Commentary on Reid Kress Weisbord and David Horton, Boilerplate and Default Rules in Wills Law: An Empirical Analysis

I.     Introduction

Reid Weisbord and David Horton have undertaken an incredibly important empirical study in an area of law that suffers from a large gap in our understanding of how people actually choose to leave their property at their death and the drafting traps that can easily lead to litigation. The study is also important for illustrating how the lawyers we teach in Trusts and Estates need to be more careful in drafting the various documents to manifest their clients’ testamentary intent. In particular, Weisbord and Horton studied 230 recently probated wills in Sussex County, New Jersey and discovered that the use of boilerplate language in certain key provisions can have troubling consequences for the testators and their beneficiaries.1 In examining these wills, Weisbord and Horton discovered that there was a consistent use of boilerplate language in four provisions: (1) the just debts clause; (2) the tax apportionment clause; (3) the method of representation for class gifts; and (4) the survivorship language necessary to defeat application of anti-lapse statutes.2 After examining the implications of the use of boilerplate in these four provisions, Weisbord and Horton argue that probate laws should make the default rules “stickier” so that it is more difficult for testators, or their attorneys, to draft around the defaults.3 This was especially true in situations where bad drafting often led to litigation.4

I found their data and arguments compelling and vital to a comprehensive understanding of how probate laws, probate lawyers, testators, and beneficiaries all interact in the effectuation of testamentary documents. In my brief commentary, I will discuss each of these four provisions (somewhat out of order) and offer additional thoughts and observations that may inform future research in this area.

II.     Just Debts

The first boilerplate provision is the just debts clause. This is a clause lawyers inserted into wills in 96% of cases, despite the fact that it is legally unnecessary and often offers absolutely no guidance when there actually are ambiguities as to which debts a decedent intended to have paid.5 As Weisbord and Horton conclude, “most lawyers clung to generic ‘just debts’ language, even though it is pure risk with no corresponding benefit.”6 Because the executor is required to pay a decedent’s debts regardless of the presence of a just debts clause, a well-phrased and intentional clause could be used to clarify which debts should take priority if a decedent has more debts than are assets in his estate, and it could help with the second provision, the testator’s intentions regarding payment of estate taxes. In most cases, however, the just debts clause offered no guidance where it was really needed, and at times it even spurred litigation.7

There are likely to be numerous reasons lawyers include a bland, unhelpful “just debts” provision. One is custom.8 Like reciting that the testator is of sound mind and over age 18, the “just debts” clause sets the mood for the rest of the will. It also indicates that the testator is morally responsible, and it expressly indicates the intention of paying off any debts he or she has assumed during life. There is something worthy about acknowledging one’s debts, for it also indicates that the testator understands that the debts come before any bequests, and the testator is warning the beneficiaries that they stand second to creditors.9

However, the fact that lawyers tend to use standard boilerplate that does not elucidate what priority should be given to certain debts over others, that often does not clarify how taxes are to be apportioned, and that can muddy an otherwise unambiguous will is problematic at best and malpractice at worst. Weisbord and Horton suggest that changes to probate law should make the default of non-exoneration stickier, so that any debts attached to certain property (like a mortgage, a car loan, and the like) would not be paid by the estate and the beneficiary would take the property subject to the debt unless the testator was absolutely clear that the debt would be paid by the estate.10 In other words, Weisbord and Horton suggest that a boilerplate just debts clause be given no meaning, and that the default rule of non-exoneration would operate regardless. I very much liked the suggestion that boilerplate just debts clauses be given no meaning with regard to exoneration unless the testator expressly identifies the debts to be paid. One problem with such a suggestion, however, is that very often a will is written years before the decedent’s death,11 and the decedent may not know what debts he or she will have and how those debts may affect the overall estate plan. An alternative that might make more sense would be that, rather than expressly listing the items of property which are to pass debt-free, the testator identify which beneficiaries should be given property debt-free. Because most testators use their residuary clause to benefit their primary beneficiaries12—like spouses and children—identifying up front which beneficiaries are privileged means the testator can identify which class of gifts will ultimately end up paying for any debts that are exonerated without having to list the property, list the debts, or even know what property or what debts the testator will have at death.

I think if we were to ask testators explicit questions about exoneration, we would be likely to find that their desires are more nuanced than simply whether they want to exonerate or not. Making it harder for them to exonerate does not address the likelihood that, if asked explicitly, their answers would likely be yes for some and no for others, or yes under certain circumstances and no under others. In other words, only if they knew how much debt they were likely to have, how much wealth would be available, and who their surviving beneficiaries are, could they answer that question relatively accurately. To add a certain amount of nuance without requiring explicit language in a will, the law could be changed to distinguish between privileged and non-privileged beneficiaries, just as the Uniform Probate Code does with anti-lapse.13 The law could provide that spouses and perhaps children receive property free of debts, but others do not, unless expressly stated otherwise.

Thus, rather than the current plan, which usually ends up requiring that the residuary beneficiaries pay any debts for property being specifically devised,14 the testator could instead state that debts should be paid on all gifts to privileged beneficiaries, or should not be paid on gifts to beneficiaries who are not related by blood or marriage. This is just one suggestion that might help testators who, at the time of drafting their will, assume their estates will have sufficient assets to pay off all debts and thus who make generous bequests and perhaps even include a direction in their just debts clause to pay all debts, but whose estates are less flush after end-of-life medical expenses are paid.

Of course, Weisbord and Horton’s proposal to make non-exoneration stickier has the benefit of continuing to operate as a default rule or, as they note, a majoritarian principle.15 Thus, Weisbord and Horton’s proposal is a change in law that will hopefully spur a change in practice. My recommendation is the change in practice. Asking testators to identify privileged classes of beneficiaries requires that they be explicit, which is precisely what they would need to do to override sticky defaults, but in a way that they may be better able to accommodate the information gap caused by the time lag. Making the defaults stickier will hopefully spur more intentioned drafting, but if that fails, we could adopt a default that attempts to deal in a more nuanced way with the information gap between will execution and death.

As with any default rule, there will be situations where exoneration or non-exoneration fails to correspond to testamentary intent. But perhaps a change in the law will spur estate planning lawyers into asking their clients directly about whether debts should be paid. This is a particularly important issue because most people will acquire significant health-care debts for their end-of-life care and are unlikely to properly anticipate the effect of those debts on their estates. Getting estate-planning lawyers to be more intentional in thinking about end-of-life debts would be one benefit of creating a sticky default that distinguishes between certain classes of beneficiaries.

III.     Tax Apportionment

The tax apportionment boilerplate is perhaps even more problematic than the just debts clause because standard boilerplate that instructs the executor to pay the estate taxes from the residuary can have tremendous tax consequences. As the famous In re Estate of Kuralt case shows, boilerplate can be more than simply superfluous, and indeed quite consequential.16 However, if this were simply a problem with estates that run into federal estate taxes, I frankly would be less concerned. Those estates are well over $5 million and comprise less than 0.2% of all estates.17 Such an estate is almost guaranteed to have had competent estate planning lawyers drafting the various succession documents and their failure to properly attribute the taxes should comprise malpractice. And is quite likely that malpractice actions will have better results in getting lawyers to pay attention to the problems with boilerplate than making the apportionment default stickier. Nonetheless, with state estate and inheritance taxes kicking in at a much lower threshold, the malpractice of the smaller estates’ attorneys can be particularly troubling. Thus, this is an issue that faces far more estates than one might initially surmise.

The question I have, however, is whether making the clause stickier will actually lead to better results. Particularly, will lawyers drafting wills actually anticipate the problems that can arise with mal-apportionment of taxes if they draft thinking that, by paying the taxes out of the provisions for the spouse, the decedent will not have any taxes? In other words, will estates lawyers actually get it right if they have to be more intentional when they opt out of the default, or will they simply use more explicit language to opt out thinking they are doing the right thing, when in fact they are not?

I would assume that every estates lawyer is now familiar with the Kuralt case and will think twice before including boilerplate tax apportionment language to pay from the residuary. But assuming that is not the case, what is the best way to get that lawyer to do it right? As with the just debts clause, it may make more sense for the lawyer drafting a will to privilege certain beneficiaries rather than assume he or she knows in advance what the tax consequences of the estate plan will be or that assigning taxes to the residuary will result in the optimal tax outcome. Of course, the point of the default rule is that apportionment to all property makes more sense than the residuary paying when the latter is going to result in more taxes. But it is the opposite when paying from the residuary would result in fewer taxes. But again, assuming a decedent and her lawyer are talking about taxes in the client meetings, my guess is that more decedents would prefer to privilege certain beneficiaries rather than identify a particular pot of assets from which to pay tax debts—especially when testators are unlikely to know exactly how many assets and in what form they will take many years in the future when the testator passes into that dark night. Thus, the identity of privileged beneficiaries is likely to be more consistent over time than the nature and identity of property. Moreover, privileged beneficiaries are more likely to remain constant than the changing tax code.

Both of these boilerplate issues force us to ask whether a testator who is expressly asked about exoneration or tax apportionment at the time of will drafting, presumably with good legal advice, is more likely to plan for the optimal outcome than a default rule, sticky or otherwise. In other words, assuming there is an optimal exoneration or tax apportionment outcome for every estate, and that testators are asked about each and identify their preferences at the time they draft their wills, are they more likely to reach their optimal outcome if they are asked explicitly and their preferences are incorporated into their wills, or are the odds just as likely that a default rule will result in the optimal outcome years later when they die and the documents take effect? Given the uncertainty of assets and priorities at the moment of death, a point sometimes far in the future, it may be that the defaults are likely to result in better outcomes than expressing some intention that may exist when the will is drafted but not when the will becomes effective. If the uncertainty of assets and priorities in the future actually results in non-optimal outcomes for those decedents who intentionally plan, then it makes sense to calibrate the default more accurately rather than push lawyers into drafting more intentionally. And again, I would argue that testators are more likely to remain fixed on who their privileged beneficiaries are than on the identity of their assets at death. If that is the case—and intentional planning is either ineffective, ill-advised, or simply boilerplate—then making the defaults stickier and more nuanced by providing privileged classes of beneficiaries might be even more in line with testamentary intent than simply making the defaults stickier.

IV.     Representation Language

The third boilerplate provision Weisbord and Horton identify is the misuse of representation language.18 This is one of those puzzles that always stuns me. Although we spend one to two days in an average trusts and estates class on the different representation schemes, it is not a very difficult concept. If students learn anything, it should be that the different representational schemes clearly apply only to gifts to multi-generational classes, not gifts to individuals. And yet nearly 20% of Weisbord and Horton’s wills used the term completely inaccurately.19 I frankly do not know what to think about this, and it is a problem that I have encountered routinely in teaching future interests and trust drafting, where the cases show testators and their lawyers have absolutely no idea what the terms per stirpes and per capita mean. On the one hand, I would not be too concerned about it when we realize that getting it wrong may simply mean that grandchildren or lineal descendants may take uneven shares when they were supposed to get even shares, or vice versa; they still take shares nonetheless. What makes the inaccuracy problematic, however, is when the term is misapplied to a gift to an individual, and then a court interprets the term to indicate intent to apply an anti-lapse statute in the absence of a qualifying relationship. When ignorance of a term’s meaning, which is apparent from the way these particular terms are misused, has distributive consequences such that some unintended people may take a gift and the intended beneficiaries do not, the problem needs to be addressed. If it were just a matter of the percentage of the bequest that each intended beneficiary receives, the problem would not be so bad. Yet I have to ask whether making the default stickier will ultimately help. To the extent the anti-lapse boilerplate issue is the real problem, then perhaps there is nothing one can do about the utter ignorance of testators and attorneys who misuse the representational language. We can focus on fixing the anti-lapse issue and hope that that 20% number falls as a result.

But I am not entirely happy with that. Part of the problem stems from the fact that we actually have different representational schemes, and not just per stirpes and per capita. Some states have a hybrid form that is per capita at certain levels and per stirpes at others.20 And of course others use a degree-of-relationship scheme rather than the more standard parentelic scheme.21 One obvious solution would be to standardize the representational scheme across the states, which is what the Uniform Probate Code is attempting to do.22 But unfortunately, the per stirpes model has been difficult to displace, in large part because beneficiaries think it is more fair. There is something intrinsically equitable about dividing class gifts initially at the first level of children.

One option would be to simply provide that all class gifts will conform to a particular representational scheme and not allow deviation except by individual gifts directly to the beneficiaries. Another would be to ignore all references to per stirpes or per capita unless the term is used correctly, thus not allowing it to trigger any anti-lapse provision. Another would be to abolish both and require that testators expressly identify how they want class gifts to be distributed. None of these options is likely to be satisfactory, but we can analogize the situation to the fate of adopted children and the gradual displacement of the stranger to the adoption rule.

In all states at one time or other, adopted children were not treated as natural children until state statutes were modified to treat them the same.23 There was also the ubiquitous stranger to the adoption rule that did not include adopted children in class gifts by donors who were not the adoptive parent.24 Yet today, it is very unusual to see a provision in a will that excludes an adopted child.25 The new rule eventually took hold and operates nearly universally.26 Of course, until the per capita scheme was adopted, per stirpes had a stranglehold in most probate codes, which meant that at least there was not a lot of unintended deviation even if the term was used for individual gifts.27 So the question is whether we should simply mandate a particular rule, like the adoption rule, and over time assume it will become nearly universally followed; or do we try to encourage donors to pick and choose, assuming that the smorgasbord option will result in greater satisfaction of testamentary intent? I do not know what the best solution would be, but certainly untying misapplication of representation language from anti-lapse would be a start in the right direction.

V.     Survivorship Language and Anti-Lapse

The fourth provision is the survivorship language used to indicate intent not to apply anti-lapse protections to certain gifts. Although the Uniform Probate Code has tried to make this default rule stickier, most states have opted against doing so.28 That suggests that legislators, and the trusts and estates bar that advises them, have some intentional basis for rejecting the basic rule that mere survivorship language alone does not indicate intent not to apply the anti-lapse protections. This dispute reminds me of many age-old property law problems, where the courts have to grapple with silence and impose meaning when the documents are either silent or ambiguous. In this case, where a testator fails to state specifically where each bequest will go in case any beneficiary predeceases her, there is no reason to assume silence means an unreasonable outcome, whether that outcome is lapse or no lapse. Lawyers certainly need to discuss with their clients the implications of lapse and anti-lapse protections and need to provide for every alternative if they want to avoid application of anti-lapse.

Perhaps the problem is with the lapse rule itself. If there were no lapse rule, there would be no need for anti-lapse statutes. If all gifts went to lineal descendants of pre-deceased beneficiaries, there would be no need to draft complicated survivorship options. If there were not complicated rules about anti-lapse application, then the misuse of representational language would have little effect.

VI.     Conclusion

Thus, while I agree wholeheartedly with Weisbord and Horton about the problems of these four boilerplate provisions, I wonder if simply making the defaults stickier is adequate. I suspect it would be adequate for the tax apportionment boilerplate because most lawyers drafting wills and trusts for high-value estates are likely to understand the consequences of getting it wrong. I do not think the problems we see with representational language, however, can be solved as easily. The misuse of the representational language clearly shows either ignorance on the part of a certain group of lawyers, or that a homemade will is being used. In the case of the latter, courts are simply going to have to grapple with the meaning of the mistaken use of representational language because making the default stickier is not likely to solve the problem. And with the ignorant lawyer, the solution seems to be better education rather than a stickier default. The just debts and anti-lapse problems lie somewhere in between. Clearly there is ignorance or laziness on the part of some lawyers, and in many cases the consequences are not terrible, so there is no incentive for them to improve. The question is whether improvement will come with better education, more severe consequences like malpractice exposure, or stickier defaults. The stickier default option is probably the least costly, but it also raises the moral hazard of not punishing those who continue to act irresponsibly. I wonder if that is the direction we want to pursue, given the finality of the decisions in this area of law. If higher taxes are paid, or an anti-lapse provision is, or is not, applied, there can be profound consequences for the living. Estates can be ruined, family harmony destroyed, and lawyers can lose their reputations. I have fewer qualms about boilerplate that has fewer consequences than boilerplate that results in bequests passing to the wrong people, or higher taxes being paid, or litigation becoming necessary. The harm the Kuralt case imposed on the decedent’s wife and children is inestimable; it cannot be made up simply by winning damages in a malpractice action.

Weisbord and Horton’s article is incredibly valuable because it shows not just the understandable places where ambiguities lie and traps that can catch the unwary estate planning lawyer. Rather, it shows the inexcusable errors, the ambiguities that are inserted because of laziness or ignorance, and the barriers to attempts to provide better default rules. In an ideal world, would we change the rules, make the defaults stickier, jettison the rules, or mandate no deviation? Those are questions none of us can answer, but Weisbord and Horton’s article should get us thinking about them.

  1. [1]. Reid Kress Weisbord & David Horton, Boilerplate and Default Rules in Wills Law: An Empirical Analysis, 103 Iowa L. Rev. 663, 668 (2018).

  2. [2]. Id. at 676–82.

  3. [3]. Id.

  4. [4]. See id. at 688.

  5. [5]. Id. at 698.

  6. [6]. Id.

  7. [7]. See id. at 680, 698 (noting that 96% of wills had a “just debts” clause and 71% were generic, offering no guidance).

  8. [8]. Id. at 679.

  9. [9]. Id. at 695–98.

  10. [10]. Id. at 705.

  11. [11]. In my own research into hundreds of wills probated in Alachua County, Florida in 2013, I found that nearly 60% of wills were executed more than five years before death. Research on file with author.

  12. [12]. See Marc S. Bekerman, Navigating a Non-Tax Estate Planning Maze with Some Tax Considerations, 27 Prob. & Prop. 9, 13 (2013) (“Apportioning estate taxes against residuary bequests will reduce the amounts passing to those beneficiaries who are usually the beneficiaries that the client intends to benefit the most.”).

  13. [13]. See Unif. Prob. Code § 2-603(b) (2010) (applying only to devisees who are either
    a grandparent, or a descendant of a grandparent—thus excluding friends, distant relatives,
    and neighbors).

  14. [14]. Jonathan G. Blattmachr & Dan T. Hastings, The Tax Apportionment Clause – Often the Most Important Provision in the Will, 60 N.Y. St. B.J. 26, 27 (1988).

  15. [15]. Weisbord & Horton, supra note 1, at 668.

  16. [16]. In re Estate of Kuralt, 68 P.3d 662, 664–68 (Mont. 2003).

  17. [17]. See Patricia Cohen, Who’d Gain From an Estate Tax Rollback: The 0.2 Percenters, N.Y. Times (Nov. 16, 2017),; Estate Tax, IRS (May 9, 2018),

  18. [18]. Weisbord & Horton, supra note 1, at 692.

  19. [19]. Weisbord & Horton, supra note 1, at 694 (finding that 48 out of 260 wills used representation language for individuals).

  20. [20]. See, e.g., Ala. Code § 43-8-42 (2018) (stating Alabama uses per capita with representation); Idaho Code § 15–2–106 (West 2018). And Pennsylvania has a per capita with representation scheme, but has a degree of relationship exception. See 20 Pa. Stat. and Cons. Stat. Ann. § 2104 (West 2018).

  21. [21]. See, e.g., Ga. Code Ann. § 53-2-1 (West 2018); La. Civ. Code Ann. arts. 888, 892, and 896 (2018); Mass. Gen. Laws. Ann. ch. 190B § 2-103(4) (West 2018); Vt. Stat. Ann. tit. 14,
    § 314(b)(4) (West 2018). Louisiana, Massachusetts, and Vermont use representation at the closest levels and then a degree of relationship scheme for more distant and collateral relatives.

  22. [22]. Unif. Prob. Code § 2-103 and comment (amended 2010).

  23. [23]. See generally W.W. Allen, Right of Adopted Child to Inherit from Kindred of Adoptive Parent, 43 A.L.R.2d 1183 (1955) (explaining the history of inheritance for adopted children).

  24. [24]. Hallie E. Still-Caris, Legislative Reform: Redefining the Parent-Child Relationship in Cases of Adoption, 71 Iowa L. Rev. 265, 284–87 (1985).

  25. [25]. Mark Strasser, Interstate Recognition of Adoptions: On Jurisdiction, Full Faith and Credit, and the Kinds of Challenges the Future May Bring, 2008 BYU L. Rev. 1809, 1830–31.

  26. [26]. Id.

  27. [27]. See Lawrence W. Waggoner, A Proposed Alternative to the Uniform Probate Code’s System for Intestate Distribution Among Descendants, 66 Nw. U. L. Rev. 626, 628 (1971).

  28. [28]. The Uniform Probate Code’s recommended anti-lapse provision, § 2-603, adopts a sticky default regarding words of survivorship that has not been adopted by most states. See also Unif. Prob. Code § 2-603 cmt. “Contrary Intention—the Rationale of Subsection (b)(3) (amended 2010).


Clarence J. TeSelle Professor of Law, University of Florida Levin College of Law.