, even to the point of giving effect to a provision regarding the rights of beneficiaries that might depart from the standard provisions of the Act, unless “otherwise contrary to law.”
Another legislative enactment that reveals a strong public policy of freedom of testation was the adoption, in 1969, of the Statute Concerning Perpetuities (765 ILCS 305/1 et seq. (West 2008)), for the purpose of modifying the common law rule that a will or trust provision that violated the rule against perpetuities was void ab initio. 765 ILCS 305/2 (West 2008). The statute permits the settlor of a trust to create a “qualified perpetual trust” by including in the instrument a provision that the rule against perpetuities does not apply and by granting certain specified powers to the trustee. 765 ILCS 305/ 3(a —-5), 4(a)(8) (West 2008). The statute also specifies other circumstances under which the rule shall not apply. 765 ILCS 305/4(a)(1) through (a)(7) (West 2008). In addition, the statute adopts a set of rules to be applied when determining whether an interest violates the rule against perpetuities. 765 ILCS 305/4 (c)(1) through (c)(3) (West 2008). With regard to trusts, the statute provides that a trust containing a provision that would violate the rule against perpetuities, as modified by the statute, shall terminate 21 years after the death of the last surviving beneficiary who was living at the beginning of the perpetuities period (765 ILCS 305/5(a)(A) (West 2008)) or else at the end of the 21-year perpetuities period if no beneficiary was living when the period began to run (765 ILCS 305/5(a)(B) (West 2008)). Thus, the trust is not rendered void ab initio, but is merely terminated by operation of law at the conclusion of the perpetuities period.
Also, in 1953, the legislature adopted the Rule in Shelley’s Case Abolishment Act (765 ILCS 345/1 et seq. (West 2008)), to abolish the common law rule that a life estate to A, with a remainder to A’s heirs, shall pass to A in fee simple.
As demonstrated by the Probate Act, the Trusts Act, the Statute Concerning Perpetuities, and the Rule in Shelley’s Case Abolishment Act, the public policy of the state of Illinois protects the ability of an individual to distribute his property, even after his death, as he chooses, with minimal restrictions under state law.
Our case law also demonstrates the existence of a public policy favoring testamentary freedom, reflected in the many cases in which a court strives to discover and to give effect to the intent of a deceased testator or settlor of a trust. See, e.g., Harris Trust & Savings Bank v. Donovan, 145 Ill. 2d 166, 172 (1991) (“The first purpose in construing a trust is to discover the settlor’s intent from the trust as a whole, which the court will effectuate if it is not contrary to public policy”); Harris Trust & Savings Bank v. Beach, 118 Ill. 2d 1, 3 (1987) (“In construing either a trust or a will the challenge is to find the settlor’s or testator’s intent and, provided that the intention is not against public policy, to give it effect”).
The record, via the testimony of Michael and Leila, reveals that Max’s intent in restricting the distribution of his estate was to benefit those descendants who opted to honor and further his commitment to Judaism by marrying within the faith. Max had expressed his concern about the potential extinction of the Jewish people, not only by holocaust, but by gradual dilution as a result of intermarriage with non-Jews. While he was willing to share his bounty with a grandchild whose spouse converted to Judaism, this was apparently as far as he was willing to go.
There is no question that a grandparent in Max’s situation is entirely free during his lifetime to attempt to influence his grandchildren to marry within his family’s religious tradition, even by offering financial incentives to do so. The question is, given our public policy of testamentary freedom, did Max’s beneficiary restriction clause as given effect by Erla’s appointment violate any other public policy of the state of Illinois, thus rendering it void?
Public Policy Regarding Terms Affecting Marriage or Divorce
The contrary law relied upon by the appellate court to invalidate Max’s beneficiary restriction clause is found in three decisions of this court: Ransdell, 172 Ill. 439, Winterland v. Winterland, 389 Ill. 384 (1945), and Estate of Gerbing, 61 Ill. 2d 503 (1975) (which overruled Winter-land in part). The appellate court concluded that the “language and circumstances” of the testamentary provisions in these cases, “which Illinois courts have found to be against public policy, are strikingly similar to the instant case.” 383 Ill. App. 3d at 996. Specifically, the appellate court invoked the “principle that testamentary provisions are invalid if they discourage marriage or encourage divorce.” 383 Ill. App. 3d at 995.
In Ransdell, the testator’s will included provisions for his wife, his son, and his daughter. At the time the will was executed, the son and his wife were separated and cross-suits for divorce were pending. The father’s bequest to the son provided that the property be held in trust, giving him use and income of the land for life, or “until such time as he *** shall become sole and unmarried,” at which time the trustee was to convey title to the land to him in fee simple. Ransdell, 172 Ill. at 440. If the son died childless while still married to the wife, the land was to go to other devisees. Several years after the father’s death, the son, who was still married but living apart from his wife, challenged the provision on public policy grounds. The circuit court granted judgment for the defendants and this court affirmed.
This court acknowledged the long-standing rule that conditions annexed to a gift that have the tendency to induce spouses to divorce or to live separately are void on grounds of public policy. Ransdell, 172 Ill. at 445. However, the testator’s purpose in this case “was simply to secure the gift to his son in the manner which, in his judgment, would render it of the greatest benefit to him in view of the relations then existing between him and his wife” (Ransdell, 172 Ill. at 445), which were strained, to say the least. “Certainly,” this court noted, “it cannot be said that the condition tended to encourage either the separation or the bringing of a divorce suit, both having taken place long prior to the execution of the will.” Ransdell, 172 Ill. at 446.
This court weighed two potentially competing public policies, stating that it was “of the first importance to society that contract and testamentary gifts which are calculated to prevent lawful marriages or to bring about the separation or divorcement of husbands and wives should not be upheld.” Ransdell, 172 Ill. at 446. On the other hand, “it is no less important that persons of sound mind and memory, free from restraint and undue influence, should be allowed to dispose of their property by will, with such limitations and conditions as they believe for the best interest of their donees.” Ransdell, 172 Ill. at 446. Because the testator had not disinherited his son if he remained married, but made one provision for him in case he remained married (a life estate) and a different provision if he divorced (taking title in fee simple), the condition was not contrary to public policy.
Finally, this court distinguished between a condition subsequent (for example, if the will devised property to the beneficiary in trust for Ufe, subject to divestment if he married), and a condition precedent, which directs that upon the fulfillment of the condition, ownership of the property is to vest in the beneficiary. Ransdell, 172 Ill. at 447. The condition subsequent, such as one that would prohibit marriage generally, would be void and the donee would retain the property, unaffected by the violation of the condition. A condition precedent would be given effect, because until the condition was met, the beneficiary’s interest was a mere expectancy. Ransdell, 172 Ill. at 447-48.
The appellate court cited Ransdell for the “general rule that testamentary provisions which act as a restraint upon marriage or which encourage divorce are void as against public policy” and distinguished Ransdell from the present case on the basis that the Ransdells’ marriage was “already in disrepair” at the time the will was executed. 383 Ill. App. 3d at 994. The appellate court noted that subsequent Illinois cases, however, have “reaffirmed the underlying principle.” 383 Ill. App. 3d at 995.
One such case was Winterland, in which the testator created a trust for his wife that, upon her death, was to be distributed equally to their 11 children. However, in a later codicil, the testator directed that the share intended for their son, George, was to be held in trust for him “ ‘so long as he may live or until his present wife shall have died or been separated from him by absolute divorce.’ ” Winterland, 389 Ill. at 385. George predeceased his wife and she and their son challenged the codicil as promoting divorce, contrary to good morals, and against public policy. Winterland, 389 Ill. at 386. This court distinguished Ransdell on the basis that the couple’s separation was “already an accomplished fact and a divorce suit was then pending” at the time the testator made his will. Winterland, 389 Ill. at 387. But where no separation was contemplated, the “natural tendency of the provision” was “to encourage divorce.” For that reason, the provision was void. Winterland, 389 Ill. at 387-88. This court announced that it is “the public policy of this state to safeguard and protect the marriage relation, and this court will hold as contrary to that policy and void any testamentary provision tending to disturb or destroy an existing marriage.” Winterland, 389 Ill. at 387.
This court further found that the codicil established two separate and divisible conditions upon which the trust would be distributed to George. First, George’s life estate was to continue until the death of his wife; second, his life estate would terminate upon their absolute divorce. This court rejected the argument that the life estate itself failed and that title to the property vested in George upon his father’s death. Rather, while the second condition was void, the first condition was not and, thus, could be given effect. Winterland, 389 Ill. at 388.
In Gerbing, this court considered the validity of a provision in a testamentary trust that would have terminated the trust and distributed the corpus to the testator’s son in the event that his wife predeceased him or the couple divorced and remained divorced for two years. Gerbing, 61 Ill. 2d at 505. This court restated the general principle that “a devise or bequest, the tendency of which is to encourage divorce or bring about a separation of husband and wife is against public policy.” Gerbing, 61 Ill. 2d at 507. However, if the “dominant motive of the testator is to provide support in the event of such separation or divorce the condition is valid.” Gerbing, 61 Ill. 2d at 507. Further, unless the couple was separated or a divorce was pending at the time the will was executed, the “exception to the general rule announced in Ransdell” was not applicable. Gerbing, 61 Ill. 2d at 508. This court found the provision void, but declined to sever the two conditions, as it had done in Winterland. Finding that it was the testator’s general intent to benefit her son and that she would have preferred that he take the corpus of the trust, even if he remained married, rather than have him take nothing, this court found the entire provision void, overruling Winterland to the extent it held otherwise. Gerbing, 61 Ill. 2d at 512.
In the present case, the appellate court found the “language and circumstances” of these three cases “strikingly similar” to the present case and saw “no reason to depart from this well-established principle” of these cases. 383 Ill. App. 3d at 996. We disagree with the appellate court’s conclusion regarding the similarity of the present case to the cited cases. The beneficiary restriction clause as given effect by Erla’s distribution scheme does not implicate the principle that trust provisions that encourage divorce violate public policy. That is, the present case does not involve a testamentary or trust provision that is “capable of exerting *** a disruptive influence upon an otherwise normally harmonious marriage” by causing the beneficiary to choose between his or her spouse and the distribution. Gerbing, 61 Ill. 2d at 508. The challenged provision in the present case involves the decision to marry, not an incentive to divorce. This court has considered the validity of restrictions affecting marriage in cases going back as far as 1857.
In Shackelford v. Hall, 19 Ill. 211 (1857), the testator, Hall, left his estate to his wife for life or until she remarried, with the remainder to his four children, subject to the condition that they not marry before the age of 21. Any child who married before his or her twenty-first birthday was to receive one dollar only. The only daughter, Eliza, married four months before her twenty-first birthday, with the approval of her eldest brother, the executor of their father’s estate. This court described the provision as a “devise with a condition subsequent,” because the remainder interest vested in the four children immediately upon the death of the testator, “subject to be defeated by their marriage before they should attain that age.” Shackelford, 19 Ill. at 213. This court noted that:
“whoever will take the trouble to examine this branch of the law attentively, will find that the testator may impose reasonable and prudent restraints upon the marriage of the objects of his bounty, by means of conditions precedent, or subsequent, or by limitations, while he may not, with one single exception, impose perpetual celibacy upon the objects of his bounty, by means of conditions subsequent or limitations. That exception is in the case of a husband in making bequests or legacies to his own wife. He may rightfully impose the condition of forfeiture upon her subsequent marriage.” Shackelford, 19 Ill. at 214-15. As for other conditions affecting marriage that might
be imposed by a testator, this court said that:
“[a]n examination of the subject, will show that the courts have very rarely held such condition void, although it might appear harsh, arbitrary and unreasonable, so as it did not absolutely prohibit the marriage of the party, within the period wherein issue of the marriage might be expected. It is enough for our present purpose, and we will go no further now, for it is not necessary, that it has been nowhere held, or pretended, that an absolute prohibition of marriage till twenty-one years of age is not reasonable and lawful, and must not be upheld, as a good condition, the violation of which may defeat a vested estate. The condition, then, annexed to this devise, was proper, reasonable and lawful, and its violation must be held to have forfeited the estate devised, unless it can be saved by some other equally well settled principle of law.” Shackelford, 19 Ill. at 215.
Further:
“The facts of the case show, that all of the devisees of the estate in remainder, now in controversy, were the children and heirs at law of the testator, and as such heirs at law, had expectations of this estate. In the absence of the will, each would have been entitled to his or her respective proportions of it, according to our statute of descent. When such is the case, the condition subsequent, the breach of which shall divest the estate which has become vested in the devisee by the will, must be shown to have been brought home to the knowledge of the devisee, before the breach, in order to mark the forfeiture.” (Emphases added.) Shackelford, 19 Ill. at 215-16.
In the end, this court found that the marriage of Hall’s daughter prior to her twenty-first birthday did not divest her of the remainder interest conveyed to her upon her father’s death. The basis for this decision was not that her father’s partial restraint upon marriage was invalid, but that her remainder interest vested upon his death and could not be divested by a subsequent act on her part, absent a showing that she had notice of the condition subsequent. Other factors supporting this result were that she would have been one of her father’s heirs at law should he have died intestate and that her brother, the executor, had unclean hands:
“And this rule is in harmony with the general principles of law, which always lean hard against a forfeiture of estates once vested, and that it will not allow such forfeiture, where there has been no laches or misconduct. In the case before us, we must assume that the defendant did not know of the existence of the will, and much less of the condition which it contained, that she should not marry till she was twenty-one years of age, under the penalty of forfeiting her interest in her father’s estate. In ignorance of the will, she supposed she was entitled to take as heir without any condition. When we look at this case as it is presented by the record, we see it would be a monstrous piece of injustice to enforce this forfeiture against her. Here was her elder brother, who was an executor named in the will, knowing of the condition of forfeiture, had an interest in keeping it from her, that she might, by doing the prohibited act, incur the forfeiture, that her portion might go to himself and the other heirs of the testator. Under the influence of this direct interest, he suffers her to go on in ignorance of the will, and marry only four months before she attained the age of twenty-one years, and now he comes forward and claims the benefit of the forfeiture, and insists upon depriving her of the portion devised to her by the will. To sustain this claim, would be to offer a premium for the commission of the most heartless frauds. *** We have not the least doubt that, upon the soundest principles of law and morality, she must take the estate devised, discharged of the condition.” (Emphasis added.) Shackelford, 19 Ill. at 217-18.
In the present case, Michael argues that the beneficiary restriction clause is a similar “reasonable and prudent restraint” that does not operate as a complete restraint upon marriage. Rather, the clause disqualifies from receipt of a share of the trust assets any grandchild who has chosen to marry outside the religious tradition their grandparents valued so highly.
More importantly, we note that, unlike Eliza Hall, the grandchildren did not receive a vested interest in the trust upon Max’s death. By creating a power of appointment in Erla, Max created a situation in which the interests of the grandchildren were contingent on whether and in what manner she would exercise her lifetime and testamentary powers of appointment. Thus, the grandchildren had a mere expectancy that they might receive some portion of the remainder at the conclusion of Erla’s life estate. No one had a vested interest in the remainder of the trust assets until Erla’s death resolved all contingencies. Further, unlike Hall’s daughter, the grandchildren in the present case were not Max’s or Erla’s heirs at law. Finally, while the record is unclear whether any or all of the grandchildren were aware of the existence of the beneficiary restriction clause, because they had no vested interest to protect, they were not entitled to notice of the condition.
More recently, the appellate court upheld the validity of a testamentary provision regarding the marriage of the intended legatee. In In re Estate of Gehrt, 134 Ill. App. 3d 308 (1985), the testator, Forrest Gehrt, originally left a portion of his estate to six named individuals who were the children of Edna Bocock, apparently his deceased sister. Upon the death of one of these individuals, Harold Bocock, he executed a codicil leaving the portion originally intended for Harold to his widow, Betty, provided that at the time of Forrest’s death, she remained unmarried. If, at the time of Forrest’s death, it was determined that Betty had remarried, the share was to go to Harold’s five siblings. Estate of Gehrt, 134 Ill. App. 3d at 309.
Betty remarried and, upon Forrest’s death, sought Harold’s share of the estate. The parties agreed that the condition operated as a condition precedent. Betty argued that it constituted an invalid restraint on marriage and asked that it be declared void as against public policy. The executor argued that the condition did not operate as a restraint because the interest “either vests or not at the date of the death of the testator depending on [Betty’s] marital status at the time, not at some later time.” Estate of Gehrt, 134 Ill. App. 3d at 310.
The appellate court invoked a rule of reasonableness, quoting a case from the state of Louisiana in support of such a rule:
“ ‘[(Conceding, without deciding, that a legacy conditioned upon the legatee remaining unmarried is against the public policy of this State, it is apt to observe here that the provision under consideration is not one forbidding the donee to marry during her lifetime or even for a fixed period of time, nor one that directs the legacy shall lapse in case the legatee should marry in the future, but rather one that is conditioned upon her status at the time of the testator’s death. Certainly, such a provision is not against good morals, and we know of no law prohibiting the same.’ ” Estate of Gehrt, 134 Ill. App. 3d at 311, quoting Succession of Ruxton, 226 La. 1088, 1091, 78 So. 2d 183, 184 (1955).
Applying this principle to the Gehrt estate, the appellate court noted the well-established principle that a will speaks as of the date of death of the testator. Estate of Gehrt, 134 Ill. App. 3d at 310. Thus, the court observed:
“[T]he testator, Forrest L. Gehrt, could have, for any reason, changed his codicil at any time prior to his death. He could have, at the time of plaintiffs remarriage, immediately executed another codicil cancelling the gift to the plaintiff, and could have given that portion of property to others. He can validly accomplish the same result by using the language that he did in the codicil in this case.” Estate of Gehrt, 134 Ill. App. 3d at 311.
The appellate court then quoted this court’s opinion in Ransdell'.
“While it is of the first importance to society that contract and testamentary gifts which are calculated to prevent lawful marriages or to bring about the separation or divorcement of husbands and wives should not be upheld, it is no less important that persons of sound mind and memory, free from restraint and undue influence, should be allowed to dispose of their property by will, with such limitations and conditions as they believe for the best interest of their donees.” Ransdell, 172 Ill. at 446.
We conclude, reading Ransdell, Shakelford, Gerbing, and Gehrt together, that no interest vested in the Feinberg’s grandchildren at the time of Max’s death because the terms of his testamentary trust were subject to change until Erla’s death. Because they had no vested interest that could be divested by their noncompliance with the condition precedent, they were not entitled to notice of the existence of the beneficiary restriction clause. Further, because they were not the Feinbergs’ heirs at law, the grandchildren had, at most, a mere expectancy that failed to materialize for four of them when, at the time of Erla’s death, they did not meet the condition established by Max.
Applicability of Restatement (Third) of Trusts
In reaching its decision, the appellate court also relied on section 29 of the Restatement (Third) of Trusts, and the explanatory notes and comments thereto. 383 Ill. App. 3d at 997.
Since the Restatement (Second) of Trusts was adopted in 1959, this court has, on several occasions, cited various sections with approval. See, e.g., Eychaner v. Gross, 202 Ill. 2d 228, 255 (2002) (requirements for creation of a trust). We have not yet had reason to consider whether any section of the Restatement (Third) of Trusts, which was adopted in 2003, is an accurate expression of Illinois law and we need not do so in this case.
The validity of a trust provision is not at issue, as the distribution provision of Max’s trust was revoked when Erla exercised her power of appointment. Her distribution scheme was in the nature of a testamentary provision, which operated at the time of her death to determine who would be entitled to a $250,000 distribution.
The appellate court mistakenly compared the present case to an illustration accompanying Comment j to section 29 of the Restatement (Third) of Trusts. 383 Ill. App. 3d at 997. The illustration concerns a trust created by an aunt to benefit her nephew, who was to receive discretionary payments until age 18, and all income and discretionary payments until age 30, at which time he would receive an outright distribution of all trust property. However, all of his rights under the trust would end if, before the trust terminated on his thirtieth birthday, he married “a person who is not of R Religion.” If he violated this condition, the remainder of the trust would be given to a college. The drafters of the Restatement called this an “invalid restraint on marriage,” and stated that the invalid condition and the gift over to the college should not be given effect. Restatement (Third) of Trusts §29, Explanatory Notes, Comment j, Illustration 3, at 62-64 (2003).
This illustration is similar to Max’s original trust provision. Under his plan, the grandchildren who were not “deemed deceased” at the time of Erla’s death would receive distributions from the trust for life, subject to termination if they should violate the marriage restriction. Erla’s scheme, however, does not operate prospectively to encourage the grandchildren to make certain choices regarding marriage. It operated on the date of her death to determine which, if any, of the grandchildren qualified for distribution on that date. The condition was either met or it was not met. There was nothing any of the grandchildren could have done at that time to make themselves eligible or ineligible for the distribution.
As this court noted in Ransdell, a condition precedent, even if a “complete restraint” on marriage, “will, if broken, be operative and prevent the devise from taking effect.” However, “[w]hen the condition is subsequent and void it is entirely inoperative, and the donee retains the property unaffected by its breach.” Ransdell, 172 Ill. at 447, quoting 2 Pomeroy, Equity Jurisprudence §933B (1881).
Max’s will and trust created no vested interests in the children or grandchildren because Erla retained a power of appointment until her death. No vested interests were created in 1997 by Erla’s exercise of her power of appointment. Her actions created a mere expectancy, contingent on her dying without further amending the distribution scheme. Because no interest vested in any of the grandchildren until Erla’s death, her appointment created a condition precedent. As we noted in Ransdell, under these circumstances, even a complete restraint on marriage (i.e., distribution only to unmarried grandchildren) would be operative.
Thus, this is not a case in which a donee, like the nephew in the illustration, will retain benefits under a trust only so long as he continues to comply with the wishes of a deceased donor. As such, there is no “dead hand” control or attempt to control the future conduct of the potential beneficiaries. Whatever the effect of Max’s original trust provision might have been, Erla did not impose a condition intended to control future decisions of their grandchildren regarding marriage or the practice of Judaism; rather, she made a bequest to reward, at the time of her death, those grandchildren whose lives most closely embraced the values she and Max cherished.
The trial court and the appellate court erred by finding a violation of public policy in this case. While the beneficiary restriction clause, when given effect via Erla’s distribution provision, has resulted in family strife, it is not “so capable of producing harm that its enforcement would be contrary to the public interest.” Kleinwort Benson, 181 Ill. 2d at 226.
Other Issues
The several other arguments made by Michele do not alter our conclusion.
First, Michele argues that even a partial restriction on marriage is void unless the “dominant motive for the restriction was to provide a benefit to the donee.” She cites Ransdell and Gerbing, neither of which involved a restraint upon marriage, in support of this assertion. We concluded above, however, that the beneficiary restriction clause does not operate as a restriction on marriage because it operated only upon Erla’s death to determine which grandchildren, if any, would share in the proceeds of the trust. Until that time, none of Max’s heirs had a vested interest in the proceeds of the trust, over which Erla held a limited power of appointment. Michele’s choices regarding when to marry and whom to marry were entirely unrestricted, even though, as it turns out, those choices did have consequences for her.
In any event, the cases cited do not condition the validity of a condition precedent regarding marriage solely on the settlor’s “dominant motive” in creating it. The provision that this court upheld in Ransdell gave the testator’s son a life estate in certain property, but provided that title would be conveyed to him upon the death of his wife or their subsequent divorce. Although a conditional gift that has the tendency to induce divorce is generally considered void, this court noted that “certain facts and circumstances” in the particular case could save such a condition. Specifically, the testator’s “purpose was simply to secure the gift to his son in the manner which, in his judgment, would render it of the greatest benefit to him in view of the relations then existing between him and his wife,” including a separation of several years’ duration and a pending divorce suit. Ransdell, 172 Ill. 2d at 445. Thus, while Ransdell makes the testator’s intent to provide a benefit to the donee a proper consideration, it does not elevate it to an absolute requirement. Hence, there is nothing illegitimate about a testator’s preference for supporting a particular cause, value, or personal interest over the interests of potential beneficiaries, so long as the condition stated in the will or trust does not, at the relevant time, violate public policy.
The result in Gerbing differed because the provision that tended to induce divorce, in violation of public policy, was not within the “exception to the general rule announced in Ransdell.” Gerbing, 61 Ill. 2d at 508. That is, the couple was not already separated and a divorce suit was not pending. After noting that a condition that might tend to induce divorce is valid “if the dominant motive of the testator is to provide support in the event of such separation or divorce” (Gerbing, 61 Ill. 2d at 507, citing Restatement of Property §427 (1944)), this court concluded that the testator’s intent was to “deprive her son of the ownership of this property as long as he remained married to his wife,” not to provide for his support in the event of divorce. Gerbing, 61 Ill. 2d at 508-09. Therefore, this court found the provision to be a “void condition precedent to the vesting of title to the property” in the son. Gerbing, 61 Ill. 2d at 510.
Gerbing is readily distinguishable from the present case. The son in that case had a vested interest in the trust, from which he was receiving income for life. Title to the trust property was withheld from him, however, so long as he remained married to the wife of whom his late mother had disapproved. The provision had a tendency to exert “a disruptive influence upon an otherwise normally harmonious marriage” and was, therefore, void as against public policy. Gerbing, 61 Ill. 2d at 508. It could not be saved by reliance on the so-called Ransdell “exception.”
Michele is mistaken when she asserts that a provision that tends to promote divorce or to restrain marriage is valid only if its dominant purpose is to benefit the potential donee. While such a dominant purpose may save such a provision, we have never stated that it is the only consideration. In any event, Michele had no interest in the trust whatsoever. Her taking title to trust property was not subject to a condition subsequent nor did she have an interest that would be divested by her marriage to a non-Jew. Her grandfather’s purpose is, therefore, not implicated.
Second, Michele argues that the beneficiary restriction clause discourages lawful marriage and interferes with the fundamental right to marry, which is protected by the constitution. She also invokes the constitution in support of her assertion that issues of race, religion, and marriage have special status because of their constitutional dimensions, particularly in light of the constitutional values of personal autonomy and privacy.
Because a testator or the settlor of a trust is not a state actor, there are no constitutional dimensions to his choice of beneficiaries. Equal protection does not require that all children be treated equally; due process does not require notice of conditions precedent to potential beneficiaries; and the free exercise clause does not require a grandparent to treat grandchildren who reject his religious beliefs and customs in the same manner as he treats those who conform to his traditions.
Thus, Michele’s reliance on Shelley v. Kraemer, 334 U.S. 1, 92 L. Ed. 1161, 68 S. Ct. 836 (1948), is entirely misplaced. In Shelley, the Supreme Court held that the use of the state’s judicial process to obtain enforcement of a racially restrictive covenant was state action, violating the equal protection clause of the fourteenth amendment. Shelley, 334 U.S. at 19, 92 L. Ed. at 1183, 68 S. Ct. at 845. This court, however, has been reluctant to base a finding of state action “on the mere fact that a state court is the forum for the dispute.” In re Adoption of K.L.P., 198 Ill. 2d 448, 465 (2002) (citing cases). Indeed, Shelley has been widely criticized for a finding of state action that was not “ ‘supported by any reasoning which would suggest that “state action” is a meaningful requirement rather than a nearly empty or at least extraordinarily malleable formality.’ ” Adoption of K.L.P., 198 Ill. 2d at 465, quoting L. Tribe, American Constitutional Law 1698 (2d ed. 1988).
Third, Michele argues that the beneficiary restriction clause is capable of exerting an ongoing “disruptive influence” upon marriage and is, therefore, void. She is mistaken. The provision cannot “disrupt” an existing marriage because once the beneficiary determination was made at the time of Erla’s death, it created no incentive to divorce.
Finally, it has been suggested that Michael and Leila have litigated this matter rather than concede to Michele’s demands because they wish to deprive the grandchildren of their inheritance. The grandchildren, however, are not the heirs at law of Max and Erla and had no expectancy of an inheritance, so long as their parents were living, even if Max and Erla had died intestate. In addition, Michael and Leila are the coexecutors of their parents’ estates and, as such, are duty-bound to defend their parents’ estate plans. Hurd v. Reed, 260 Ill. 154, 160 (1913) (“It is the duty of an executor to defend the will”), citing Pingree v. Jones, 80 Ill. 177, 181 (1875) (executor is “bound, on every principle of honor, justice and right” to defend the will, he “owes this, at least, to the memory of the dead who placed this confidence in him”). Although those plans might be offensive to individual family members or to outside observers, Max and Erla were free to distribute their bounty as they saw fit and to favor grandchildren of whose life choices they approved over other grandchildren who made choices of which they disapproved, so long as they did not convey a vested interest that was subject to divestment by a condition subsequent that tended to unreasonably restrict marriage or encourage divorce.
CONCLUSION
It is impossible to determine whether Erla’s distribution plan was the product of her own wisdom, good legal advice, or mere fortuity. In any case, her direction that $250,000 of the assets of Trust B be distributed upon her death to each of the then-living grandchildren of Max who were not “deemed deceased” under the beneficiary restriction clause of Max’s trust revoked his plan for prospective application of the clause via a lifetime trust. Because no grandchild had a vested interest in the trust assets and because the distribution plan adopted by Erla has no prospective application, we hold that the beneficiary restriction clause does not violate public policy.
Therefore, we reverse the judgment of the appellate court and remand to the circuit court for further proceedings.
Reversed and remanded.