Boilerplate and Default Rules in Wills Law: An Empirical Analysis

I.     Introduction

On April 5, 2010, a New Jersey resident named Robert Paulson signed his last will and testament.1 In this short, professionally drafted document, Robert left a house to his daughter, Sharon, and the rest of his property to his wife, Rebecca.2 Five years later, Robert died.3


Figure 1: Robert Paulson’s Will

 Screen Shot 2018 02 05 at 6.19.06 PM

Screen Shot 2018 02 05 at 6.20.22 PM


Although Robert’s will seems straightforward, a closer look reveals that one key provision (which we have highlighted in Figure 1) differs from his probable intent. Suppose the house was encumbered by mortgage debt. Would Sharon, who was living in the house, or Rebecca, the residuary beneficiary and executor, be responsible for discharging the outstanding balance? Similarly, who should pay the estate and inheritance taxes imposed on the house? The state probate code presumes that testators want the recipient of land—here Sharon—to bear both expenses.4 But these presumptions are mere background principles, which testators can override. And in the second paragraph of his will, Robert appears to do exactly that, instructing his executor to pay his “debts . . . [and] inheritance and estate taxes . . . from my [residuary] estate.”5 This “just debts” clause, which Robert probably did not read and likely could not understand, had the potential to alter the ultimate disposition of his property by forcing Rebecca to subsidize Sharon’s inheritance.

Robert’s will highlights a neglected tension in wills law. One of the most important concepts in fields such as contracts, corporations, and inheritance is the default rule. In the last three decades, scores of articles in leading journals have considered how best to calibrate these gap-filling doctrines.6 The conventional wisdom is that most default rules are—and should be—majoritarian: they reflect what most parties want.7 By mimicking widely shared preferences, these background principles minimize the transaction costs of drafting instruments that would otherwise have to address every contingency.8 However, as Robert’s “just debts” clause illustrates, default rules suffer from a glaring vulnerability. Because default rules are so deferential, they can easily be overridden by boilerplate. Indeed, the stock phrase that Robert employed was not custom-tailored for his will; to the contrary, it has been common in estate plans for centuries.9

This Article addresses the friction between boilerplate and default rules in the realm of decedents’ estates. It does so by analyzing a unique dataset of 230 wills that were probated in 2015 in Sussex County, New Jersey. This trove of empirical evidence reveals that recycled language is endemic in wills. This problem is particularly acute among what we call “non-salient” clauses: those that govern important but obscure topics, such as what happens to the share of a beneficiary who dies before the testator, how to distribute property among multi-generational classes, and whether the recipient of a specific devise or the residuary beneficiaries are liable for mortgage payments and estate and inheritance taxes.10 Of course, because we can only guess about the genesis of any particular will—we can neither observe the drafting process nor eavesdrop on conversations between the estate planner and the client—we only have circumstantial evidence that these provisions were cut and pasted from previous instruments. Nevertheless, we show that wills often deal with non-salient matters through language that sounds authoritative, but makes little sense in context.11 Even worse, we demonstrate that these readymade terms usually opt out of majoritarian default rules, thus thwarting efforts by judges and policymakers to create a body of intent-serving background principles.

The Article then draws on these findings to prescribe policy. Most default rules in wills law are “simple,” meaning that they can be displaced by any contrary textual command.12 The Article contends that this rubric is not appropriate for complex and poorly understood topics because simple defaults are particularly vulnerable to the scourge of boilerplate. In fact, we found that boilerplate provisions so routinely displace simple defaults that most of the majoritarian defaults in our sample did not actually apply to a majority of wills. Accordingly, we argue that background principles governing non-salient issues should be “sticky” (harder to draft around).13 By anchoring defaults in this fashion, our proposal would re-align wills with what most testators want. It would also encourage lawyers to ascertain their clients’ wishes, rather than relying on recycled text.

At the outset, we should clarify what we mean by the term “boilerplate.” As will become apparent during our discussion, courts and commentators define this word in subtly different ways. For example, in contract law, “boilerplate” can be an epithet for unfair fine print.14 Alternatively, in both contract and corporate law, it can mean something more benign: “standard legal language that is identical in instruments of a like nature.”15 Our definition of “boilerplate” borrows from both camps. We use the word to mean text that is unlikely to leap out to testators and is copied wholesale from other documents. Throughout the Article, we will try to be attuned to the ways in which our usage both mirrors and deviates from other definitions, and how this impacts the normative dimensions of the boilerplate problem in wills.

The Article contains five Parts. Part II provides background on the relationship between default rules and boilerplate. It reveals that, despite the paucity of scholarship on these topics in wills law, courts and policymakers have long struggled with how to insulate background rules from mindless form provisions. Part III describes our study. It uses six months of probate records to document the fact that terms addressing non-salient issues tend to be boilerplate. In particular, it examines four consequential but abstruse topics: survivorship, representation, lien exoneration, and tax apportionment. Part IV discusses the policy implications of our findings. It argues that courts and legislatures should fortify vulnerable default rules against boilerplate by making them “sticky”: only susceptible to highly specific expressions of intent or external symbols of authenticity such as separate signatures. Part V concludes.

II.     Default Rules and Boilerplate in Wills Law

This Part sets the stage by discussing the relationship between default rules and boilerplate. It reveals that both of these issues have been thoroughly canvassed in contract and corporate scholarship, but virtually ignored in the context of wills. Yet it also demonstrates that efforts to reform wills law have been quietly struggling with these topics.

A.     Default Rules

Some legal rules are mandatory. For instance, parties cannot draft around minimum wage laws16 or the prohibition on contracting with individuals who lack mental capacity.17 These immutable rules restrain negative externalities (like the race to the bottom that might occur if employers could set salaries unilaterally) and protect the vulnerable or ill-informed from bad deals.18      

But areas that place a premium on autonomy—such as contracts, corporations, and wills and trusts—consist largely of default rules, which are waivable.19 These gap-filling tenets govern in the absence of an expressed preference.20 Default rules are generally calibrated to reflect what most parties want, sparing the transaction costs of haggling over and addressing every possible contingency.21

In the last three decades, commentators have explored default rules in contract and corporate law from every conceivable angle.22 For instance, in a seminal piece, Ian Ayres and Robert Gertner argue that not all defaults are majoritarian; rather, some are “penalty defaults”: tenets that are deliberately designed not to mimic what the parties want.23 These penalty defaults force individuals and entities to draft around the unpopular rule and thereby divulge information to each other or the courts.24 Likewise, Alan Schwartz and Robert Scott have criticized the idea that the drafters of the Uniform Commercial Code and the Restatement (Second) of Contracts are capable of creating majoritarian default rules.25 Schwartz and Scott contend these sources actually contain default standards—holistic guidelines like reasonableness and good faith—rather than default rules.26 According to Schwartz and Scott, because most parties prefer crystalline rule-like commands to muddy standard-like benchmarks, they draft around default law, defeating its purpose of saving transaction costs.27 Finally, in a recent article, Ayres turns his attention to “altering rules”: those that regulate how to displace a default.28 In particular, Ayres examines how lawmakers can reduce the risk of parties inadvertently opting out of a majoritarian default by making the rule “stickier.”29 He discusses how the law can align text with intent by conditioning the ousting of a default on an instrument reciting magic words or bearing a separate signature, or even by insisting that one or both of the parties pass a test about the impact of a particular clause.30

In sharp contrast to this rich and rewarding literature, scholarship in wills law has generally overlooked default rule theory.31 One exception is Adam Hirsch’s work.32 In the only comprehensive examination of default rules in decedents’ estates, Hirsch considers whether teachings from other fields can be transplanted into this specialized domain.33 He observes that estate planners may be less likely to take advantage of the gap-filling function of default rules than contract drafters.34 This is because wills often do not kick in until long after they are signed.35 Given the odds that the law will change, it has become “a credo of estate planning that a well-drafted will should anticipate contingencies and never rely on default rules.”36

Nevertheless, Hirsch also cites several ways in which the default rule paradigm fits inheritance law snugly. In contracts, majoritarian defaults must reflect what both parties desire.37 Because contractual partners often have diametrically opposed interests, determining what it means for a default to be “majoritarian” can be hard to pin down.38 Yet wills defaults involve no such obstacle.39 Because wills are unilateral, formulating majoritarian defaults requires ascertaining what most testators desire—an inquiry with fewer variables than in the contracts context.40 Overall, then, Hirsch concludes that the default rule scholarship in contract and corporate law “is readily adaptable to inheritance defaults and points the way to their ideal composition.”41   

However, Hirsch adds an important caveat: He is skeptical about penalty and sticky defaults in wills law. Hirsch notes that “the intent of the testator is ‘the pole-star by which the courts must steer.’”42 He argues that penalty defaults are impossible to square with this objective because they thwart the wishes of uninformed (typically low-income) testators who do not know to draft around the unpopular rule.43 Likewise, he contends that sticky defaults penalize property owners without access to competent legal advice by causing them incorrectly to believe that they have displaced the default rule.44 Accordingly, he asserts that all default rules in the sphere should be both majoritarian and “simple.”45 

Conversely, Shelly Kreiczer-Levy argues that some wills law defaults are—and should be—intent-defeating.46 Kreiczer-Levy defends what she calls “[d]eliberative accountability rules,” which “require the decision-maker to give reasons, make a direct statement of her intentions, and consider other options.”47 She gives two related examples. The first is pretermitted child statutes,48 which give a son or daughter who is accidentally omitted from a will a slice of the estate.49 The second is the negative will, which bars a particular heir from taking in intestacy.50 In general, to draft around pretermitted child protections or to create a valid negative will, testators must explain why they want to disinherit their kin.51 Thus, Kreiczer-Levy concludes that these principles prompt a testator “to think her decision through, give reasons, and face the relational consequences of her act.”52

Although Hirsch’s and Kreiczer-Levy’s contributions are valuable, they have only started the conversation about default rules in wills law. Hirsch focuses largely on intestacy statutes, which are the default rules that distribute property not controlled by will.53 Kreiczer-Levy confines her work to the exceedingly narrow issue of disinheritance.54 As a result, no one has yet analyzed default principles that cover critical aspects of will drafting: matters such as survivorship, representation, debts, and taxes. Moreover, as we discuss next, academics have yet to examine a kind of anti-matter force that undermines the intent-serving function of majoritarian default rules: boilerplate.

B.     Boilerplate

Default rules yield when an instrument expresses a different intent. This pliability promotes autonomy by minimizing the effort necessary to individualize a transaction. Yet the relative ease with which a party can opt out makes default rules vulnerable to boilerplate that distorts their true wishes. This Subpart surveys the clash between default rules and boilerplate and then argues that the same problem lurks beneath the surface in wills law. 

Few issues in contract law are more controversial than fine print. Contract’s marquee doctrines, including mutual assent and consideration, were born at a time when markets were dominated by face-to-face negotiation. But as Edwin Patterson observed in 1919, some exchanges did not fit that mold.55 Rather than arising from bargaining, they were promulgated by the drafter, forcing the other party to choose between walking away or accepting the deal and adhering to the preprinted form.56 Borrowing a phrase from French jurist Raymond Saleilles,57 Patterson called these immutable bundles of text “contracts of ‘adhesion.’”58

Since then, adhesion contracts have swept through the economy, washing away vast segments of default law. Carmakers use fine print to disclaim the implied warranties of merchantability and fitness for a particular purpose.59 Banks use fine print to reverse the American rule and recover attorneys’ fees in litigation against borrowers.60 Railroads, airlines, and trucking companies use fine print to limit their liability as common carriers.61 Even one of the most basic ideals of American civil justice—that every plaintiff is entitled to her day in court—has been eclipsed by the lockstep use of fine print arbitration clauses.62 Critics have long argued that these “agreements” are riddled with “terms whose consequences are often understood only in a vague way, if at all.”63 As a result, “[f]ew topics in recent decades have attracted more attention in contract scholarship than standard-form contracts, and rightly so.”64

Conversely, boilerplate in wills has flown beneath the scholarly radar.65 This might be because wills seem to be impervious to boilerplate. Unlike consumer and employee contracting, where businesses lace their documents with one-sided terms, wills are written by attorneys who are supposed to serve their clients’ interests. Rather than being adversarial, the drafting process is collaborative. In fact, some testators read multiple iterations of their wills and suggest revisions, reducing the risk of important terms going unread or misunderstood.

Likewise, the benefits of boilerplate in wills might seem to dwarf the risks. As contract and corporate scholars have observed, boilerplate can have a silver lining. First, it reduces transaction costs.66 Indeed, borrowing language from previous instruments allows attorneys to churn out estate plans quickly and cheaply. Likewise, estate planning attorneys maintain propriety form books containing blocks of text that they can paste into a will, saving time and money.67 Second, standardized language promotes certainty.68 For example, if a provision has been around for a long time, it confers “network benefits”: an accumulation of judicial rulings that allows parties and their attorneys to predict its impact with greater clarity.69 This can be particularly valuable in a risk-adverse field like will-drafting, where certainty is the coin of the realm. Third, boilerplate sometimes reflects a consensus within a firm, subfield, or jurisdiction regarding best practices. In this way, boilerplate can “represent[] the accumulated wisdom of prior drafters.”70

Finally, one might believe that boilerplate in wills is innocuous because there is no need for a testator to read—let alone grasp—every nuance in their estate plan. A similar viewpoint has emerged in the debate over adhesion contracts. Some commentators reject the charge that adhesion contracts are non-consensual by questioning the value of consent itself. Seen through this lens, it is “quaint” to insist “that a weaker party’s acquiescence in market power can only be legitimated by some transcendent insight or internal transformation.”71 As a result, “contractual consent is a transaction cost to be minimized, not a good to be maximized.”72 In the same vein, it seems quixotic to force attorneys to educate their clients about the minutiae of their wills. Instead, both parties expect that the lawyer will make some decisions on the testator’s behalf without securing the testator’s informed consent.73 Perhaps for these reasons, boilerplate in wills—not to mention the pernicious relationship between boilerplate and default rules—has never received sustained attention.

Nevertheless, there is one context—the doctrine of antilapse—in which scholars in the field of decedents’ estates have recognized the fraught relationship between default rules and boilerplate. Antilapse is an oft-invoked default rule that addresses the problem of a beneficiary who predeceases the testator. Under the common law rule of lapse, a beneficiary must be alive at the testator’s death to inherit.74 For instance, suppose T’s will leaves her valuable watercolor to A, and the residue to B. A has a daughter, AA, and B has a son, BB. At common law, if A dies before T, the devise of the painting lapses, falls into the residue, and passes to B. Likewise, if B dies before T, then B’s residuary devise lapses and passes to T’s intestate heirs.

The overwhelming majority of states have partially modified the common law lapse rule by enacting antilapse legislation. Antilapse laws provide that, when a predeceased beneficiary is a close relative of the testator (usually a descendent of the testator’s grandparents), then the gift does not lapse.75 Instead, the property flows to the predeceased beneficiary’s descendants.76 Accordingly, in the example above, if A was T’s predeceased brother—a relationship triggering antilapse—then AA (T’s niece) would take the watercolor. Conversely, if A was merely T’s predeceased friend, antilapse would not apply, and the art would fall into the residue for B.77 Similarly, if B was T’s predeceased son—another relationship triggering antilapse—then BB (T’s grandson) would take the residuary estate. 

However, antilapse statutes are default rules that can be displaced by expressions of contrary intent—and here the trouble begins.78 One way of drafting around antilapse is to expressly address the contingency of a predeceased beneficiary by making an alternative devise, such as “to my brother, A, but if A dies first, to my cousin, C.” In that circumstance, even though A would normally qualify for antilapse protection, the testator has written around the statute by making a substitute gift to C. Unfortunately, it is less clear whether language in a will that merely requires survival—“to my brother A, if he is then-living”—is specific enough to manifest the testator’s intent to opt out of antilapse. Because most estate planners believe that these survivorship conditions should be read literally (e.g., the devise to A is effective only if A is alive),79 the traditional approach is to interpret them as overriding antilapse.80 Thus, in the example above, if T had given the watercolor “to my brother A, if he survives me,” the gift would lapse, antilapse would not kick in, and B—and not AA—would receive the painting. 

Yet the drafters of the 1990 amendments to the Uniform Probate Code (“UPC”) and the 1999 Restatement (Third) of Property: Wills and Other Donative Transfers took a more jaundiced view of survivorship conditions. They worried that stock phrases, such as “to X, if X survives me,” have become so common in wills that they are no longer reliable evidence of the testator’s intent:

Because such a survival provision is often boiler-plate form-book language, the testator may not understand that such language could disinherit the line of descent headed by the deceased devisee. When the testator is older than the devisee and hence does not expect the devisee to die first, or if the devisee was childless when the will was executed, it seems especially unlikely that a provision requiring the devisee to survive the testator was intended to disinherit the devisee’s descendants.81

Accordingly, to solve this perceived problem, they changed antilapse from its traditional status as a mere default rule to a harder-to-displace sticky default:

[W]ords of survivorship, such as in a devise to an individual “if he survives me,” or in a devise to “my surviving children,” are not, in the absence of additional evidence, a sufficient indication of an intent contrary to the application of th[e antilapse statute].82

This proposal has been polarizing. A handful of states signed onto the UPC and Restatement’s approach.83 Yet leading academics summarily rejected the notion that words of survivorship should be viewed as boilerplate and criticized the UPC’s sticky default for disregarding an instrument’s plain language. Mark Ascher put this point tartly: “Apparently, the revisers believe their own antilapse provisions are likely to reflect any particular testator’s intent more faithfully than the testator’s own will.”84 Other policymakers agreed and went so far as to pass laws expressly rejecting the novel rules.85 In this one domain, then, there has been at least some acknowledgement that improper use of boilerplate can erode majoritarian defaults.86

But upon closer inspection, antilapse is the tip of the proverbial iceberg. Courts and lawmakers have struggled with other kinds of stock testamentary language for decades. The common thread among these patterns of boilerplate is that they govern what we call “non-salient” issues: topics that represent important estate planning considerations but are unlikely to leap out to most testators as part of the will-drafting process.

One of the most striking examples is the “just debts” clause, a simple incantation at the beginning of most wills that says something like: “I direct that my executors hereinafter named shall pay all of my just debts as soon as practicable . . . .”87 As early as 1795, a North Carolina court noted that these clauses were “common in almost all wills.”88 Today, nothing has changed; indeed, as the Minnesota Supreme Court remarked in 2012, “just debts” provisions are a shining example of “boilerplate will language.”89

The prevalence of these “just debts” clauses is deeply puzzling. Even if a will does not include a “just debts” clause, one of an executor’s most elementary responsibilities is to reimburse a testator’s creditors.90 Thus, decade after decade, judges have observed that a “direction to pay all the testator’s just debts adds nothing to the duty imposed upon all executors by law.”91 Indeed, such a clause “is a mere formality”92 that is “wholly unnecessary”93 and “legally meaningless.”94

At the same time, “just debts” provisions have a tendency to spark disputes about a testator’s intent.95 Consider the common situation described at the beginning of this Article, when a testator dies owning real estate that is subject to a mortgage. Jurisdictions are divided about whether to assume that the testator wanted the recipient of a bequest of real property or the residuary beneficiaries to pay off any such loan.

To make this topic concrete, suppose T’s will gives her home to her son, A, and the remainder of her estate to her daughter, B. T dies owning the house, on which there is a mortgage of $50,000, and $300,000 in cash. Under the common law, unless T’s will expressly states otherwise, A takes the real property free and clear because A is entitled to have the executor “exonerate” the lien.96 Exoneration requires the executor to repay the mortgage on real property passing to A by using cash from B’s residuary estate.97 As a result, the value of A’s specific devise increases by $50,000, and the value of B’s residuary devise decreases by an equal amount to $250,000.

Over the course of the 20th century, several judges and legislatures observed that this “exoneration doctrine”—which privileges recipients of real property at the expense of residuary beneficiaries—reflected the primacy of landownership in feudal England, and has become outmoded.98 In an attempt to align the benefits of inheritance with the burdens, they adopted the non-exoneration doctrine, which presumes that real property descends to beneficiaries along with the duty to pay off any mortgage.99

Nevertheless, the widespread use of rote “just debts” clauses has muddied the waters around the non-exoneration default. Mortgages are debts, and when a testator demands the satisfaction of her “just debts”—but does not specify where the money should come from—she seems to instruct the executor to drain the residue.100 Then again, because “just debts” provisions are so prevalent, giving them this power would have the effect of reversing the majoritarian default and requiring most residuary beneficiaries to repay outstanding loans on property passing to other beneficiaries. Accordingly, courts have never agreed on how “just debts” provisions should factor into the exoneration analysis. Some hold that “just debts” clauses “cannot be regarded as meaningless.”101 Others brush this language aside as “a direction . . . to do what the law requires to be done [that] can throw no material light upon the meaning of the will.”102 And still others find that “just debts” terms “should be given weight” but are “not conclusive.”103

Likewise, courts and policymakers have struggled with boilerplate in yet another context: tax apportionment clauses. Testators enjoy the power to allocate death tax liability among the various recipients of the estate—a choice that can dramatically affect the net inheritance received by each beneficiary. At common law, residuary takers paid all death taxes unless the will provided otherwise.104 However, this “burden on the residue” principle proved out-of-step with prevailing norms. Residuary beneficiaries tend to be a surviving spouse or minor children, so it seemed both unfair and inconsistent with the testator’s probable intent to leave them with the entire tax bill.105 Thus, in 1958, the Uniform Law Commission promulgated the Uniform Estate Tax Apportionment Act, which reverses the burden on the residue rule in favor of “equitable apportionment,” a regime in which each beneficiary presumptively pays a proportionate share of estate taxes imposed on the probate estate.106 Since then, equitable apportionment has become the runaway majority approach.107

Consistent with this perspective, estate planners believe that, in most contexts, spreading the burden of taxes equally among beneficiaries is superior to penalizing residuary beneficiaries disproportionately.108 For one, they doubt that testators truly understand the degree to which a “residue-pays” arrangement can distort the ultimate distribution of assets.109 Residuary beneficiaries are often the people whom the testator loves the most, so it would be odd to burden them with a disproportionately large share of the estate tax liability.110 In addition, a residue-pays arrangement tends to contradict established norms of estate tax planning. Married testators often leave the residue to their surviving spouse, who inherits free of federal estate tax under the marital deduction.111 As a result, a residue-pays regime actually increases the testator’s total tax liability by reducing the amount of property that qualifies for the marital deduction.112 For these reasons, practitioners sometimes regard “a direction in a will that all death taxes shall be paid from the residue of the estate [as] an example of malpractice per se.”113

Paradoxically, though, it appears that wills often override the equitable apportionment default, causing unintended consequences. The estate of CBS News icon Charles Kuralt offers a famous example. In 1994, Kuralt executed a formal attested will that left the residue of his property to his wife and his children.114 This instrument expressly opted out of the equitable apportionment presumption by stating that “[a]ll estate, inheritance . . . and other death taxes” should be borne by the residuary beneficiaries.115 In 1997, Kuralt signed a codicil giving land in Montana to his longtime mistress (about whom his wife did not know).116 There was strong evidence that Kuralt did not realize that his wife and children would have to pay the taxes attributable to the Montana property.117 Yet the Montana Supreme Court enforced the tax apportionment clause in the 1994 will, adding insult to injury by forcing Kuralt’s surviving spouse and children to underwrite his secret lover’s inheritance.118

Tax apportionment clauses are also often ambiguous. For instance, many instruments contain vague statements about taxes in the same paragraph as the “just debts” clause:

FIRST: I direct that all lawful debts I owe at the time of my death, including funeral and administration expenses and the expense of my last illness . . . and [a]ll estate and inheritance taxes, be paid as soon after my death as can lawfully and conveniently be done.119

Courts have faced a torrent of disputes about these provisions. Most have held that these terms trump the equitable apportionment default, citing their “implied direction” that taxes, like debts and other costs, should be paid from the residue,120 and refusing to reduce words in a will to “pure surplusage.”121 Yet a vocal minority of courts disagree, explaining that this “stock language”122 does not “unambiguously . . . state[] an intent not to apportion.”123

To summarize, courts and lawmakers have sometimes expressed skepticism that language in a will faithfully embodies the testator’s intent. But to date, this view has been founded on conjecture, not hard evidence. In the next Part, we move beyond anecdote. By surveying 230 recently probated wills, we demonstrate that boilerplate is a serious and systemic problem.  

III.     Empirical Results

This Part reports the results of our study. It first describes how we collected and analyzed our data. It then explains why our findings suggest that boilerplate is endemic in wills.  

A.     Data Description

We chose to study wills from Sussex County, New Jersey, a rural area about 50 miles west of Newark. In recent years, the region has transitioned from an agricultural community to a suburb, with many of its residents commuting into New York City.124 Its total population of about 150,000 has a median annual household income of about $86,565,125 which makes it wealthier than the overall U.S. population.126 Despite this discrepancy, we were excited to gather data from Sussex County for several reasons. For one, to gauge whether language in wills remains static even when the law changes, we wanted to find a jurisdiction that had adopted the most recent amendments to the UPC within the last two decades.127 New Jersey fit the bill: It adopted much of the revised UPC in 2005.128 As a result, New Jersey probate courts (known as “surrogate courts”) are currently processing instruments executed both before and after this exogenous shift in the law. In addition, the Sussex County Surrogate, Gary Chiusano, kindly facilitated our ease of access to all files from the public record for a six-month period beginning February 1, 2015 and ending July 31, 2015.129

This dataset originally consisted of more than 5,000 pages of court files. It broke down into 260 testate cases, 78 intestate administrations, and 53 affidavit matters (which involve estates of less than $20,000). To focus narrowly on the topic of boilerplate, we excluded intestacies (which do not involve wills) and affidavit cases (which do not require wills to be lodged in the court files). In addition, we cut two kinds of wills from the initial collection. First, we eliminated any document executed outside New Jersey. Because these wills were drafted under the auspices of another state’s law, they could have distorted our findings. Second, we purged all “pour over” wills. Decedents who transmit their assets through revocable trusts often execute pour over wills: those that leave everything to a trust, rather than individual beneficiaries.130 The trust, not the will, serves as the primary estate planning device. This pour over technique ensures that any asset that the decedent fails to title in the name of the trust during life nevertheless passes under the terms of the trust at death.131 Because pour over wills are not the primary estate planning vehicle for the testators who create them, they tend to be relatively short and simple, and thus are less relevant for our purposes. Together, these adjustments left us with 230 wills. 

We then collected about 25 variables from each estate. We gathered general information such as the identity of surviving family members and the dates of will execution,132 death, and the filing of the probate case. Then we focused on the nitty-gritty in each will: arcane but consequential provisions that govern issues such as the payment of the testator’s debts, the apportionment of tax liability, the requirement that the executor post bond, and the period by which a beneficiary must outlive the testator to inherit. 

At the outset, we acknowledge a few caveats. First, for the most part, our sample does not include contested matters. By statute in New Jersey, the Surrogate lacks jurisdiction to resolve complex or disputed estates.133 Any case that degenerates into litigation must be transferred to the Superior Court Chancery Division Probate Part for adjudication.134 However, Sussex County only reported 20 such matters during the half-year period under our microscope.135 Because our combined dataset contains 391 estates, it is unlikely that small gap in our data distorts it.

Second, one might contend that selecting New Jersey threatens to make our findings unrepresentative of will-making patterns elsewhere in the country. Six decades ago, the Garden State pioneered an idiosyncratic approach to the interpretation of wills known as the “probable intent” doctrine.136 Unlike many other jurisdictions, which refuse to admit extrinsic evidence when an instrument is clear on its face,137 the probable intent doctrine allows judges to divine the testator’s intent through other sources:

[T]he judicial inquiry must focus on the subjective intent of the testator as evidenced not merely by the text of the will but, primarily, by the testator’s “dominant plan and purpose as they appear from the entirety of his will when read and considered in the light of the surrounding circumstances,” ascribing to the testator, “[s]o far as the situation fairly permits ‘those impulses which are common to human nature . . . .’”138

Arguably, then, this searching inquiry reduces pressure on New Jersey testators to ensure that their wills are airtight expressions of testamentary intent.

Nevertheless, we are not persuaded that the probable intent doctrine changes testators’ incentives. For one, some recent New Jersey appellate decisions have throttled back on the doctrine, using it only when a will is vague or confusing.139 Seen this way, the probable intent doctrine is not unique; to the contrary, it is merely a different label for the widely acknowledged principle that extrinsic evidence is admissible to uncover and resolve ambiguity.140 And even if this were not so, it seems far-fetched that testators would pay less attention to the clarity of their wills because of the probable intent doctrine. Nobody wants their estate to become bogged down in litigation, and a well-drafted instrument remains the first line of defense against a dispute.

B.     Results   

At the outset, our data provide some insight into the debate over whether default rule theory in contract and corporate law—in particular, the notion that majoritarian defaults reduce transaction costs—can be neatly incorporated into wills law. Recall that one bone of contention is whether estate planners are as likely as contract drafters to minimize transaction costs by leaving gaps in their instruments, thereby allowing default rules to govern the unaddressed contingencies.141 For instance, Adam Hirsch argues that probate lawyers should not rely heavily on defaults because they are not locked-in as governing law until the testator dies:

Although a theorist might conjecture that default rules can function to reduce the marginal cost of preparing a will by removing the need to spell out contingencies, a competent practitioner in the field would scoff at that idea. . . . [T]he potentially long latency period before a will matures, coupled with the possibility that the benefactor will migrate to a different jurisdiction in the interim, raises the prospect that the original default rule upon which a drafter relied may not in due course govern the testamentary instrument.142

Our findings suggest that Hirsch is correct that wills are often drafted under one set of rules but probated under another. For starters, 16 of the original 260 wills (6%) in our raw data—a small but non-negligible number—were created in states other than New Jersey. In addition, many wills do indeed have long fuses.143 As Figure 2 elucidates, the average span between the date of execution and the testator’s death was 9.5 years. Even more to the point, 86 wills (37%) predated New Jersey’s adoption of the most-recent revisions to the UPC.144 Thus, there is no assurance that the relevant default rule when the testator signs her estate plan will be the same one that applies when she dies. We will return to this point in our discussion of policy implications in Part III.B. But now we examine how the Sussex County wills dealt with matters such as survivorship conditions, multi-generational classes, exoneration of loans, and death taxes. 


Figure 2: Number of Wills Executed in Each Year

Screen Shot 2018 02 05 at 6.33.42 PM

1.     Survivorship Language

Recall that the only mention of boilerplate in wills scholarship relates to the issue of antilapse.145 In most states, including New Jersey, survivorship conditions—“to A, if he survives me”—override the antilapse statute.146 Conversely, the UPC and Restatement believe that references to survivorship are usually “form-book language” and should be ignored.147 This Subpart explains why our data offers qualified support to both camps. In addition, it highlights a related problem that has entirely eluded scholarly attention: boilerplate that erroneously invokes language of representation for multigenerational gifts.

i.     Survivorship Conditions and Antilapse

Our data render a mixed verdict on the debate over survivorship conditions and antilapse. On the one hand, we found that survival clauses are relatively uncommon, and thus may not require a sticky default to prevent them from inadvertently interfering with the antilapse statute. But on the other hand, some of the survivorship conditions we did discover have the telltale signs of boilerplate.

Contrary to the UPC and Restatement’s concern that testators often overlook the impact of survivorship clauses, we found that the Sussex County wills were attuned to the possibility that a beneficiary might die first. One hundred ninety-seven of the 230 (86%) instruments expressly addressed predeceasing beneficiaries by including at least one alternative devise: for example, “to my wife, but if she dies before me, to my children equally.” Just 40 testators (17%) relied solely on clauses containing bare survival language, such as “to X, if X survives me.” Further, a close examination reveals that 24 of these 40 wills (60%) included bare survival conditions only to control a contingent devise of tangible personal property,148 which rarely comprises the bulk of an estate.149 The paucity and relative unimportance of survivorship conditions suggests that they may not be a pressing problem. 

Our second finding, however, provides qualified support for the reformers’ concern about boilerplate survival clauses intruding upon the antilapse doctrine. Some of the survival clauses in our sample are, in fact, mindless word balloons. Recall that the antilapse statutes only govern devises to descendants of the testator’s grandparents: for instance, the testator’s siblings, nieces, or grandchildren.150 This means that antilapse does not cover devises to neighbors or far-flung relatives. If a testator makes a gift to one of these individuals, but wants it to lapse if the beneficiary dies first, there is no reason to use a survivorship condition to draft around the antilapse statute. Return to the example in Part II.B, where T leaves her watercolor to her friend A and the residue to B. If T wants B to take if A dies first, T does not need to add the words “to A, if A survives me” or “to A, if A is then-living,” because antilapse does not apply to A, a friend whose relationship does not trigger antilpase treatment. Thus, if A predeceases T, B takes the painting even if the will does not contain a survival mandate because the background principle of lapse applies to the devise. Yet we unearthed four wills that imposed survivorship conditions on beneficiaries who did not qualify for antilapse protection. Because this language serves no purpose, it was almost certainly boilerplate.151

Although this discovery hardly justifies a rule that disregards bare survivorship conditions, it does imply that the UPC and Restatement’s concerns are not baseless.152 As we discuss next, we also find that illogical boilerplate is a significant problem in a related context that, to date, has gone unnoticed.

ii.     Language of Representation

Despite the attention focused on boilerplate survivorship conditions, we found that a similar but even more significant issue has gone unnoticed. Our sample contains a large number of wills that erroneously invoke language of “representation,” the method for distributing property among multiple generations of intestate heirs and will beneficiaries.  

To illustrate the concept of representation, suppose a widow, W, dies intestate, leaving two predeceased children, C1 and C2. In turn, C1 had one living child (W’s grandchild), GC1. C2 left two living children (also W’s grandchildren), GC2 and GC3. How should a probate court distribute W’s property? 


Figure 3: Representational Schemes

Screen Shot 2018 02 05 at 7.35.49 PM


As a general rule, when an intestate decedent outlives some of her offspring, her surviving relatives “represent”—in other words, step into the shoes of—the ones that have already passed away.153 There are several different ways of defining “representation.” The traditional perspective, known as “English per stirpes” or “strict per stirpes,” divides the estate into shares among the decedent’s children.154 After this initial allocation, survivors of each line of descent split whatever their predeceased ancestor would have taken.155 Thus, in the hypothetical above, the widow’s estate would be divided into two, half for C1’s stock and half for C2’s stock. GC1 then would represent his predeceased parent, C1, and take half of W’s estate. GC2 and GC3 would represent their predeceased parent, C2, and divide C2’s half two ways, each taking one quarter.

Conversely, many UPC jurisdictions, including New Jersey, follow a more modern system of representation called “per capita at each generation.”156 The animating force behind per capita at each generation is “equally near, equally dear”: the idea that each surviving relative of a particular class—for instance, the decedent’s grandchildren—should receive the same amount of property.157 Per capita at each generation achieves this horizontal symmetry by dividing the estate at the first level in which there are living descendants and also giving each member of the level the same share. Thus, in the hypothetical above, the shares of the predeceased children C1 and C2 would be distributed equally among the members of the next generation, leaving each grandchild—GC1, GC2, and GC3—with one-third of W’s estate.

Critically, representational schemes apply not only to intestate estates, but also to wills that make gifts to multi-generational classes.158 In the examples above, suppose W executed an instrument that left her property “to my descendants by representation.” In a state that follows English per stirpes, GC1 would take one half of W’s property, while GC2 and GC3 could each receive one quarter. But in a jurisdiction following per capita at each generation, GC1, GC2, and GC3 would each take a third. Likewise, regardless of the background law, testators sometimes affirmatively select a particular system of representation by leaving their assets to their descendants “per stirpes” or “per capita.”    

Bizarrely, though, many of the Sussex County wills invoke language of representation in gifts to individual beneficiaries, not multi-generational classes. For instance, four testators left property to single recipients “per capita.” Because “per capita” means “[d]ivided equally among all individuals . . . in the same class,”159 it is a non-sequitur in a devise to one person.160 An additional 44 wills—nearly one out of every five in our sample—contained a similarly illogical devise to an individual “per stirpes.”

These inapposite references suggest that neither the drafter nor the testator understood the meaning of “per capita” or “per stirpes,” but nonetheless chose to use those terms. In some wills, the scrivener’s ignorance of the law is evident. For example, one will used “per stirpes” five times, but defined it in a fashion betraying the drafter’s complete misunderstanding of the term: “Per Stirpes, for the purposes set forth herein, shall mean any child born or adopted by any of my children prior or subsequent to the signing of this my Last Will and Testament and living at the time of my death.”161 Another testator included a dizzying pastiche of representation language alongside other specialized terms, leaving money to her son-in-law and sister, “absolutely and in fee simple, per capita,” and the residue to her six children “in equal shares, one sixth each, share and share alike, absolutely and in fee simple, per stirpes.”162

This haphazard vocabulary makes it hard to discern a testator’s intent with respect to the antilapse statute. One way to make sense of a devise to a single person “per stirpes” or “per capita” would be to read it as demanding antilapse treatment regardless of the beneficiary’s relation to the testator. Suppose a testator adds language of representation to a gift to someone not covered by the antilapse statute, such as “to my co-worker, C, per stirpes” or “to my son-in-law, S, per capita.” This additional phrase selects a particular system of representation for distributing property to the beneficiary’s descendants, information that would only be relevant if the beneficiary predeceased and the testator intended to make a substitute devise to that beneficiary’s descendants (e.g., antilapse treatment). Thus, courts sometimes interpret a bequest to a single person “per capita” or “per stirpes” as opting into antilapse if the named beneficiary dies before the testator.163 But other judges have rejected this reading, calling such language a “legalistic flourish, devoid of any expression of intent” and “mischievous.”164 No matter which view is correct, the rampant misuse of representational terminology creates needless ambiguity.165

In sum, the UPC and Restatement appear to be right that some survivorship conditions are only “in the testator’s will merely because the testator’s lawyer used a will form . . . .”166 At the same time, though, the scope of this problem pales in comparison to the robotic misuse of representation language. And, as we discuss next, these are hardly the only examples of testamentary boilerplate in our study. 

2.     “Just Debts”

As noted above, “just debts” clauses have vexed courts for centuries.167 Today, these provisions sow confusion about whether a testator intended to opt out of the non-exoneration doctrine and force residuary beneficiaries to pay off mortgages encumbering specifically devised real property.168

This problem has been particularly acute in New Jersey. In 1981, the state legislature attempted to abolish the exoneration doctrine. It passed a statute that replaced the common law exoneration doctrine with a default rule that a beneficiary who inherits real property is responsible for any outstanding mortgage or security interest on the property:

When property subject to a mortgage or security interest . . . passes to a devisee, [she] shall not be entitled to have the mortgage or security interest discharged out of any other property of the . . . testator, but the property . . . shall be primarily liable for the mortgage or secured debt, unless the will of the testator shall expressly or impliedly direct that the mortgage or security interest be otherwise paid.169

Although the statute applies only to transfers by wills, in 1997, a New Jersey appellate court extended the non-exoneration doctrine to nonprobate transfers.170 In In re Estate of Zahn, the court held that a generic “just debts” clause in the will did not give a surviving joint tenant (the decedent’s girlfriend) a right to have the residuary beneficiaries (the decedent’s children from a different relationship) exonerate an unpaid mortgage on the property.171

Seven years later, when New Jersey updated its probate code to reflect the amendments to the UPC, policymakers revisited the non-exoneration statute. They instituted two changes that made it even harder for the recipient of real estate to argue that a generic just debts clause saddled the residuary beneficiaries with the unpaid mortgage debt. First, the 1981 version of the law had established non-exoneration as the default “unless the will of the testator shall expressly or impliedly direct . . . otherwise. . .”172 The 2004 iteration deleted the italicized text—in particular, the word “impliedly”—and thus appeared to foreclose the claim that a will could tacitly authorize exoneration.173 Second, the legislature emphasized this point by adding a final sentence: “A general direction in the will to pay debts shall not be deemed a direction to pay the mortgage . . . .”174 These amendments seemed to drive the last nail into the coffin of the exoneration rule.

Nevertheless, in 2006, a New Jersey Supreme Court opinion, In re Estate of Payne,175 cut sharply in the other direction. Ted Payne held interests in two pieces of real estate: a vacation home in Maine that he shared as joint tenants with right of survivorship with his former lover, Frederick Wohlfarth, and a house in New Jersey, where he lived with his partner, Don Burton.176 In 1998, Payne executed a will that included a generic “just debts” clause and left his stake in the Maine property to Wohlfarth, along with enough money to satisfy its mortgage.177 Wohlfarth “divided the residue among nieces, nephews, godchildren, charities, and educational institutions,” but left nothing to Burton.178 In 2002, however, Payne revised his estate plan to give his New Jersey home to Burton.179 In a letter to his lawyer, Payne expressed his understanding that both Wohlfarth and Burton would receive their bequests without any outstanding mortgage, writing: “As may be evident from my will, I want the debt encumbering my real estate liquidated by whatever means so that it passes to the beneficiaries free and clear.”180 A divided state high court read the “just debts” clause together with the letter to hold that Payne intended to exonerate the loan on the New Jersey house:

[T]he interpretation that Payne intended the “all my just debts” clause in his will to satisfy the mortgage debts on the New Jersey property is consistent with the opening phrase in his letter that “as may be evident from my will.” That is, it was only “evident” from his will that the property would pass to his beneficiary free and clear if the just debts clause required his estate to pay the mortgage debts on the New Jersey property.181

Thus, despite the 2004 modifications to the non-exoneration statute, the majority interpreted a generic “just debts” provision to shift the costs of a mortgage to the residue.182 The dissent, however, found Payne’s letter more ambiguous than the will.183 Whereas the letter did not expressly mention either the Maine or New Jersey properties, the will unequivocally treated exoneration of the two properties differently.184

After Payne, one thing seemed clear: It no longer made sense to use a generic “just debts” clause that did not expressly state whether the residuary beneficiaries or the recipients of real property were responsible for any mortgage. While unadorned just debts clauses have never served a useful purpose, under Payne, they are no longer a mere superfluity. Instead, they inject needless uncertainty about the testator’s intent and invite the admission of contestable extrinsic evidence. A testator would be better off either expressly addressing the question of exoneration or omitting the just debts clause altogether.

However, our data reveal how boilerplate can weather changes in the law. A whopping 220 of the 230 (96%) instruments in our sample contained a “just debts” clause. By itself, this is striking: As noted, for more than a century, these terms have been condemned for spawning litigation and yet “add[ing] nothing to the will.”185 Even more remarkably, 157 of the 220 wills (71%) with “just debts” provisions were “generic”: they did not also address exoneration of loans. To be sure, the incidence of these clauses decreased after April 20, 2006, when Payne was decided. Before this date, 81 of 92 wills (88%) included generic “just debts” language; afterwards, 86 of 138 wills (62%) did so
(a statistically significant difference, p<0.01).186 This suggests that some attorneys modified their instruments to keep pace with developments in the law. Nevertheless, despite this small step in the right direction, most lawyers clung to generic “just debts” language, even though it is pure risk with no corresponding benefit.

3.     Tax Apportionment

As we have mentioned, testators can also greatly enlarge or reduce the size of a beneficiary’s net inheritance by allocating tax liability.187 Yet as we show in this Subpart, these provisions sometimes also seem to be boilerplate.

We expected the testators in our sample to be crystal clear about the apportionment of death taxes. For one, the topic is especially important to New Jersey residents. The state has been branded “[t]he worst place to die” from a fiscal perspective.188 During the period of our research, it was one of only two jurisdictions to impose both an estate and an inheritance tax.189 Also, its estate tax exemption, $675,000,190 was the lowest in the country.191 Roughly 11% of New Jersey deaths triggered at least one of these two taxes;192 by comparison, about 0.2% of decedents pay federal estate tax, and there is no federal inheritance tax.193

Yet we were surprised to discover that many tax apportionment provisions did not seem like the product of careful consideration. For starters, in New Jersey, it is an open question whether a generic tax apportionment clause—one that merely instructs the executor to pay death taxes in the same breath as it instructs the executor to pay debts and funeral expenses—is enough to rebut the presumption of equitable apportionment.194 Despite this uncertainty, 22 wills (14%) featured generic tax apportionment clauses. This vague language simply invites litigation. Even worse, some of the tax apportionment provisions were marred by logical inconsistencies. The most extreme example was a clause that contradicted itself by making a general bequest of $200,000 and then stating: “Said bequest shall be responsible for the [i]nheritance tax due and owing. Any [i]nheritance taxes owed shall be paid from my residuary estate.”195

* * *

Boilerplate is not regarded as a major problem in wills law. However, our empirical survey reveals that non-salient issues in wills are often governed by inartfully drafted or poorly selected text. In the next Part, we discuss the policy implications of this finding.

IV.     Policy Implications

 This Part uses our data to prescribe policy. We argue that the widespread use of boilerplate makes simple defaults a poor choice for non-salient issues such as representational schemes, exoneration of mortgages, and tax apportionment. Instead, we argue that courts and lawmakers should insulate these majoritarian defaults by making them “stickier.”     

A.     Simple Defaults 

Most default rules are “simple”: They govern in the face of silence, but yield to any expression of contrary intent. Nevertheless, in this Subpart we argue that simple defaults are generally not appropriate for obscure but important issues.  

Wills law consists largely of simple default rules. Indeed, this is the default rule about default rules. This constellation of background principles includes issues such as the effect of a divorce on a devise to a spouse,196 what happens when the testator and a beneficiary die at roughly the same time,197 the rights of someone who was supposed to receive a specific item that the testator no longer owns,198 and whether a bequest to a creditor satisfies the debt.199 On each of these topics, the law presumes that testators intend a particular result, but does not place a thumb on the scale against a party who wishes to prove otherwise.  

Simple defaults have many advantages. First, they make wills more user-friendly than sticky defaults, which govern unless the testator takes elaborate steps or invokes magic words to opt out. As such, sticky defaults “make[] it harder for a lay person to understand the words an estate planner has chosen on his or her behalf.”200 For instance, the average person might assume that a bequest of $100,000 “to my son, A, if he survives me” means that A’s descendants do not take the cash if A passes away before the testator. Yet the UPC and Restatement’s sticky antilapse rule compels a different and counterintuitive result. Second, simple defaults impose lower transaction costs than sticky defaults. It takes time and effort to draft around a sticky default; alternatively, dislodging a simple default does not require extensive legal know-how or custom-tailored language.

Nevertheless, our research illustrates that these virtues of simple defaults can also be vices. Because simple defaults are so fragile, they can be washed away by the relentless tides of boilerplate. Consider the issue of representational schemes. As noted, every jurisdiction has a background rule that distributes property among a multi-generational class.201 This regime—be it per stirpes or per capita at each generation—is a simple default. As a result, a stray reference in the will to a complex topic can dramatically alter the distribution of the estate.202

Likewise, simple defaults may not be appropriate for the non-exoneration and equitable apportionment doctrines. These background principles are relatively uncontroversial policy choices about what most informed decedents want. They stem from efforts by forward-looking courts and reform-minded institutions like the Uniform Law Commission to modernize misguided strands of the common law. But despite their pedigree, these default rules did not control a sizeable percentage of the wills in our sample. Although unsettled New Jersey law makes it difficult to generalize, only 70 of the wills in our sample (30%) were clearly subject to the non-exoneration default. Likewise, despite the fact that the equitable tax apportionment rule applies either if a testator expressly adopts it or does not include a clause that speaks to the allocation of death taxes, this default applied only to a mere 97 (42%) wills. Accordingly, these majoritarian default rules do not, in fact, govern a majority of wills.  

Of course, it is possible that the Sussex County testators deliberately rejected these principles. We uncovered at least one example of the routine displacement of what seems to be a wildly unpopular background rule. In New Jersey, as in every state, a decedent’s personal representative must post a bond unless the will waives it.203 The purpose of the bonding requirement is to insure the heirs and beneficiaries against the personal representative looting the estate. However, bonds are expensive, and most testators pick a close relative or trusted friend to manage their affairs.204 As a result, 223 of the wills (97%) in our sample waived the bonding requirement. In the same vein, one might argue that the number of testators who drafted around the per capita at each generation, non-exoneration, and equitable apportionment defaults did so because they were dissatisfied with them.

Yet two important clues point in the opposite direction. First, there are reasons to doubt that testators truly understand these jargon-laced provisions.205 For instance, of the 144 wills that contained intelligible tax apportionment provisions, a whopping 124 (86%) defied the conventional wisdom by requiring the residuary beneficiaries to pay all death taxes.206 Many of these clauses seemed to be rank boilerplate. For example, 69 of the 124 instruments (56%) featured no general or specific devises and instead only included a residuary clause. These wills did not need language mandating that the residuary beneficiaries pay death taxes for a simple reason: There were no other beneficiaries.207 At best, this bears out skeptics’ claims that “[f]or many lawyers, directing that the taxes be paid from the residue appears to be a kind of habit or reflex, like wearing a coat and tie.”208 At worst, it implies that potentially important but obscure language in wills goes unnoticed until it is too late to change it.   

Second, we also found proof that attorneys recycle boilerplate. Ninety-one estates in our sample revealed the identity of the drafter. In this subsample, several firms popped up at least three times. As Figure 4 demonstrates, there was literally no variation within each entity’s “just debts” and tax apportionment clauses. This standardization of provisions within firms stands in stark contrast to the diversity of provisions between firms. For example, all of Firm 1’s wills (1) used a “just debts” clause with a non-exoneration mandate; (2) opted out of the equitable tax apportionment default; and (3) opted out of per capita at each generation by affirmatively selecting the per stirpes representational system. Conversely, each of Firm 3’s instruments pointed in the opposite direction, (1) employing a “just debts” clause without mentioning non-exoneration; (2) selecting equitable tax apportionment; and (3) not tinkering with the representational scheme. These patterns suggest that it is robotically inserted, ready-made language—not reasoned deliberation—that drives estate planning “choices” about non-salient matters.


Figure 4: Standardization in Wills Drafted by Repeat-Playing Law Firms

Screen Shot 2018 02 05 at 7.15.56 PM


To conclude, there is a pathological relationship between boilerplate and simple default rules. Allowing shopworn text to displace a majoritarian default can do violence to a testator’s intent. Accordingly, as we discuss next, sticky defaults are a better choice for non-salient issues.  

B.     Sticky Defaults 

This Subpart argues that sticky defaults can ameliorate the boilerplate problem in wills law. It first considers conventional ways of making majoritarian defaults harder to displace, such as clear statement rules. It then urges courts and lawmakers to go further and experiment with exotic altering rules, like reason-giving and external badges of assent.

Sticky defaults are in vogue. Over the last decade, theorists have devised innovative regulatory techniques to “nudge” people into making smarter choices.209 One weapon in their arsenal is the sticky default.210 When policymakers believe that a default is majoritarian, they can guide individuals toward it by conditioning the displacement of the rule on compliance with onerous or obscure procedures.211 In this way, sticky defaults are a hybrid of immutable and default rules: They provide more freedom than iron-clad mandates, but also restrict autonomy by discouraging particular decisions.212

There is precedent for giving wills defaults additional armor. As noted, the best-known sticky default is the UPC and Restatement’s disregard of survivorship conditions. This tactic helps ensure that testators do not blithely disinherit branches of their family tree.213 As mentioned above, our data suggest that the need for this intervention may be less urgent than believed.214 Perhaps because survivorship is more likely to be on testators’ minds than debts and taxes, very few wills rely on the default rule of antilapse.215 Instead, even in a state like New Jersey that has eschewed the UPC and Restatement’s approach, most testators treat the antilapse statute as though it were a sticky default, stating precisely who takes the share of any predeceasing beneficiary.216

Conversely, testators’ unfamiliarity with representational schemes makes the default approach for dividing property among multi-generational classes a prime candidate for a sticky default. We were startled to discover that 82 of the 230 instruments (36%) in our sample select per stirpes instead of New Jersey’s background rubric of per capita at each generation. We find it hard to believe that the authors of these wills explained to their clients the subtleties of these rival systems of representation—a topic that even the best trusts and estates students need at least a week to grasp. Reinforcing this conclusion, an alarming eight of these 82 wills (10%) also contain a nonsensical devise to an individual beneficiary “per stirpes.” This carelessness suggests that both testators and their attorneys often pay little heed to language of representation. Consequently, the bar for drafting around a state’s default system for distributing assets among multigenerational classes should be higher. For instance, rather than merely reciting the empty phrases “per stirpes” or “per capita,” wills could be forced to spell out in plain English how the probate court should drop property down the lines of the family tree.

Likewise, states should make it harder to override the non-exoneration and equitable apportionment defaults. Here, again, the seeds of this approach can be found in existing law. In general, sticky defaults already govern the associated question of whether a “just debts” or tax apportionment provision in a will can affect nonprobate transfers (which pass outside the will).217 Suppose T leaves the residue of her estate to A, does not mention exoneration in her “just debts” clause, and requires the residue to pay estate and inheritance taxes. T then conveys her house—on which there is an unpaid mortgage—to B by trust, joint tenancy, or transfer-on-death deed. Who pays the loan and the death taxes on the real estate: A (the residuary beneficiary of the will) or B (who received the land outside of probate)? The answer is that there is a strong presumption that provisions in a will do not apply to nonprobate assets. The logic here is that because decedents frequently pass most of their wealth through nonprobate devices, forcing the residue to bear these expenses has the potential to disfigure an estate plan.218 For instance, a bare “just debts” clause does not “evinc[e] an intent to exonerate property passing outside probate.”219 Instead, a testator must “clearly and unambiguously” state that she wants the residuary beneficiaries to assume responsibility for the mortgage.220 Likewise, a mere “residue pays” apportionment clause in a will is not sufficient to make a residuary beneficiary pay death taxes on nonprobate assets.221 Only an “unequivocal direction” can “require the [residue] to bear the burden of taxation for property passing outside the will.”222 Thus, it would be a mere baby step for courts and lawmakers to extend these sticky defaults—which serve as a kind of sincerity check—from nonprobate transfers to their probate counterparts.

Admittedly, sticky defaults are no panacea. For one, because they hinge on the text of the will, they can be just as susceptible to being overridden by boilerplate as simple defaults. For instance, in states that follow the UPC and Restatement’s antilapse statute, there were reports that practitioners did not abandon boilerplate. Instead, they merely changed the content of their boilerplate. Instead of using cookie-cutter survivorship conditions—which were no longer sufficient to displace antilapse—they began to reflexively insert language that “entirely disclaims the antilapse statute.”223

Sticky defaults also increase the risk of ambiguity about a testator’s desires. For instance, courts have struggled to determine whether a “just debts” clause in a will is supposed to cover nonprobate assets. The Tennessee Supreme Court recently held that a testator did not mean to exonerate real property that passed by joint tenancy with the right of survivorship when his will directed his executor “to pay all my just debts and funeral expenses” but allowed “any installment debts secured by real estate . . . to be paid on an installment basis.”224 Conversely, the Oklahoma Supreme Court reached the opposite conclusion when faced with a provision that required “all my debts . . . [to] be paid by my executor, except that the payment of any debt secured by mortgage . . . may be postponed until payable by its terms.”225 As these opinions elucidate, increasing the proof required to rebut the default creates a great gray zone of uncertainty.

We found that this problem is particularly acute in the tax apportionment context. New Jersey has not clarified how specific testators must be to task the residue with paying death taxes on nonprobate transfers. A state statute provides that a tax apportionment clause in a will “shall be limited in its operation to the property passing thereunder unless the will . . . otherwise directs.”226 Yet the case law fails to establish what kind of language “directs” to the contrary.227 As a result, many tax apportionment clauses were ambiguous. Indeed, 59 wills in our sample (38%) gestured toward forcing the residuary beneficiaries to pay taxes on nonprobate transfers, but did not necessarily “evidence[] a clear intent to overcome the presumption of apportionment.”228 

Yet there are also several ways to address these challenges. One might be to force a testator to jump through unique textual hoops to reject a default. This rubric would make a particular rule the norm unless a testator employs language that is custom-tailored for her estate plan. The goal here would be to demand a reference that cannot be cut and pasted from a previous document. For example, a state could insist that a testator list each particular piece of property on which a residuary beneficiary would need to pay loans or taxes. This specificity requirement would weed out the kinds of vague statements that currently populate non-salient clauses. Moreover, the extra effort required to pinpoint each item of property that qualified for exoneration or special tax treatment could serve a cautionary function, ensuring that testators recognize the gravity of their choices.

Alternatively, jurisdictions could insist on an external manifestation of a testator’s assent, such as initialing next to a specific clause, to change a default rule.229 Many other areas of law employ this tool in an effort to call a party’s attention to particular language in a document. For example, some provisions in the Uniform Commercial Code, statutes that govern real estate contracts, and state arbitration laws tie the validity of a term to the non-drafter “separately sign[ing]” next to it.230 Likewise, representational schemes, non-exoneration, and equitable apportionment could become mandatory rules that can only be drafted around if the testator has specifically indorsed the contrary language.   

Of course, these separate signing rules only go so far. As anyone who has reflexively signed page after page of a mortgage or even clicked “I agree” when presented with a pop up box on a website knows too well, external badges of assent are often meaningless.231 It is easy to imagine that testators would rush through the separate signing process without giving due regard to what they were doing. Thus, assent-heightening defaults might be a step in the right direction, but they should not be the only step: It would make sense to use them in conjunction with a clear statement or plain English default.232 

Finally, one might object to our proposals on three other grounds. First, a naysayer might argue that adopting sticky defaults could thwart the wishes of low-income testators, who lack access to top-shelf counsel. As noted above, Adam Hirsch has cited this concern to contend that sticky defaults are a “trap[] for the unwary”233 and “have no place in our inheritance law.”234

However, even if sticky defaults make it harder for uninformed individuals to opt out, it does not follow that they frustrate testamentary intent. If lawmakers have done their job properly and crafted the underlying default to be majoritarian, then stickiness actually facilitates most decedents’ wishes. Indeed, rather than allowing indiscriminate boilerplate language in a will to override a majoritarian rule, it “nudges” wills in the right direction.235 

Second, one might wonder whether our proposals are destined for the same fate as the UPC and Restatement’s approach to survivorship conditions. As noted, these efforts to transform antilapse into a sticky default rule have fizzled.236 Likewise, lawmakers might see our embrace of clear statement rules and their ilk as a pretentious attempt to tell testators what they were really trying to say.237

We do not believe that this is an apt comparison. One reason why the UPC and Restatement’s rubric has been so divisive is that it governs a high-profile issue. Most testators give serious thought to whom should inherit if their front-line beneficiaries die before they do.238 As a result, ignoring the plain language of a survivorship condition can wreak havoc by driving a wedge between what a will says and what it means. Conversely, our interventions involve debts, taxes, and multi-generational classes.239 Because owners are more likely to defer to their estate planners on these topics, there is less of a risk of thwarting a testator’s intent by imposing a sticky default. And indeed, with little controversy, many jurisdictions have already adopted sticky defaults to govern the impact of just debts and tax apportionment provisions on nonprobate transfers.240 For these reasons, our thesis should be more palatable than the UPC and Restatement’s attempted revamp of antilapse.

Third, a critic could complain that sticky defaults increase transaction costs. This is not an idle concern in a sphere where drafters charge hundreds of dollars per hour. Arguably, it would be unfair to force testators with idiosyncratic preferences to pay their lawyer to explain their choices or allocate debts and taxes for every right or item in their estate.

But on closer inspection, it is not clear that sticky defaults would dramatically inflate the cost of legal services. As we have noted above, because wills often lie dormant for years—raising the specter of a testator moving to a different state or the law changing—estate planners try to avoid relying too heavily on default rules.241 Thus, rather than reducing fees by leaving strategic gaps, good practitioners already take the time and effort to spell out a testator’s wishes. As a result, drafting around a sticky default merely requires finessing existing language, not the full-fledged addition of a particular provision. As such, sticky defaults would encourage attorneys to become more competent and diligent.

Also, transaction costs are not quite the boogeyman in wills law that they are elsewhere. For instance, the will-execution process is already shot-through with time-consuming and laborious formality.242 In many states, a will must be written, signed by the testator, and also signed by two witnesses who, present at the same time, either saw the testator sign the will or acknowledge her previous signature.243 In fact, estate planners often go further and also have the testator sign or initial each page of the instrument. For instance, 136 of the 230 Sussex County wills (59%) took this additional step. There is a good reason for these extra precautions. Because estate planning is a singular activity—one that can deeply affect one’s friends and family on both pecuniary and psychological levels—it is imperative to get things right.244 Thus, even if sticky defaults do make the drafting process slightly more cumbersome, this detriment is well worth the price of keeping intent-defeating boilerplate in check.     

V.     Conclusion

Carrying out a decedent’s wishes is nothing less than “[t]he organizing principle of the American law of donative transfers . . . .”245 As a result, judges and legislators take pains to ensure that the field’s default rules dovetail with testators’ likely preferences. Yet our empirical study reveals that the widespread use of stock language in wills can prevent these efforts from taking root. This problem is particularly acute for non-salient matters such as representational schemes, exoneration of loans, and tax liability. Making these defaults stickier would shield them from the widespread use of boilerplate clauses that do not reflect a testator’s intent.


  1. [1]. See Will of Robert C. Paulson at 3 (Sussex Cty. Surr. Ct. Apr. 5, 2010) (on file with authors) [hereinafter Paulson Will].

  2. [2]. See id. at 1–2.

  3. [3]. See Application Probate at 1, In re Estate of Paulson (Sussex Cty. Surr. Ct. Apr. 2, 2015) (on file with authors).

  4. [4]. See N.J. Stat. Ann. § 3B:25-1 (West Supp. 2017); see also id. § 3B:24-4.

  5. [5]. Paulson Will, supra note 1, at 1. Courts often interpret a general reference to the testator’s “estate” to mean “residuary estate.” See infra note 118. And unless otherwise provided in the will, estate debts are paid from the residuary estate before being charged against specific devises. See infra note 98 and accompanying text.

  6. [6]. See generally, e.g., Ian Ayres, Empire or Residue: Competing Visions of the Contractual Canon, 26 Fla. St. U. L. Rev. 897 (1999) [hereinafter Ayres, Contractual Canon]; Ian Ayres & Robert Gertner, Filling Gaps in Incomplete Contracts: An Economic Theory of Default Rules, 99 Yale L.J. 87 (1989) [hereinafter Ayres & Gertner, Incomplete Contracts]; Ian Ayres & Robert Gertner, Majoritarian vs. Minoritarian Defaults, 51 Stan. L. Rev. 1591 (1999); Ian Ayres, Menus Matter, 73 U. Chi. L. Rev. 3 (2006); Ian Ayres, Regulating Opt-Out: An Economic Theory of Altering Rules, 121 Yale L.J. 2032 (2012) [hereinafter Ayres, Altering Rules]; Ian Ayres & Robert Gertner, Strategic Contractual Inefficiency and the Optimal Choice of Legal Rules, 101 Yale L.J. 729 (1992); Omri Ben-Shahar & John A.E. Pottow, On the Stickiness of Default Rules, 33 Fla. St. U. L. Rev. 651 (2006); David Charny, Hypothetical Bargains: The Normative Structure of Contract Interpretation, 89 Mich. L. Rev. 1815 (1991); Richard Craswell, Contract Law, Default Rules, and the Philosophy of Promising, 88 Mich. L. Rev. 489 (1989); Frank H. Easterbrook & Daniel R. Fischel, The Corporate Contract, 89 Colum. L. Rev. 1416 (1989); Charles J. Goetz & Robert E. Scott, The Limits of Expanded Choice: An Analysis of the Interactions Between Express and Implied Contract Terms, 73 Calif. L. Rev. 261 (1985); Jason Scott Johnston, Strategic Bargaining and the Economic Theory of Contract Default Rules, 100 Yale L.J. 615 (1990); Russell Korobkin, The Status Quo Bias and Contract Default Rules, 83 Cornell L. Rev. 608 (1998); Brett H. McDonnell, Sticky Defaults and Altering Rules in Corporate Law, 60 SMU L. Rev. 383 (2007); Ariel Porat & Lior Jacob Strahilevitz, Personalizing Default Rules and Disclosure with Big Data, 112 Mich. L. Rev. 1417 (2014); Eric A. Posner, There Are No Penalty Default Rules in Contract Law, 33 Fla. St. U. L. Rev. 563 (2006); Alan Schwartz & Robert E. Scott, The Common Law of Contract and the Default Rule Project, 102 Va. L. Rev. 1523 (2016) [hereinafter Schwartz & Scott, Default Rule Project]; Alan Schwartz & Robert E. Scott, Contract Theory and the Limits of Contract Law, 113 Yale L.J. 541, 598 (2003) [hereinafter Schwartz & Scott, Limits of Contract Law]; J.H. Verkerke, Legal Ignorance and Information-Forcing Rules, 56 Wm. & Mary L. Rev. 899 (2015); Lauren E. Willis, When Nudges Fail: Slippery Defaults, 80 U. Chi. L. Rev. 1155 (2013).

  7. [7]. See, e.g., Ayres & Gertner, Incomplete Contracts, supra note 6, at 89–90.

  8. [8]. See Richard A. Posner, Economic Analysis of Law 98 (6th ed. 1998).

  9. [9]. See, e.g., Anonymous, 2 N.C. (1 Hayw.) 243, 243 (Super. Ct. of Law & Eq. 1795) (per curiam).

  10. [10]. See infra Part III.B.

  11. [11]. See infra Part IV.A.

  12. [12]. See infra Part IV.A.

  13. [13]. See infra Part IV.B.

  14. [14]. See generally Margaret Jane Radin, Boilerplate: The Fine Print, Vanishing Rights, and the Rule of Law (2012) (exploring the normative issues raised by the widespread use of form terms).

  15. [15]. In re Greenfield Direct Response, Inc., 171 B.R. 848, 857 n.5 (Bankr. N.D. Ill. 1994) (citing Black’s Law Dictionary (6th ed. 1990)).

  16. [16]. See, e.g., Carter v. Countrywide Credit Indus., Inc., 189 F. Supp. 2d 606, 610 n.1 (N.D. Tex. 2002) (discussing the minimum wage established by the Fair Labor Standards Act).

  17. [17]. See, e.g., Hauer v. Union State Bank of Wautoma, 532 N.W.2d 456, 466 (Wis. Ct. App. 1995) (holding that there was sufficient evidence that the borrower lacked capacity to enter into loan).

  18. [18]. See Ayres, Contractual Canon, supra note 6, at 901.

  19. [19]. See Schwartz & Scott, Default Rule Project, supra note 6, at 1525.

  20. [20]. See id.

  21. [21]. See, e.g., Duncan v. TheraTx, Inc., 775 A.2d 1019, 1021 (Del. 2001) (“[D]efault damages rules, like other contract rules, should generally reflect the contract term that most parties would have bargained for at the time of the agreement.”).

  22. [22]. See supra note 6.

  23. [23]. Ayres & Gertner, Incomplete Contracts, supra note 6, at 91.

  24. [24]. For example, Ayres and Gertner cite U.C.C. § 2-201 (Am. Law Inst. & Unif. Law Comm’n 1976), which refuses to enforce a purported contract unless the parties state the quantity of goods sold. See id. at 95–97. As Ayres and Gertner explain, this penalty default serves a valuable purpose “because it is cheaper for the parties to establish the quantity term beforehand than for the courts to determine after the fact what the parties would have wanted.” Id. at 96.

  25. [25]. Schwartz & Scott, Default Rule Project, supra note 6, at 1532.

  26. [26]. See id. at 1527–29.

  27. [27]. See id. at 1530; see also Schwartz & Scott, Limits of Contract Law, supra note 6, at 594 (“Firms would prefer the state not to create inefficient defaults because firms will contract out of them; thus, the only effect these defaults will have is to increase transaction costs.”).

  28. [28]. Ayres, Altering Rules, supra note 6, at 2032.

  29. [29]. See id. at 2084–96. One potential source of confusion is that the default rule literature uses “sticky” to signify two different things. First, some commentators describe default rules as “sticky” when they have a tendency to govern even when parties might prefer some other principle. See, e.g., Ben-Shahar & Pottow, supra note 6, at 651 (describing a “sticky” default as one that the “parties might choose not to opt out of ... even when a better provision can easily be identified” (emphasis omitted)). Second, “sticky” sometimes means a default rule that is simply “hard[er] to opt out of.” McDonnell, supra note 6, at 385.

  30. [30]. See Ayres, Altering Rules, supra note 6, at 2076–77, 2100, 2112–13.

  31. [31]. See, e.g., Adam J. Hirsch, Default Rules in Inheritance Law: A Problem in Search of Its Context, 73 Fordham L. Rev. 1031, 1033 (2004) (“[S]cholarship on inheritance law has largely neglected to explore, or even to notice, the potential relevance of general default rule theory ... .”);
    cf. generally John H. Langbein, Mandatory Rules in the Law of Trusts, 98 Nw. U. L. Rev. 1105 (2004) (surveying mandatory rules in trust law and concluding that they either prevent settlors from harming the interests of the beneficiaries or facilitate settlors’ likely intent).

  32. [32]. See Adam J. Hirsch, Incomplete Wills, 111 Mich. L. Rev. 1423, 1438–39 (2013). See generally Hirsch, supra note 31 (considering how to formulate default rules in wills law). In roughly the same vein, Naomi Cahn and Amy Ziettlow have argued that the default rules governing child legitimacy and revocation of a will upon divorce may not comport with majoritarian preferences. See generally Naomi Cahn & Amy Ziettlow, “Making Things Fair”: An Empirical Study of How People Approach the Wealth Transmission System, 22 Elder L.J. 325, 332–70 (2015) (noting discrepancies between the assumptions that underlie these doctrines and preferences expressed by interviewees in Baton Rouge, Louisiana). Likewise, Ariel Porat and Lior Jacob Strahilevitz begin their fascinating article about using big data to personalize default rules with an example of using “readily observable characteristics (e.g., wealth, health, time of marriage, age of children, and occupation) that could predict default rules in intestacy for population subgroups.” Porat
    & Strahilevitz, supra note 6, at 1419.

  33. [33]. Hirsch, supra note 31, at 1033.

  34. [34]. See id. at 1039.

  35. [35]. See id. at 1040.

  36. [36]. Id. at 1039 (emphasis omitted).

  37. [37]. See id. at 1039–42.

  38. [38]. See id. at 1040–41.

  39. [39]. See id. at 1041.

  40. [40]. See id. Ascertaining majoritarian preferences about testamentary intent presents its own challenges. See id. at 1069 (“The mind of a decedent is the ultimate sanctum sanctorum. It refuses to yield itself to view.”).

  41. [41]. Id. at 1033.

  42. [42]. Id. at 1042 (quoting 4 James Kent, Commentaries on American Law 537 (photo. reprint 1971) (1826–1830)).

  43. [43]. Id. at 1060.

  44. [44]. See Hirsch, supra note 32, at 1438.

  45. [45]. See Hirsch, supra note 31, at 1061.

  46. [46]. Shelly Kreiczer-Levy, Deliberative Accountability Rules in Inheritance Law: Promoting Accountable Estate Planning, 45 U. Mich. J.L. Reform 937, 938 (2012).

  47. [47]. Id. at 951.

  48. [48]. See id. at 954–57.

  49. [49]. Kreiczer-Levy explains that some pretermitted child statutes are intent-facilitating while others are intent-defeating. See id. at 955–57. For example, the UPC’s pretermission rule applies only when a testator makes a will, then has a child, and then dies without updating her estate plan. See, e.g., Unif. Prob. Code § 2-302 (amended 2010) (regulating the situation in which “a testator fails to provide in his [or her] will for any of his [or her] children born or adopted after the execution of the will” (alterations in original)). The logic of giving the pretermitted child a share of the estate is to carry out the decedent’s likely intent by recognizing “that the omission was unintentional.” Dotson v. Dotson, 372 S.W.3d 398, 401 (Ark. Ct. App. 2009) (quoting Robinson v. Mays, 610 S.W.2d 885, 887 (Ark. 1981)). Conversely, other states require testators to expressly disinherit children. See Kreiczer-Levy, supra note 46, at 957. Kreiczer-Levy contends that this mandate prompts testators to take accountability by addressing the matter directly. See id.

  50. [50]. Kreiczer-Levy, supra note 46, at 958–62.

  51. [51]. See id. at 957.

  52. [52]. Id. at 964.

  53. [53]. See generally Hirsch, supra note 31 (focusing largely (albeit not exclusively) on intestacy). Hirsch has also addressed the issue of partial intestacy that occurs when a will makes an incomplete disposition of the testator’s estate. See generally Hirsch, supra note 32 (analyzing the problem of wills that fail to devise all of the testator’s property).

  54. [54]. See Kreiczer-Levy, supra note 46, at 954–62.

  55. [55]. See Edwin W. Patterson, The Delivery of a Life-Insurance Policy, 33 Harv. L. Rev. 198, 222 (1919).

  56. [56]. See id.

  57. [57]. See Edwin W. Patterson, The Interpretation and Construction of Contracts, 64 Colum. L. Rev. 833, 856 (1964) (citing Raymond Saleilles, De la Déclaration de Volonté § 89, at 229–30 (1901)).

  58. [58]. Patterson, supra note 55, at 222.

  59. [59]. See, e.g., Fed. Motor Truck Sales Corp. v. Shanus, 250 N.W. 713, 714 (Minn. 1933).

  60. [60]. See, e.g., W. David Slawson, Mass Contracts: Lawful Fraud in California, 48 S. Cal. L. Rev. 1, 9 (1974).

  61. [61]. See, e.g., Lichten v. E. Air Lines, Inc., 87 F. Supp. 691, 697 (S.D.N.Y. 1949), aff’d sub nom., Lichten v. E. Airlines, Inc., 189 F.2d 939 (2d Cir. 1951).

  62. [62]. See generally David Horton & Andrea Cann Chandrasekher, After the Revolution: An Empirical Study of Consumer Arbitration, 104 Geo. L.J. 57 (2015) (discussing the spread of mandatory arbitration); Judith Resnik, Diffusing Disputes: The Public in the Private of Arbitration, the Private in Courts, and the Erasure of Rights, 124 Yale L.J. 2804 (2015) (surveying the shift away from public courts to private dispute resolution).

  63. [63]. Friedrich Kessler, Contracts of Adhesion—Some Thoughts About Freedom of Contract, 43 Colum. L. Rev. 629, 632 (1943); see Karl N. Llewellyn, What Price Contract?—An Essay in Perspective, 40 Yale L.J. 704, 731 (1931) (“Law, under the drafting skill of counsel, now turns out a form of contract which resolves all questions in advance in favor of one party to the bargain. It is a form of contract which ... amounts to the exercise of unofficial government of some by others ... .”).

  64. [64]. Eyal Zamir, Contract Law and Theory: Three Views of the Cathedral, 81 U. Chi. L. Rev. 2077, 2096 (2014). For a handful of notable contributions to this debate, see generally Margaret Jane Radin, Boilerplate: The Fine Print, Vanishing Rights, and the Rule of Law (2012); Ian Ayres & Alan Schwartz, The No-Reading Problem in Consumer Contract Law, 66 Stan. L. Rev. 545 (2014); Douglas G. Baird, The Boilerplate Puzzle, 104 Mich. L. Rev. 933 (2006); Randy E. Barnett, Consenting to Form Contracts, 71 Fordham L. Rev. 627, 638 (2002); Kessler, supra note 63; Todd D. Rakoff, Contracts of Adhesion: An Essay in Reconstruction, 96 Harv. L. Rev. 1174 (1983); W. David Slawson, Standard Form Contracts and Democratic Control of Lawmaking Power, 84 Harv. L. Rev. 529 (1971).

  65. [65]. Cf. David Horton, Unconscionability in the Law of Trusts, 84 Notre Dame L. Rev. 1675, 1724–31 (2009) (exploring the use of boilerplate in trusts, but focusing largely on terms that a drafter deliberately inserted to try to expand her own power).

  66. [66]. See, e.g., Gregory H. Shill, Boilerplate Shock: Sovereign Debt Contracts as Incubators of Systemic Risk, 89 Tul. L. Rev. 751, 765 (2015).

  67. [67]. See, e.g., Lawrence P. Keller, Wills § 1:3 (2016) (stating that “[t]here are several ‘Articles’ which, with subtle variations, should be components of virtually every Will” and providing standardized forms for each important “Article”).

  68. [68]. See Michael Klausner, Corporations, Corporate Law, and Networks of Contracts, 81 Va. L. Rev. 757, 761 (1995); see also Mark A. Lemley & David McGowan, Legal Implications of Network Economic Effects, 86 Calif. L. Rev. 479, 565 (1998) (observing that “frequent adoption of a given term will increase the probability that the term will be subject to future interpretation by courts” which, in turn, “will presumably lead to greater clarity in potentially ambiguous terms”).

  69. [69]. See, e.g., Marcel Kahan & Michael Klausner, Standardization and Innovation in Corporate Contracting (or “The Economics of Boilerplate”), 83 Va. L. Rev. 713, 722, 726–27 (1997) (“Judicial opinions can reduce uncertainty regarding the validity and meaning of a term and the interaction of the term with relevant legal requirements, such as those contained in corporate, securities, and bankruptcy laws.”).

  70. [70]. See Mark Weidemaier et al., Origin Myths, Contracts, and the Hunt for Pari Passu, 38 Law
    & Soc. Inquiry 72, 82 (2013). For instance, Weidemaier et al. interviewed leading corporate attorneys who recounted multiple conflicting explanations (or origin myths) for the persistent inclusion of an obscure problematic term (the “pari passu” clause) in sovereign debt instruments. Id. at 72–73. Many attributed the pari passu clause to the inattentive copying of terms from one instrument to the next. See id. at 77–78. But closer examination revealed that its language had evolved over time and, in fact, served salutary purposes. See id. at 84–85.

  71. [71]. Alan Scott Rau, Arbitral Power and the Limits of Contract: The New Trilogy, 22 Am. Rev. Int’l Arb. 435, 521 n.285 (2011).

  72. [72]. Joshua Fairfield, The Cost of Consent: Optimal Standardization in the Law of Contract, 58 Emory L.J. 1401, 1405 (2009).

  73. [73]. See id. at 1426.

  74. [74]. This survival requirement reflects the plausible assumption that testators generally want property to pass directly to a living beneficiary rather than through the estate of a predeceased beneficiary. See Restatement (Third) of Prop.: Wills & Donative Transfers § 1.2 cmt. a (Am. Law Inst. 1999) (“[A]n individual who fails to survive the decedent cannot take as a devisee ... .”).

  75. [75]. See id. § 5.5 (“Antilapse statutes typically provide, as a rebuttable rule of construction, that devises to certain relatives who predecease the testator pass to specified substitute takers, usually the descendants of the predeceased devisee who survive the testator.”).

  76. [76]. See id.

  77. [77]. In this way, antilapse tries to prevent unintentional disinheritance of a branch of the testator’s family. For instance, in the hypothetical above, T probably assumed that the watercolor would eventually pass from A to B when A died, which is precisely what antilapse accomplishes. See id. reporter’s note n.1 (noting that antilapse statutes “prevent unintended disinheritance of one or more lines of descent, by presumptively creating an alternative or substitute gift in favor of the descendants of certain of the decedent’s predeceased relatives” (quoting Edward C. Halbach, Jr. & Lawrence W. Waggoner, The UPC’s New Survivorship and Antilapse Provisions, 55 Alb. L. Rev. 1091, 1099 (1992))).

  78. [78]. See id. cmt. f (“Antilapse statutes establish a strong rule of construction, designed to carry out presumed intention.”).

  79. [79]. See, e.g., Mark L. Ascher, The 1990 Uniform Probate Code: Older and Better, or More Like the Internal Revenue Code?, 77 Minn. L. Rev. 639, 651 (1993) (“[M]ost estate planners believe that if a bequest contains language such as, ‘if he survives me,’ the antilapse statute cannot apply.”).

  80. [80]. See, e.g., Restatement (Third) of Prop.: Wills & Donative Transfers § 5.5 cmt. h (Am. Law Inst. 1999) (“An often litigated question is whether language requiring the devisee to survive the testator, without more, constitutes a sufficient expression of a contrary intent to defeat the antilapse statute. The majority view is that such language signifies a contrary intent.”).

  81. [81]. Restatement (Third) of Prop.: Wills & Donative Transfers § 5.5 cmt. h (Am. Law Inst. 1999). The drafters of the UPC echoed this concern:

    Another objection to applying the antilapse statute is that mere words of survivorship somehow establish a contrary intention. The argument is that attaching words of survivorship indicates that the testator thought about the matter and intentionally did not provide a substitute gift to the devisee’s descendants. At best, this is an inference only, which may or may not accurately reflect the testator’s actual intention. An equally plausible inference is that the words of survivorship are in the testator’s will merely because the testator’s lawyer used a will form with words of survivorship.

    Unif. Prob. Code § 2-603 cmt. (amended 2010) (Contrary Intention-the Rationale of Subsection (b)(3)).

  82. [82]. Unif. Prob. Code § 2-603(b)(3) (emphasis added); see also Halbach & Waggoner, supra note 77, at 1101–03 (linking the UPC’s approach to the assumption that “[w]hen a deceased child leaves children or more remote descendants, most parents would not want to disinherit that child’s line of descent”).

  83. [83]. See, e.g., Alaska Stat. § 13.12.603(a)(3) (2016); Colo. Rev. Stat. § 15-11-603(2)(c) (2016); Haw. Rev. Stat. Ann. § 560:2-603(b)(3) (LexisNexis 2006); Mich. Comp. Laws Ann.
    § 700.2603(1)(c) (West 2002); Mont. Code Ann. § 72-2-613(2)(c) (West 2009); N.M. Stat. Ann.
    § 45-2-603B(3) (West 2014); Ruotolo v. Tietjen, 890 A.2d 166, 173, 177 (Conn. App. Ct. 2006).

  84. [84]. Ascher, supra note 79, at 654.

  85. [85]. See, e.g., Minn. Stat. Ann. § 524.2-603(2) (West 2012) (“[W]ords of survivorship ... are a sufficient indication of an intent contrary to the application of [the antilapse statute].” (emphasis added)); S.C. Code Ann. § 62-2-603(C) (2009 & Supp. 2013); Tex. Est. Code Ann.
    § 255.151 (West 2014); Utah Code Ann. § 75-2-603(2)(c) (LexisNexis 1993 & Supp. 2016); see also McGowan v. Bogle, 331 S.W.3d 642, 646 (Ky. Ct. App. 2011) (rejecting the claim that the 1990 UPC amendments “warrant a deviation from ... [traditional] law”).

  86. [86]. In addition, there are other dark flickers in the case law of a boilerplate problem in wills. See, e.g., Minary v. Citizens Fid. Bank & Tr. Co., 419 S.W.2d 340 (Ky. 1967). Amelia Minary’s 1932 testamentary trust requires its property to eventually be given “to [her] then surviving heirs, according to the laws of descent and distribution then in force in Kentucky.” Id. at 341 (emphasis added). This provision is odd because it pins the distribution of her assets to the content of future rules, which Amelia could only guess at, rather than existing rules, which she would have known. Nevertheless, reported opinions in Kentucky reveal at least six other litigated trusts containing the same or nearly identical language. This suggests that the unusual provision had been mindlessly copied by attorneys in the region from one instrument to the next. See, e.g., Ky. Tr. Co. v. Sweeney, 163 F. Supp. 450, 451 (W.D. Ky. 1958); Bedinger v. Graybill’s Ex’r & Tr., 302 S.W.2d 594, 596 (Ky. 1957); Commonwealth v. Fid. & Colum. Tr. Co., 146 S.W.2d 3, 4 (Ky. 1940); Zinsmeister’s Tr. v. Long, 61 S.W.2d 887, 888 (Ky. 1933); Cecil’s Ex’rs v. Anhier, 195 S.W. 837, 839 (Ky. 1917); Barclay’s Tr. v. Commonwealth, 161 S.W. 510, 511 (Ky. 1913).

  87. [87]. Anderson v. Menefee, 174 S.W. 904, 905 (Tex. Civ. App. 1915). For instance, George Washington’s will contains the following language: “All my [deb]ts, of which there are but few, and none of magnitude, are to be punctu[al]ly and speedily paid.” George Washington’s Last Will and Testament (July 9, 1799),

  88. [88]. Anonymous, 2 N.C. (1 Hayw.) 243, 243 (Super. Ct. of Law & Eq.1795) (per curiam).

  89. [89]. In re Pamela Andreas Stisser Grantor Tr., 818 N.W.2d 495, 503 (Minn. 2012).

  90. [90]. See, e.g., François-Xavier Martin, Martin’s Treatise on the Powers and Duties of Executors and Administrators: According to the law of North Carolina 101 (1820).

  91. [91]. Bacon v. Pomeroy, 104 Mass. 577, 585 (1870).

  92. [92]. In re Reynolds’ Estate, 109 A. 60, 62 (Vt. 1920).

  93. [93]. Perkins v. Corkey, 159 A.2d 166, 168 (Conn. 1960).

  94. [94]. In re Mayer’s Estate, 137 A. 627, 629 (Pa. 1927).

  95. [95]. For example, in a rash of cases beginning in the late 1700s, creditors tried to collect obligations from the estate that should have been barred by the statute of limitations. See, e.g., Peck v. Botsford, 7 Conn. 172, 181 (1828); Collamore v. Wilder, 19 Kan. 67, 82 (1877); Jones’ Ex’r v. Jones, 122 S.W.2d 779, 781 (Ky. 1938); Bacon, 104 Mass. at 585; Rogers v. Rogers, 3 Wend. 503, 503 (N.Y. 1829); Anonymous, 2 N.C. (1 Hayw.) 243, 243 (Super. Law & Eq. 1795) (per curiam); Smith v. Porter, 1 Binn 209, 209 (Pa. 1807); Johnston v. Wilson’s Adm’r, 70 Va. 379, 387 (1877). These individuals and entities argued that a “just debts” clause revealed that the testator wanted to honor all “just” claims—those that were well-founded and authentic—whether they were timely or not. See, e.g., Collamore, 19 Kan. at 71; Jones, 122 S.W.2d at 780. As one party contended: “[i]t is necessary to give some operation to the words adopted by the testator ... ; and if they have not this effect they have none.” Smith, 1 Binn at 209 (emphasis added). A few courts were persuaded. See, e.g., Anonymous, 2 N.C. at 243 (reasoning that a “just debts” clause requires “the payment of all just debts, whether recoverable at law or not” (citation omitted)). But most reached the opposite conclusion, reasoning that a “just debts” provision was nothing more than a kind of linguistic ritual:

    The testator adopts this language . . . in compliance with a custom almost universal, and, perhaps, having its origin in the solemnity, which attends a final disposition of his earthly concerns. It is not credible, that he thereby intends to direct the payment of any particular debt; much less, to deprive his representative of the right of interposing a legal defen[s]e . . . .

    Peck, 7 Conn. at 176–77, 181 (“[A] clause in a will, directing all just debts to be paid, . . . will [not] defeat the operation of a beneficial statute . . . .”); see, e.g., Jones, 122 S.W.2d at 780; Smith, 1 Binn at 209. Thus, even though “just debts” clauses “ha[ve] no authoritative force,” Collamore, 19 Kan. at 82, they proved to be a fount of expensive and time-consuming litigation.

  96. [96]. See, e.g., Hannibal Tr. Co. v. Elzea, 286 S.W. 371, 376 (Mo. 1926); Ruston’s Ex’rs v. Ruston, 2 Yeates 54, 54 (Pa. 1796); Estate of Fussell v. Fortney, 730 S.E.2d 405, 410–11 (W. Va. 2012); Note, Exoneration of Specific Property from Incumbrances Existing at the Death of the Testator or Ancestor, 40 Harv. L. Rev. 630, 631 (1927).

  97. [97]. See Estate of Fussell, 730 S.E.2d at 409.

  98. [98]. See, e.g., Hannibal, 286 S.W. at 378 (“No doubt the common-law [sic] rule grew out of the feudal system of land ownership in England ... .”); Smith v. Wilson, 81 A. 851, 853 (N.J. Ch. 1911) (“[T]he policy of the English law which protected the heir and had tender regard for landed estates ... .”).

  99. [99]. See, e.g., Mass. Gen. Laws Ann. ch. 191, § 23 (West 2012) (repealed 2008); Neb. Rev. Stat. § 30-2347 (2016); N.Y. Est. Powers & Trusts Law § 3-3.6(a) (McKinney 2012); Okla. Stat. Ann. tit. 46, § 5 (West 2014); S.D. Code § 56.0227 (1939); Wash. Rev. Code § 11.12.070 (2016); Succession of Rabasse, 17 So. 597, 597–98 (La. 1895).

  100. [100]. By default, estate debts are charged against the residuary estate before specific devises. Unif. Prob. Code § 3-902(a) (amended 2010).

  101. [101]. In re Cline’s Estate, 227 P.2d 157, 162 (Kan. 1951); cf. In re Wilson’s Estate, 66 Pa. D.
    & C. 308, 313 (Orphans’ Ct. 1949) (“A direction to pay debts is often given great weight.”). Likewise, several jurisdictions continue to follow the exoneration doctrine. In these states, a “just debts” clause can reinforce the assumption that a testator wanted the recipient of a specific bequest of real property to take it without any unpaid mortgage balance. See, e.g., Wilson v. Smith, 360 S.W.2d 78, 88 (Tenn. Ct. App. 1962); Estate of Fussell, 730 S.E.2d at 410–11; cf. Lemp v. Keto, 678 A.2d 1010, 1019 (D.C. 1996) (applying the exoneration doctrine even though will did not contain a “just debts” clause).

  102. [102]. Meyer v. Cahen, 18 N.E. 852, 853 (N.Y. 1888); see also In re Porter, 72 P. 173, 174 (Cal. 1903) (reasoning that just debts clauses “are much like the formal, meaningless terms of endearment and pious phrases printed in the formal part of blanks for making wills”); Caruthers v. Buscher, 382 A.2d 608, 615 (Md. Ct. Spec. App. 1978) (“If a particular piece of realty is to be exonerated, the Will must make expressly clear that that piece of realty is, indeed, to be exonerated.”); Sav. Tr. Co. of St. Louis v. Beck, 73 S.W.2d 282, 286 (Mo. Ct. App. 1934) (holding that just debts clause “is insufficient to show an intention to exonerate the specifically devised property from the mortgage debt”). The Uniform Probate Code has endorsed this position, providing that “[a] specific devise passes subject to any mortgage interest existing at the date of death, without right of exoneration, regardless of a general directive in the will to pay debts.” Unif. Prob. Code § 2-607 (amended 2010). A handful of states have adopted this rule wholesale. See, e.g., Ala. Code § 43-8-228 (1982); Alaska Stat. § 13.12.607 (2016); Colo. Rev. Stat. § 15-11-607 (2016); Haw. Rev. Stat. § 560:2-607 (LexisNexis 2006); Me. Rev. Stat. Ann. tit. 18-A § 2-609 (2012); Mass. Gen. Laws Ann. ch. 190B § 2-607 (West 2012); S.D. Codified Laws § 29A-2-607 (2004); Utah Code Ann. § 75-2-607 (LexisNexis 2016). Others have adopted similar, but not identical, laws. See, e.g., Neb. Rev. Stat. § 30-2347 (2016); N.M. Stat. Ann.
    § 45-2-607 (West 2014); N.Y. Est. Powers & Trusts Law § 3-3.6(a) (McKinney 2012).

  103. [103]. Nawrocki’s Estate v. Kirkpatrick, 268 P.2d 363, 367 (Or. 1954); see also Ashkenazy v. Estate of Ashkenazy, 140 So. 2d 331, 336 (Fla. Dist. Ct. App. 1962) (reasoning that a just debts clause “does not, of itself, indicate a positive intent to exonerate the real property”).

  104. [104]. See, e.g., Eugene F. Scoles & Richard B. Stephens, The Proposed Uniform Estate Tax Apportionment Act, 43 Minn. L. Rev. 907, 915 (1959).

  105. [105]. See id.

  106. [106]. See Unif. Estate Tax Apportionment Act of 1958 § 2 (Unif. Law Comm’n 1958). Estate tax apportionment is generally a question of state law. See Riggs v. Del Drago, 317 U.S. 95, 102 (1942). However, the Internal Revenue Code also establishes a default rule of equitable apportionment for beneficiaries receiving the proceeds of life insurance, property subject to a general power of appointment, or certain qualified terminable interest property. See I.R.C.
    §§ 2205–07 (2016).

  107. [107]. See, e.g., Ark. Code Ann. § 28-54-104 (2016); Cal. Prob. Code § 20110 (West 2011); Colo. Rev. Stat. § 15-12-916 (2016); Conn. Gen. Stat. Ann. § 12-401(a) (West 2008); Del. Code Ann. tit. 12 § 2901 (2007); Fla. Stat. Ann. § 733.817 (West 2017); La. Stat. Ann. § 9:2432 (West 2016); Md. Code Ann., Tax–Gen. § 7-308 (West 2010); Miss. Code. Ann. § 27-10-7 (West 2012); Mont. Code Ann. § 72-16-603 (West 2009); Neb. Rev. Stat. Ann. § 77-2108 (West 2016); Nev. Rev. Stat. Ann. § 150.310 (LexisNexis 2014); N.H. Rev. Stat. Ann. § 88-A-2 (West 2016); N.J. Stat. Ann. § 3B:24-4 (West 1983 & Supp. 2017); N.Y. Est. Powers & Trusts Law § 2-1.8 (McKinney 2012 & Supp. 2017); N.D. Cent. Code Ann. § 30.1-20-16 (2010); Ohio Rev. Code Ann. § 2113.86 (West 2014 & Supp. 2017); Or. Rev. Stat. Ann. § 116.313 (West 2016); S.D. Codified Laws § 29A-3-916 (2004 & Supp. 2017); Tex. Est. Code Ann. § 124.005 (West 2014); Utah Code Ann. § 75-3-916 (LexisNexis 1993); Va. Code Ann. § 64.2-540 (West 2012); Wyo. Stat. Ann. § 2-10-103 (2015); see also Estate of Mcvey v. Dep’t of Revenue, 480 S.W.3d 233, 242 (Ky. 2015) (“Ordinarily, inheritance taxes are paid out of the shares received by the beneficiaries.”); In re Estate of Roe v. Gray, 426 N.W.2d 797, 799–800 (Mich. Ct. App. 1988) (“There is said to be a strong policy in favor of apportionment and the burden of proof rests with those arguing against it.”). But see In re Estate of Maierhofer, 767 N.E.2d 850, 852 (Ill. App. Ct. 2002) (continuing to apply the “burden on the residue” rule).

  108. [108]. See, e.g., Daniel B. Evans, Tax Clauses to Die For, 20 Prob. & Prop., July/Aug. 2006, at 39.

  109. [109]. See id. at 40.

  110. [110]. Cf. Reid Kress Weisbord, Essay, The Advisory Function of Law, 90 Tul. L. Rev. 129, 155 (2015) (“[T]he residuary clause constitutes a central organizing component of nearly all donative instruments ... .”).

  111. [111]. See I.R.C. § 2056 (2016).

  112. [112]. See, e.g., Mark R. Siegel, Who Should Bear the Bite of Estate Taxes on Non-Probate Property?, 43 Creighton L. Rev. 747, 751 (2010).

  113. [113]. Evans, supra note 108, at 38.

  114. [114]. See In re Estate of Kuralt (Kuralt I), 15 P.3d 931, 932 (Mont. 2000).

  115. [115]. In re Estate of Kuralt (Kuralt II), 68 P.3d 662, 666 (Mont. 2003) (omission in original).

  116. [116]. Kuralt I, 15 P.3d at 932.

  117. [117]. See Kuralt II, 68 P.3d at 663–64 (observing that honoring the tax apportionment clause in the 1994 will led to “adverse tax consequences” related to “the marital deduction”).

  118. [118]. See id. at 668.

  119. [119]. Johnson v. Hall, 392 A.2d 1103, 1107 (Md. 1978).

  120. [120]. Lynchburg Coll. v. Cent. Fid. Bank, 410 S.E.2d 617, 620–21 (Va. 1991); see Estate of Mumby v. Caldwell, 982 P.2d 1219, 1227 (Wash. Ct. App. 1999) (“The vast majority of courts ... have concluded that when the non-specific tax clause is grouped along with a provision for the payment of debts and other expenses of administration of the estate, the result is to shift the tax burden to the same fund ... .”); see also Estate of Semmes v. Comm’r of Internal Revenue, 288 F.2d 664, 665 (6th Cir. 1961) (applying Tennessee law); Morris v. Dosch, 106 S.W.2d 159, 160 (Ark. 1937) (opining that this result “is in exact compliance with the will”); Starr v. Watrous, 165 A. 459, 460 (Conn. 1933) (reasoning that if the testator “provide[s] for payment of the tax in like manner as debts of the estate and funeral expenses, ... the tax ultimately comes out of the residuary estate”); Univ. of Louisville v. Liberty Nat’l Bank & Tr. Co., 499 S.W.2d 288, 289 (Ky. 1973) (explaining that a just debts clause “would be pure surplusage or ‘boilerplate’ if not construed as evidencing the intent that the taxes be paid ‘off the top’”); Succession of Jones, 172 So. 2d 312, 315 (La. Ct. App. 1965) (noting that the testator “directed his executors to pay the taxes mentioned along with his debts”); In re Estate of Morris, 838 P.2d 402, 404–05 (Mont. 1992) (holding that a just debts clause overrode the equitable apportionment rule); Gaither v. U.S. Tr. Co. of N.Y., 97 S.E.2d 24, 26 (S.C. 1957) (“[W]here there is a general direction to the executor to pay all debts, costs of administration and taxes, there is an implied direction that the taxes are to be paid from the fund which also bears the burden of debts and expenses of administration.”).

  121. [121]. Univ. of Louisville, 499 S.W.2d at 289.

  122. [122]. First Nat’l Bank of Morgantown v. McGill, 377 S.E.2d 464, 469–70 (W. Va. 1988); see also In re Estate of Carrington, 136 N.E.2d 182, 185 (Ohio Prob. Ct. 1956) (“Nearly all wills have stock language of varying types requiring the payment of debts, taxes and costs of administration.”).

  123. [123]. Johnson, 392 A.2d at 1109–10 (Md. 1978); see also In re Grondin’s Estate, 100 A.2d 160, 163 (N.H. 1953) (explaining that a will must “definitely show a testamentary intent to impose the inheritance tax upon the residuary estate and this intent is not to be inferred from a doubtful phrase or word in the will” (citing Sherman v. Moore, 93 A. 241 (Conn. 1915))).

  124. [124]. See Urbitran Assocs., Sussex County Ten-Year Mobility Study 11, http://www.sussex.nj.

  125. [125]. Quick Facts, Sussex County, New Jersey, U.S. Census Bureau,
    quickfacts/table/PST045215/34037 (last visited Oct. 23, 2017).

  126. [126]. The overall median household income in the United States is $53,889. Quick Facts, United States, U.S. Census Bureau, (last visited Oct. 23, 2017). In addition, the testators in our data tended to live in relatively stable nuclear families. Most either were or had been married: 111 were widowed (48%), 80 were the first spouse in the couple to die (35%), 26 were divorced (11%), and 12 had always been single (5%). Two hundred (87%) had at least one child. The average age at death ranged from 41 to 106, with a mean of 84.

  127. [127]. The Uniform Law Commission unveiled an ambitious revision of the UPC in 1990. See Unif. Probate Code (amended 2010); see generally Ascher, supra note 79 (criticizing several of the new rules). However, many of its most innovative provisions proved controversial. See, e.g., id.

  128. [128]. See, e.g., Glenn A. Henkel et al., The New New Jersey Probate Code, N.J. Law.: Wkly. Newspaper, May 2, 2005, at A3, 2005 WLNR 27258532.

  129. [129]. A matter was included in our sample if it was coded in the Surrogate Court’s electronic database with a “date of proceedings” between February 1, 2015 and July 31, 2015. “The date of proceedings” refers to the day on which the Surrogate signs the Judgment appointing a personal representative for the decedent’s estate. See E-mail from Jeanne Woodhouse, Sussex Cty. Surrogate, to Jordan Doppelt, Rutgers Univ. (Aug. 26, 2015, 9:05 AM) (on file with authors).We are especially grateful for Surrogate Chiusano’s assistance because, even though probate files are a matter of public record, it can be difficult for researchers (and burdensome for court officials) to reproduce the critical mass of probate files necessary for research. Cf. Reid Kress Weisbord, The Connection Between Unintentional Intestacy and Urban Poverty, Rutgers L. Rev. Comments., Aug. 22, 2012, at 1, 7,
    archive/commentaries/2012/Weisbord_TheConnectionBetweenUnintentionalIntestacyAndUrbanPoverty.pdf (describing the difficulty of obtaining probate files from the public record in New Jersey and constitutional challenges to some surrogate courts that charge five dollars per page). For our project, we paid the statutory rate of $0.05 per page of public records reproduced.

  130. [130]. See Michael J. Gau, A Practical Guide to Estate Planning and Administration 61 (2005).

  131. [131]. See id.

  132. [132]. We coded any estate with a codicil as though the will had been executed on the date of the codicil (not the original will). We did so for two reasons: codicils (1) formally republish wills as a matter of black-letter law, see First Nat’l Bank & Tr. Co. v. Baker, 1 A.2d 283, 286 (Conn. 1938); and (2) also give the testator the chance to update her estate plan in light of new legal developments.

  133. [133]. See N.J. Stat. Ann. § 3B:2-5 (West 2007). Despite this exclusion, some cases in our sample involved disputes. See, e.g., Verified Complaint for Removal of Administrator
    & Appointment of Successor Co-Administrators, In re Estate of Vozza, No. SSX-P299-15 (N.J. Super. Ct. Ch. Div. Mar. 9, 2015) (on file with author). These matters were then transferred to the Superior Court Chancery Division for resolution of the conflict. See id.

  134. [134]. See Civil Practice Div., Admin. Office of the Courts, New Jersey Probate Part and Surrogate’s Court Statistical Reference Guide 1–2 (2008),

  135. [135]. See E-mail from Jeanne Woodhouse, Sussex Cty. Surrogate, to Jordan Doppelt, Rutgers Univ. (Aug. 26, 2015) (on file with authors).

  136. [136]. See, e.g., Fid. Union Tr. Co. v. Robert, 178 A.2d 185, 187 (N.J. 1962); Hirsch, supra note 32, at 1473 (noting the “unique[ness]” of the probable intent principle).

  137. [137]. See, e.g., Ex parte Adams, 168 So. 3d 40, 45 (Ala. 2014); Estate of McCreath v. McCreath, 240 P.3d 413, 423 (Colo. App. 2009); In re Estate of Heider, 496 S.W.3d 118, 121–22 (Tex. App. 2016).

  138. [138]. In re Estate of Tateo, 768 A.2d 243, 246 (N.J. Super. Ct. App. Div. 2001) (internal citations omitted) (quoting Fid. Union Tr. Co., 178 A.2d at 187); see also N.J. Stat. Ann. § 3B:3-33.1 (West 2007) (“The intention of a testator as expressed in his will controls the legal effect of his dispositions ... unless the probable intention of the testator, as indicated by the will and relevant circumstances, is contrary.”).

  139. [139]. See, e.g., In re Residuary Tr. Created Under the Will of Watson, No. M-2796-07, 2013 WL 1296284, at *7 (N.J. Super. Ct. App. Div. Apr. 2, 2013) (“The concept of presumed probable intent must be applied sparingly and only where necessary to give effect to the intent of the will or trust without varying the terms of the document.” (citing In re Estate of Munger, 309 A.2d 205 (N.J. 1973))); In re Estate of Flood, 9 A.3d 1086, 1088 (N.J. Super. Ct. App. Div. 2010) (reasoning that the probable intent doctrine “cannot be used to write a will that the testator did not write” (quoting In re Estate of Gabrellian, 859 A.2d 700, 705 (N.J. Super. Ct. App. Div. 2004))); In re Estate of Zahn, 702 A.2d 482, 487 (N.J. Super. Ct. App. Div. 1997) (declining to invoke the probable intent rule in the absence of an ambiguity).

  140. [140]. See, e.g., Estate of Russell v. Quinn, 444 P.2d 353, 358 (Cal. 1968) (en banc). In addition, there appears to be a burgeoning trend in other jurisdictions toward permitting extrinsic evidence to alter the terms of an unambiguous will. See, e.g., In re Estate of Duke, 352 P.3d 863, 878 (Cal. 2015); Erickson v. Erickson, 716 A.2d 92, 98 (Conn. 1998); In re Irrevocable Tr. Agreement of 1979, 331 P.3d 881, 888 (Nev. 2014); In re Estate of Herceg, 747 N.Y.S.2d 901, 904 (N.Y. Surr. Ct. 2002); Restatement (Third) of Prop.: Wills & Donative Transfers § 12.1 (Am. Law Inst. 2003) (“A donative document, though unambiguous, may be reformed to conform the text to the donor’s intention if it is established by clear and convincing evidence
    (1) that a mistake of fact or law ... affected specific terms of the document; and (2) what the donor’s intention was.”). Thus, even the more muscular version of New Jersey’s probable intent doctrine may soon be less of an outlier.

  141. [141]. See supra text accompanying notes 34–36.

  142. [142]. Hirsch, supra note 31, at 1039–40.

  143. [143]. The instruments in our sample were executed between May 20, 1972 and May 13, 2015.

  144. [144]. On the other hand, 109 wills (47%) had been created within the last five years.

  145. [145]. See supra notes 72–84 and accompanying text.

  146. [146]. See supra notes 75–83; see also N.J. Stat. Ann. § 3B:3­­-35 (West 2007).

  147. [147]. Restatement (Third) of Prop.: Wills & Donative Transfers § 5.5 cmt. h (Am. Law Inst. 1999).

  148. [148]. An additional six wills contained bare survival clauses accompanying devises of both tangible personal property and other devises, such as money or a residuary estate.

  149. [149]. But cf. In re Last Will & Testament & Tr. Agreement of Moor, 879 A.2d 648, 654–55 (Del. Ch. 2005) (“One can easily imagine persons who possess items of personal property-works of art, period piece furniture, sports memorabilia-that are more valuable than their cash and securities.”).

  150. [150]. See N.J. Stat. Ann. § 3B:3-35 (West 2007).

  151. [151]. Admittedly, this is one issue where the probable intent doctrine fuzzes things up. A party can invoke the rule to show that the testator intended antilapse to apply to a beneficiary who falls outside the scope of the statute. See, e.g., In re Estate of Mincer, 492 A.2d 1052, 1054 (N.J. Super. Ct. App. Div. 1985) (opining that trial court should have considered extrinsic evidence that testator intended antilapse to apply to bequest to brother-in-law). Conceivably, a testator could add a survivorship condition to a devise to an unrelated beneficiary to foreclose such a holding and drive home the point that antilapse should not apply.

  152. [152]. The results from Sussex County on survivorship issues are quite similar to those from another recent empirical study that one of us conducted in Alameda County, California. See David Horton, Wills Law on the Ground, 62 UCLA L. Rev. 1094, 1152–55 (2015) (finding that only five of 71 lapsed bequests contained survivorship conditions but that three of these bequests were to beneficiaries who did not fall within the ambit of the antilapse statute).

  153. [153]. See, e.g., Restatement (Third) of Prop.: Wills & Donative Transfers § 2.3 cmt. c (Am. Law Inst. 1999).

  154. [154]. This is also sometimes called “strict per stirpes” or “English per stirpes.” See, e.g., Restatement (Third) of Prop.: Wills & Donative Transfers § 2.3 cmt. d (Am. Law Inst. 1999). Modern per stirpes, a variation of English per stirpes, makes the initial division of shares at the nearest generation with a living relative. See, e.g., In re Estate of Evans, 827 N.W.2d 314, 321 (Neb. Ct. App. 2013) (“The difference between strict per stirpes and modern per stirpes is the generation at which shares of the estate are divided: Strict per stirpes begins at the generation closest to the decedent, regardless of whether there are any surviving individuals in that generation, whereas modern per stirpes begins at the first generation where there is living issue.”).

  155. [155]. See, e.g., In re Estate of Evans, 704 P.2d 35, 38 (Mont. 1985) (“The words ‘per stirpes’ mean by the root or stock ... . Persons who take per stirpes do so in a representative capacity and, standing in the place of a deceased ancestor, take only what he would have taken had he lived.” (citations omitted)).

  156. [156]. See N.J. Stat. Ann. § 3B:5-6 (West 2007); Unif. Prob. Code § 2-106(b) (amended 2010) cmt.

  157. [157]. See, e.g., Restatement (Third) of Prop.: Wills & Donative Transfers § 2.3 cmt. g (2017); see also generally 4 Jeffrey A. Schoenblum, Page on the Law of Wills § 36.6 (2017) (“A distribution per capita is an equal division of the property to be divided among the beneficiaries, each receiving the same share as each of the others, without reference to the intermediate course of descent from the ancestor.”); Lawrence W. Waggoner, A Proposed Alternative to the Uniform Probate Code’s System for Intestate Distribution Among Descendants, 66 Nw. U. L. Rev. 626 (1971) (proposing this innovation).

  158. [158]. See, e.g., N.J. Stat. Ann. §§ 3B:3-35, 3B:3-41 (West 2007). Property passing to descendants of a predeceased beneficiary under the antilapse statute is distributed “by representation”—meaning per capita at each generation—unless the will provides otherwise. Id.

  159. [159]. Per Capita, Black’s Law Dictionary (9th ed. 2009).

  160. [160]. See, e.g., Johnson v. Swann, 126 A.2d 603, 606 (Md. 1956) (“[T]his phrase has no meaning at all, with regard to named beneficiaries.”).

  161. [161]. Will of Eleanor Farber 1–2 (Jan. 2, 2001) (on file with authors). Moreover, the use of “per stirpes” (as the phrase is conventionally defined) would have been unnecessary, as the will neither contains a devise to grandchildren nor expressly provides for the contingency of a predeceased child survived by descendants.

  162. [162]. Last Will and Testament of Loretta Zuzich 1–2 (Nov. 20, 2007) (on file with authors).

  163. [163]. See, e.g., Richland Tr. Co. v. Becvar, 339 N.E.2d 830, 834 (Ohio 1975).

  164. [164]. In re Estate of Walters, 519 N.E.2d 1270, 1273 (Ind. Ct. App. 1988). In Walters, the testator’s wife, Jessie, had two children from a previous relationship, Bolin and Lucas. Id. at 1271. The testator left the residue of his estate “to [his] beloved wife, J[essie], ... per stirpes.” Id. After Jessie predeceased the testator, Bolin and Lucas argued that the bequest to her “per stirpes” was a clumsy way of providing that they—her heirs—should take her share if she died first. Id. An Indiana appellate court disagreed, noting that the phrase “per stirpes” applies “only to the mode of distribution of a bequest among a designated class” and thus does not “establish[] ... the class who shall take.” Id. at 1273; see also In re Estate of Winslow, 934 P.2d 1001, 1004 (Kan. Ct. App. 1997) (rejecting the argument that testator’s use of “per stirpes” manifested her intent that antilapse apply to bequest to the named individual); cf. Kirchner v. Buschling, 895 S.W.2d 180, 182 (Mo. Ct. App. 1995) (holding that bequest to named individuals “per capita” did not manifest intent to opt out of antilapse statute).

  165. [165]. Our sample confirms that at least some references to representation language should be treated as careless boilerplate and not as an expressed preference for antilapse. For instance, one testator devised the residuary estate to her daughter “absolutely and in fee simple per capita.” Will of Virginia Schwinn 3 (Dec. 21, 2010) (on file with authors). However, the testator then provided that, if the daughter died first, the property would pass to the testator’s son-in-law. Id. The testator could not have intended “per capita” as a shorthand for subjecting the bequest to the daughter to the antilapse statute. That would mean that the daughter’s descendants would receive the devise if the daughter passed away before the testator. But as the next passage of the will illustrated, the testator wanted someone else—her son-in-law—to step into her daughter’s shoes.

  166. [166]. Unif. Prob. Code § 2-603 cmt. (amended 2010) (Contrary Intention-the Rationale of Subsection (b)(3)).

  167. [167]. See supra Part II.B.

  168. [168]. See supra Part II.B.

  169. [169]. N.J. Stat. Ann. § 3B:25-1 (West 1982).

  170. [170]. See In re Estate of Zahn, 702 A.2d 482, 488 (N.J. Super. Ct. App. Div. 1997). The non-exoneration statute does not speak to this issue because it only governs “heir[s] or devisee[s]”—not individuals who receive property through joint tenancy. See id. at 484 (quoting N.J. Stat. Ann. § 3B:25-1).

  171. [171]. Id. at 487–88.

  172. [172]. N.J. Stat. Ann. § 3B:25-1 (emphasis added).

  173. [173]. N.J. Stat. Ann. § 3B:25-1 (West 2007).

  174. [174]. Id. (emphasis added).

  175. [175]. In re Estate of Payne, 895 A.2d 428 (N.J. 2006).

  176. [176]. Id. at 430–31.

  177. [177]. Id.

  178. [178]. Id.

  179. [179]. See id. at 430–32.

  180. [180]. Id. at 431 (alteration in original) (emphasis omitted).

  181. [181]. Id. at 435.

  182. [182]. See id. at 434–36.

  183. [183]. Id. at 437 (Rivera-Soto, J., dissenting) (“[I]t is the letter that is ambiguous while the will itself is patently clear: the mortgages on the Maine vacation home were to be satisfied from the estate, while no such provision was made in respect of the New Jersey home.”).

  184. [184]. See id.

  185. [185]. State v. Bryan, 185 P. 25, 26 (Kan. 1919).

  186. [186]. For a similar finding in a different context, see Stephen J. Choi & G. Mitu Gulati, Innovation in Boilerplate Contracts: An Empirical Examination of Sovereign Bonds, 53 Emory L.J. 929, 932–34 (2004). Choi and Gulati studied the impact of an “interpretive shock” on provisions that govern the modification of sovereign bonds. Id. at 932–33. Until 2000, sovereign bonds that selected New York law insisted on unanimous consensus among bondholders for any change to the principal and interest terms. See id. Then, however, Ecuador managed to modify its terms without unanimous agreement. See id. Choi and Gulati find that this surprising maneuver did not have an immediate impact, demonstrating “the gravitational pull of contract term standards in the market.” Id. at 934.

  187. [187]. See supra Part I.B.

  188. [188]. Bill Bischoff, Estate Taxes: The Worst Places to Die, Wall St. J. (Feb. 16, 2011, 11:17 AM),

  189. [189]. See, e.g., Ashlea Ebeling, New Jersey Repeals Estate Tax, Not Inheritance Tax, Forbes (Oct. 7, 2016, 5:08 PM), The other state is Maryland. See id. The New Jersey inheritance tax is complex, and varies based on the relationship between the decedent and the heir or beneficiary. See Jonathan L. Triantos, Note, Don’t Die in New Jersey: A Push for Effective Death Tax Reform Legislation, 13 Rutgers J.L. & Pub. Pol’y 411, 420–21 (2016). In the fall of 2016, New Jersey lawmakers voted to gradually phase out the estate tax. See, e.g., Robert F. Morris, New Jersey Estate Tax Repeal, Nat’l L. Rev. (Oct. 6, 2016),

  190. [190]. See Publications by Type, Inheritance and Estate Tax, State of N.J., http://www.state.nj.
    us/treasury/taxation/pdf/other_forms/inheritance/o10c.pdf (last visited Oct. 23, 2017).

  191. [191]. See, e.g., Ebeling, supra note 189.

  192. [192]. David J. Rosen, Legislative Budget & Fin. Officer, Remarks to the Assembly Budget Committee 4 (Mar. 30, 2015),

  193. [193]. See Alicia Adamczyk, How Many People Pay the Estate Tax?, Money: Everyday Money (Aug. 9, 2016),

  194. [194]. See Brauburger v. Sheridan, 72 A.2d 363, 366 (N.J. Super. Ct. Ch. Div. 1950).

  195. [195]. Last Will and Testament of Mary Elinor Eppler 1 (June 29, 2004) (on file with authors).

  196. [196]. See, e.g., Nichols v. Suiter, 78 A.3d 344, 346 (Md. 2013) (noting that divorce revokes such a bequest “in the absence of a contrary intention expressed by the testator”).

  197. [197]. See, e.g., In re Leete Estate, 803 N.W.2d 889, 902 (Mich. Ct. App. 2010) (noting that simultaneous death statutes do not apply when the will “declares ... a contrary intent”);
    cf. Stephens v. Beard, 485 S.W.3d 914, 917 (Tex. 2016) (explaining that the same rubric governs common-disaster provisions).

  198. [198]. See, e.g., Johnston v. Estate of Wheeler, 745 A.2d 345, 350 (D.C. 2000) (“[I]t is ‘presumed’ that if the testator made a specific bequest, he intended that bequest to fail if the designated asset is not part of the estate, unless the will in its entirety evinces a contrary intent.”); Harris v. Hines, 137 S.W.3d 898, 903–04 (Tex. App. 2004) (“A specific devise of realty is adeemed because the testator sold it before his or her death, absent a contrary intent expressed in the will ... .”).

  199. [199]. See, e.g., In re Horowitz, 961 N.Y.S.2d 854, 867 (N.Y. Surr. Ct. 2013) (“A testator who intends to fulfill his obligation [to pay a debt] by a bequest in his will can easily recite his intent to do so in his will ... .”).

  200. [200]. Ascher, supra note 79, at 642.

  201. [201]. See supra Part II.B.1.b.

  202. [202]. See, e.g., Cole v. Bailey, 146 A.2d 14, 15 (Md. 1958) (“[W]e find more than a faint glimpse of an intention that the distribution should not be per capita in the will before us.”); In re Edwin Meissner Testamentary Tr., 497 S.W.3d 860, 865 (Mo. Ct. App. 2016) (explaining that to override Missouri’s default presumption of per capita, a will must only “show[] [a] contrary intention”); cf. In re Estate of Goodwin, 739 N.Y.S.2d 239, 247 (N.Y. Surr. Ct. 2002) (“[D]ecedent’s choice of the phrase ‘share and share alike’ was evidence of his intent to rebut the then-existing statutory presumption of per stirpital distribution ... .”); Estate of Eggl v. Bjorge, 783 N.W.2d 36, 41 (N.D. 2010) (following plain language of will to distribute some of testator’s property per capita and some per stirpes). But see Wachovia Bank & Tr. Co. v. Livengood, 294 S.E.2d 319, 321 (N.C. 1982) (“[T]he term per stirpes (which the testator spelled per stripes) was not intended to be given its technical meaning.”).

  203. [203]. See, e.g., N.J. Stat. Ann. § 3B:15-1 (West 1983); see also Unif. Prob. Code § 3-601 (amended 2010).

  204. [204]. See generally Reid Kress Weisbord, Wills for Everyone: Helping Individuals Opt Out of Intestacy, 53 B.C. L. Rev. 877 (2012) (noting the importance of handpicking a trusted executor).

  205. [205]. In addition, wills have unique qualities which problematize the relationship between what they say and what testators understand them to say. “Just debts” clauses are a prime example. As the Georgia Supreme Court once observed, these provisions may be so popular because testators perceive them as signifying their “intent to leave the world with [their] accounts paid,” which enables them “to be remembered as an upright and respectable person.” Manders v. King, 667 S.E.2d 59, 61 (2008) (citation omitted). Words like “per stirpes” may also be ubiquitous because they sound like sophisticated legalese. See, e.g., Lawrence M. Friedman, A History of American Law 67 (2d ed. 1985) (noting that wills tend to employ “singsong, almost balladlike phrases”).

  206. [206]. A grand total of 156 wills (68% of the total sample of 230) contained tax apportionment clauses. However, as we discuss in the body text, 22 of these instruments did not clearly indicate whether the testator intended to override the default rule of equitable apportionment for assets passing under the will. Because we were unable to code these wills as either opting into or displacing the default rule, we have omitted them from this calculation.

  207. [207]. Admittedly, “residue pays” tax apportionment clauses can serve a purpose even if a will only contains a residuary clause. As noted above, such a provision can liberate the beneficiaries of nonprobate transfers from responsibility for taxes. However, there are two reasons to be skeptical that this explains the prevalence of these clauses in our data. First, as noted above, estate planners generally believe that testators are better off allowing each recipient of a nonprobate transfer to pay their fair share of taxes. See, e.g., Carolyn Burgess Featheringill, Estate Tax Apportionment and Nonprobate Assets: Picking the Right Pocket, 21 Cumb. L. Rev. 1, 22 (1990) (commenting on “[t]he inequity of takers of nonprobate assets receiving a tax-free windfall at the expense of the residuary takers”). Second, 16 of the 69 (23%) wills that included both (1) a “residue pays” clause and (2) distributed the entire estate through the residue did not mention nonprobate transfers. Accordingly, these tax apportionment clauses only applied to the probate estate, making their requirement that residuary takers pay all taxes superfluous.

  208. [208]. Evans, supra note 108, at 39.

  209. [209]. See generally, e.g., Richard H. Thaler & Cass R. Sunstein, Nudge: Improving Decisions About Health, Wealth, and Happiness (2008). Likewise, in 2011, President Obama issued an Executive Order directing all federal agencies to search for ways that existing regulation could be improved by appropriate choice-enhancing default rules. See Exec. Order No. 13,563, 76 Fed. Reg. 3821 (Jan. 18, 2011)

  210. [210]. See, e.g., Schwartz & Scott, Default Rule Project, supra note 6, at 1568 & n.142.

  211. [211]. See Ayres, Altering Rules, supra note 6, at 2087–88.

  212. [212]. See id.

  213. [213]. See Restatement (Third) of Prop.: Wills & Donative Transfers § 5.5 cmt. f (Am. Law Inst. 1999); see also supra Part II.B.1.a. (explaining the policy behind the “sticky” approach to antilapse).

  214. [214]. See supra Part II.B.1.a.

  215. [215]. See supra Part II.B.1.a.

  216. [216]. See supra Part II.B.1.a.

  217. [217]. In addition, there are other, less-prominent examples. If a testator dies without enough property to satisfy all of her devises, the probate court reduces general bequests of money before specific devises of particular things unless “the intent to change the order of abatement ... [is] ‘clearly indicated.’” Estate of Jenanyan, 646 P.2d 196, 200 (Cal. 1982) (en banc). Similarly, a gift to a “child” includes adopted children “unless it ‘plainly appears’ that the testator had a contrary intent.” Watson v. Baker, 829 N.E.2d 648, 653 (Mass. 2005).

  218. [218]. See, e.g., Featheringill, supra note 207, at 20–21.

  219. [219]. In re Estate of Vincent, 98 S.W.3d 146, 149 (Tenn. 2003); see also In re Estate of Dolley, 71 Cal. Rptr. 56, 62–63 (Ct. App. 1968) (refusing to find a “just debts” clause in a will sufficient to property passing by joint tenancy); In re Estate of Keil, 145 A.2d 563, 564 (Del. 1958) (reaching the same conclusion for assets passing by tenancy by the entirety); Manders v. King, 667 S.E.2d 59, 60 (Ga. 2008) (same for joint tenancy with right of survivorship).

  220. [220]. In re Estate of Carlson, 367 P.3d 486, 495 (Okla. 2016); see Estate of Young v. Phillip, No. A-96-423, 1997 WL 426191, at *3, *6 (Neb. Ct. App. July 1, 1997) (holding that provision in “will ordering that ‘all mortgages on any real property or interest therein titled in my name’ be paid” was sufficient to require residuary beneficiaries to pay off mortgage on real property that passed by joint tenancy).

  221. [221]. See, e.g., Featheringill, supra note 207, at 9–10.

  222. [222]. Ferrone v. Soffes, 558 So. 2d 146, 147 (Fla. Dist. Ct. App. 1990); see also Clarke v. United States, 94 F. Supp. 543, 547 (E.D. Pa. 1950) (“A direction in a will which applies only to the payment of the taxes on the legacies and devises therein has no significance or effect in regard to the payment of the taxes on extra-testamentary property.”); In re Estate of Shoemaker, 917 P.2d 897, 900 (Kan. Ct. App. 1996) (“The failure to specifically mention the nonprobate or nontestamentary property in the tax clause indicates that the testator did not intend for the estate to pay the inheritance taxes on the joint tenancy property.”); Chase Nat’l Bank of City of N.Y. v. Tomagno, 14 N.Y.S.2d 759, 761 (N.Y. Sup. Ct. 1939) (“A provision in a will that all taxes be paid out of the residuary or general estate applies only to property passing under the will unless it specifically refers to other property.”); In re Estate of Baltic, 946 N.E.2d 244, 246 (Ohio Ct. App. 2010) (“Any intent on the part of a testator or settlor that estate taxes are to be paid in a manner contrary to the apportionment method set forth in [the apportionment statute] ... ‘must be clear, specific, and unambiguous.’” (citation omitted)); Union Tr. Co. v. Watson, 68 A.2d 916, 919 (R.I. 1949) (“The law requires clarity in the statement of such a purpose.”). Some state legislatures have also established similar rules by statute. See, e.g., Nev. Rev. Stat. Ann. § 150.320 (LexisNexis 2014); Or. Rev. Stat. Ann. § 116.313 (West 2016). But see Estate of Sheppard ex rel. McMorrow v. Schleis, 782 N.W.2d 85, 96 (Wis. 2010) (holding that in the absence of a statutory or common law equitable apportionment rule, the residue presumptively pays taxes on nonprobate transfers).

  223. [223]. Jesse Dukeminier et. al., Wills, Trusts, and Estates 367 (Vicki Been et. al. eds., 9th ed. 2013).

  224. [224]. Vincent, 98 S.W.3d at 148.

  225. [225]. Carlson, 367 P.3d at 495.

  226. [226]. N.J. Stat. Ann. § 3B:24-5 (West 1983).

  227. [227]. For instance, in a 2008 case called McAuliffe v. Benish, Alan Burghardt executed a will that named his daughters as residuary beneficiaries and required his executors to pay “estate and inheritance taxes imposed by reason of my death, ... with respect to any property, whether disposed of by this [w]ill or otherwise.” McAuliffe v. Benish, No. C-109-06, 2008 WL 2020181, at *2 (N.J. Super. Ct. App. Div. May 13, 2008) (emphasis added). Burghardt then purchased a house in joint tenancy with right of survivorship with his girlfriend. Id. at *1. After Burghardt died, his girlfriend argued that the plain language of the apportionment provision covered nonprobate transfers and thus required the daughters to pay taxes on the home. Id. at *1. The appellate court held that it could not determine “whether [Burghardt’s] will evidences a clear intent to overcome the presumption of apportionment,” and remanded to trial court for more factfinding. Id. at *5. McAuliffe was not included in the official reports, which means that it is citable but not binding precedent. See N.J. Rule Ct. 1:36-3 (West 2017). However, reported cases also fail to resolve the clarity needed to task the residue of the will with paying death taxes on nonprobate transfers. Compare Bankers Tr. Co. v. Hess, 63 A.2d 712, 714 (N.J. Ch. Div. 1949) (finding a clause in a will that required taxes “with respect to any property required to be included in my gross estate under the provisions of any tax law and whether or not passing hereunder ... shall be paid out of my general estate ... and that there shall be no proration” governed nonprobate transfers), with Palmer v. Palmer, 39 A.2d 438, 440, 444 (N.J. Ch. 1944) (holding that a provision that mandated “taxes which may be imposed, chargeable or payable upon my estate or any legacy, bequest or devise herein, ... whether in trust or otherwise, shall be ... paid from my residuary estate” did not govern nonprobate transfers).

  228. [228]. McAuliffe, 2008 WL 2020181, at *5.

  229. [229]. See, e.g., Ayres, Altering Rules, supra note 6, at 2074–76.

  230. [230]. See, e.g., U.C.C. § 2-205 (Am. Law Inst. & Nat’l Conference of Comm’rs on Unif. State Laws 2014) (mandating that clauses that make offers irrevocable be “separately signed” by the offeror when they appear “on a form supplied by the offeree”); id. at § 2-209 (imposing a similar requirement for some provisions that require any modification of the contract to be memorialized in a signed writing); see also, e.g., Cal. Civ. Code § 1677(a) (West 2011) (requiring terms that entitle the seller to liquidated damages clauses for the buyer’s breach of a real estate contract to be “separately signed or initialed by each party”); Tenn. Code Ann. § 29-5-302(a) (West 2001) (stating that in arbitration clauses in “contracts relating to farm property, structures or goods, or to property and structures utilized as a residence of a party, the clause providing for arbitration shall be additionally signed or initialed by the parties”).

  231. [231]. Cf. Ayres, Altering Rules, supra note 6, at 2074–75 (endorsing altering rules that “requir[e] more extensive manifestations of assent” but also noting that some parties “may quickly initial at the indicated X’s without pausing to think whether the associated provision is objectionable”).

  232. [232]. A more outlandish option would be “train-and-test” defaults. Ayres, Altering Rules, supra note 6, at 2076–80. This scheme “require[s] nondrafting (and possibly even drafting) parties to pass a test before giving effect to a particular provision.” Id. at 2076. Policymakers have imposed train-and-test rules in contexts as diverse as the Health Insurance Portability and Accountability Act, mortgages, and student loans. See id. at 2076–80. In wills law, even if states balk at train-and-test principles, private companies might be able to institute similar rubrics. For instance, if there is demand for “smart” do-it-yourself estate planning, firms like Nolo Press and LegalZoom might be able to weave questions about the effect of particular provisions into their software. Similarly, high-end programs marketed to lawyers are now boasting of their ability to “warn you if you pick choices ... that are inconsistent or could lead to potential malpractice.” Estate Planning Drafting Software, Law. With Purpose, (last visited Oct. 23, 2017).

  233. [233]. Hirsch, supra note 32, at 1438. Hirsch first makes this claim in the context of intestacy. See Hirsch, supra note 31, at 1061. Specifically, he refutes the idea that intestacy statutes should distribute property randomly in order to encourage people to execute wills. See id. at 1058–61. In a later article, Hirsch extends this analysis to sticky defaults and negative wills. See Hirsch, supra note 32, at 1438.

  234. [234]. Hirsch, supra note 31, at 1061.

  235. [235]. See Willis, supra note 6, at 1157 (discussing the concept of “nudges”).

  236. [236]. See supra text accompanying notes 79–84.

  237. [237]. See supra text accompanying note 82.

  238. [238]. Indeed, as noted above, the vast majority of testators in our sample go out of their way to name substitute takers. See supra text accompanying notes 148, 213–14.

  239. [239]. See supra notes 114­–22 and accompanying text.

  240. [240]. See supra notes 217–22.

  241. [241]. See Hirsch, supra note 31, at 1039–40.

  242. [242]. See, e.g., Ashbel G. Gulliver & Catherine J. Tilson, Classification of Gratuitous Transfers, 51 Yale L.J. 1, 6–13 (1941) (discussing the functions of these formalistic demands).

  243. [243]. These are the requirements in the Wills Act, which the British Parliament passed in 1837, and which migrated to the United States. See Wills Act 1837, 7 Will. 4 & 1 Vict. c. 26; Jesse Dukeminier et al., Wills, Trusts, and Estates 226–27 (Vicki Been et al. eds., 8th ed. 2009).

  244. [244]. Cf. Kreiczer-Levy, supra note 46, at 951 (noting the importance of “careful deliberation [in the realm] of legal transfers, especially gratuitous ones”).

  245. [245]. Restatement (Third) of Prop.: Wills & Donative Transfers § 10.1 cmt. a (Am. Law Inst. 2003).


Vice Dean, Professor of Law, and Judge Norma L. Shapiro Scholar, Rutgers Law School.


Professor of Law and Chancellor’s Fellow, University of California, Davis, School of Law (King Hall).

Thanks to Mark Ascher, Naomi Cahn, Joel Dobris, David Hoffman, John Langbein, David Noll, and Stephen Urice for valuable comments. We are also especially grateful to Surrogate Gary Chiusano of the Sussex County, New Jersey Surrogate’s Court for facilitating access to the public records.