Partial Harmless Error for Wills: Evidence From California

I.     Introduction

Gerard Caspary was born in 1929 in a prosperous Jewish neighborhood in Frankfurt, Germany.1 When Caspary was four years old, Hitler came to power, and the Caspary family fled to Paris.2 In 1940, “the Nazis invaded the city.”3 Caspary’s parents were sent to Auschwitz, where they were killed.4 Caspary eluded capture and for three years lived underground in South France.5

When the war ended, Caspary immigrated to America.6 He attended Swarthmore and Harvard, won a Guggenheim Fellowship, and became a beloved medieval history professor at the University of California, Berkeley.7 Yet as he aged, he increasingly felt compelled to examine his own dark past.8 Shortly before he retired, he began to teach an undergraduate seminar on the Holocaust.9 He also wrote a book that combined his childhood memories with translations of his family’s wartime letters.10

In the spring of 2005, Caspary scheduled a meeting with an estate planner.11 On May 25, the day before the appointment, Caspary typewrote a one-page document entitled “Last Will.”12 In it, Caspary named two executors and expressed his desire to leave $10,000 to his godson, $5,000 to his housekeeper, and the “[b]ulk of [his] estate” to the Holocaust Museum in Washington, D.C.13 However, other portions of the writing seemed tentative, such as the directive that the “[e]xecutors [are] to receive 5% (?) of [the] estate from the top.”14 In shaky handwriting at the foot of the page, Caspary added his name, address, phone number, and email.15

Figure 1: Caspary’s Purported Will

Figure 11 

On May 26, Caspary met with the estate planner.16 Caspary mentioned his experiences in France during the war.17 He also said that because he was unmarried, childless, and had no close family, he wanted to leave most of his estate to the Holocaust Museum in honor of his parents.18 After the appointment, the attorney prepared a draft estate plan for Caspary and contacted Caspary several times to set a date to execute it.19 Despite these attempts, Caspary never responded.20 On April 6, 2008, Caspary died.21

Under centuries of settled law, Caspary had failed to make a valid will. Throughout the western world, the rules that govern the creation of a testamentary instrument are profoundly unforgiving.22 In former British colonies, such as Australia, Canada, and the United States, an ancient statute called the Wills Act requires posthumous dispositions of property to be in writing, signed by the testator, and attested by two witnesses who, present at the same time, saw the testator either sign or acknowledge the instrument.23 Traditionally, courts insisted on strict compliance with these elements. They cited minor departures—a misunderstanding, a misplaced signature, or a witness who glanced away at the pivotal moment—to nullify writings that decedents “doubtless[ly] intended to be [their] last will.”24 Measured by these bright lines, Caspary’s May 25 document was not even a close call. After all, the instrument had no witnesses. Accordingly, on July 7, 2008, the Public Administrator for Alameda County, California—who handles the affairs of decedents who have no known relatives—sought to probate Caspary’s estate as an intestacy.25

But, five months later, while Caspary’s case was still pending, the legal landscape shifted. The California legislature added a doctrine known as “harmless error” to its probate code.26 Harmless error has been called “the most significant change in what constitutes a will since [the] enactment of the Statute of Frauds.”27 It validates a document that does not comply with the Wills Act if there is clear and convincing evidence that a decedent wanted it to be her will.28 When this innovation became effective on January 1, 2009,29 it breathed life into Caspary’s May 25 instrument. A distant relative filed a petition asking the court to admit the May 25 writing to probate.30 Three of Caspary’s friends filed declarations asserting that the typewritten page set forth Caspary’s wishes.31 The case proceeded to trial and then settled, giving the Holocaust Museum a share of the estate.32

The harmless error rule has long been one of the few hot-button issues in the staid field of wills. It emerged more than 40 years ago in South Australia.33 About a decade later, John Langbein championed it in a famous article.34 The doctrine soon spread throughout Australia and to parts of Canada.35 In the United States it earned the endorsement of two influential secondary sources: the 1990 amendments to the Uniform Probate Code (“UPC”) and the Restatement (Third) of Property: Wills and Other Donative Transfers.36 Yet legislatures have been less enthusiastic. Only nine states other than California have embraced the doctrine.37 Moreover, some of these jurisdictions (including California) have adopted what I call “partial” harmless error—they forgive some deviations from the Wills Act (for example, those involving witnesses) but not others (such as those relating to the testator’s signature).38

Harmless error has struggled because lawmakers have been spooked by the quintessential American boogeyman: the specter of litigation. Strictly interpreting the Wills Act generates predictable results. A piece of paper that is not signed by two witnesses or by the decedent is just that—a piece of paper. Because judges do not bend the rules of will execution, it is pointless to try to probate a defective writing. But harmless error destroys this buffer. Without the bright lines of conventional law, “the concern arises that proponents of noncompliant wills will flood the courts.”39

Compounding these problems, we know very little about harmless error’s real-world impact. Even today, Langbein’s article provides the clearest window into this issue. Langbein predicted that harmless error would decrease litigation by eliminating incentives for disappointed heirs to predicate will contests on rank technicalities.40 At first blush, though, one of his findings seemed to point in the opposite direction. Langbein examined eight years of trial court decisions in South Australia and unearthed 41 harmless error cases.41 Even more starkly, a follow-up piece by Stephanie Lester two decades later discovered that the volume of harmless error cases in the tiny state had “ballooned” to several dozen each year.42 With these as our only reference points, it is easy to see why commentators have worried that relaxing the Wills Act formalities might cause “an explosion of previously foreclosed litigation”43 and why American lawmakers have been hesitant to upend centuries of settled law.  

This Article offers a new perspective on these issues by presenting the results of an empirical study of how the harmless error rule has affected an American probate court. Its centerpiece is a dataset of 2,453 estates that came on calendar in Alameda County, California, between January 1, 2008, and December 31, 2010. Because the Golden State’s partial harmless error rule became effective on January 1, 2009, my research provides a glimpse of how it played out in practice.

This Article draws two conclusions from this data. First, California’s embrace of partial harmless error had little impact on the volume of probate litigation in Alameda County. Parties invoked the new statute just five times in the 1,543 cases that were eligible for it. Second, by retaining certain Wills Act elements, partial harmless error limits mind-bending dilemmas about a decedent’s intent. For example, because California’s diluted version of the rule requires the decedent’s signature, it weeds out a broad range of tentative, informal writings. For these reasons, partial harmless error is a good option for legislatures that want to sweep away the cobwebs of formalism but have qualms about uninhibited functionalism.

Part II of this Article lays the foundation for my study by tracing the evolution of the rules that govern the execution of wills, and Part III details my research methodology and discusses my results.

II.     The Law of Will Execution

This Part provides background about the harmless error rule. It begins by describing conventional will-execution doctrine, which was reviled for defeating a decedent’s intent. It then describes how South Australia challenged this legacy of formalism by passing the first harmless error statute. Finally, it details harmless error’s struggle for acceptance in the United States.

A.     Traditional Law

Will execution takes place in the shadow of two ancient English laws. In 1677, the English Parliament passed the Statute of Frauds, which requires wills that transmit real property be in writing, signed by the testator, and subscribed in the testator’s presence by at least three witnesses.44 Then, in 1837, the Wills Act created a single set of formalities that governed all testamentary dispositions.45 The Wills Act mandates that wills be written, signed by the testator, and subscribed by two witnesses who were “present at the same [t]ime” when they saw the testator either sign the will or ratify her previous signature.46 These elements became law in places as far-flung as America and Australia.47

Courts implemented the Wills Act without mercy. Although the purpose of the statute was to safeguard the interests of the decedent, judges elevated the law’s letter over its spirit by nullifying writings that were clearly intended to be wills.48 Even when the errors were benign and banal—for example, when there were two witnesses, but they were not “present at the same time”—they were fatal.49 For example, in In Re Lynch’s Estate, the decedent signed an attempted will at his desk in front of a friend, who added his own signature and left the office.50 The decedent then asked his secretary, who had been working a few feet away, to sign the document.51 She did, and then asked what she had signed.52 Although both witnesses had been in the same room, the California appellate court denied probate, reasoning that “[a] will is not signed in the presence of one who is attending to another matter and does not know what is taking place until [s]he is told later.”53 This tradition of strict compliance made the law of will execution “notorious for its harsh and relentless formalism.”54

Some American states carved out an exception to the Wills Act by recognizing holographic wills.55 The hallmark of a holograph is that its “material provisions” must be in the testator’s handwriting.56 In this way, holographs serve as a small pocket of informality in the law of will-making. Indeed, they can be inscribed in letters,57 recipes,58 or even the fender of a tractor.59 At the same time, though, courts were unblinking in their insistence that the key terms of a holograph be inscribed by the testator’s hand.60 In In re Thorn’s Estate, the decedent owned a rubber stamp that bore the title of his second home, Cragthorn.61 He wrote his will manually, but used the stamp to identify “Cragthorn.”62 The California Supreme Court denied probate, reasoning that the “intent of the deceased [was] obvious” but “beside the question.”63

In the middle of the 20th century, cases like Lynch and Thorn became more controversial as progressive academics began to reconsider the role of formalism in private law. Seminal articles by Ashbel Gulliver and Catherine Tilson (writing together) and Lon Fuller (writing alone) christened this debate by identifying and critiquing the underlying purposes of the Wills Act.64 First, the authors argued that the statute generates proof of a decedent’s intent (the “evidentiary function”).65 As Gulliver and Tilson elaborated, the writing and signature elements are vital because “the testator will inevitably be dead and therefore unable to testify” during any subsequent trial.66 Second, Gulliver, Tilson, and Fuller contended that the pomp and circumstance of obtaining two witnesses reinforces the gravity of estate planning.67 Gulliver and Tilson called this the “ritual function,” while Fuller dubbed it the “cautionary function.”68 Third, Fuller claimed that formalities serve a “channeling function.”69 As he observed, unbending doctrines “furnish[] a simple and external test of enforceability” by standardizing the appearance of particular legal instruments.70 As later authors would put it, the telltale signs of a signed and witnessed writing means that “[c]ourts are seldom left to puzzle whether the document was meant to be a will.”71 Fourth, Gulliver and Tilson noted that the safeguard of witnesses might deter third parties from trying to trick or strong-arm the testator (the “protective function”).72 However, they were skeptical of this function, noting that it made sense when deathbed wills were common, but had little relevance now that “wills are probably executed . . . in the prime of life and in the presence of attorneys.”73

Gulliver, Tilson, and Fuller’s insight that testamentary formalities are mere steps toward a goal—not a goal themselves—made a deep impression. In the last quarter of the 20th century, a new generation of scholars took aim at the practice of demanding strict compliance with the Wills Act. Some echoed Gulliver and Tilson’s calls to abandon the attestation requirement, arguing that it was nothing more than a hangover from previous centuries.74 Others suggested that courts stop demanding strict compliance with the statute and instead focus on whether the circumstances surrounding the execution of a will had nevertheless satisfied the evidentiary, ritual, channeling, and protective objectives.75

These arguments were bolstered by a dramatic change in the nature of estate planning. Inspired by Norman F. Dacey’s unlikely best-seller, How to Avoid Probate!, property owners began going out of their way to bypass the court-supervised administration process.76

In particular, they began to structure their posthumous dispositions around revocable inter vivos trusts. The use of trusts as “will substitutes” creates a paradox because trusts do not need to adhere to the Wills Act.77 The popularity of these unwitnessed, contract-like mechanisms raised fresh questions about the steps necessary to transmit property after death.

Thus, near the end of the 20th century, there was a budding consensus that orthodox will-creation doctrine needed to change.78 At the same time, though, there was no consensus about how to change it. The most promising solution to this conundrum would stem from an unlikely place.

B.     Harmless Error  

1.     South Australia

In 1974, the Law Reform Committee of South Australia released a 12 page report about the country’s inheritance laws.79 The document discussed the nitty-gritty of intestate succession, including advancements and the share of a surviving spouse.80 As though it were an afterthought, the Committee also noted that there would be fewer intestacies if

in all cases where there is a technical failure to comply with the Wills Act . . . the Court or a Judge [could] declare that the will in question is a good and valid testamentary document if he is satisfied that the document does in fact represent the last will and testament of the testator.81

A year later, Parliament enshrined this recommendation in section 12(2) of the probate code:

A document purporting to embody the testamentary intentions of a deceased person shall, notwithstanding that it has not been executed with the formalities required by this Act, be deemed to be a will of the deceased person if the Supreme Court, upon application for admission of the document to probate as the last will of the deceased, is satisfied that there can be no reasonable doubt that the deceased intended the document to constitute his will.82

Initially called the “dispensing power,” this novel rule gradually came to be known as “harmless error.”83

In 1987, John Langbein threw his weight behind the South Australian experiment.84 He examined every Australian Supreme Court opinion applying section 12(2) between 1975, when the statute was enacted, and June 1986.85 He found no harmless error cases within the first three years of the law’s existence.86 Afterwards, he found 41.87

Parsing these decisions, Langbein demonstrated that some deviations from the Wills Act are more innocuous than others. For instance, the Supreme Court bent over backwards to forgive attestation errors.88 Indeed, judges routinely overlooked the fact that two witnesses were not together when the testator authenticated the will.89 This leniency extended to what Langbein called “number” defects: instruments that had been attested by just one person, or nobody at all.90 Conversely, no litigant had tried to probate an oral will, and the Supreme Court was more hesitant to enforce documents that the testator had not signed.91 As Langbein explained, this indicated that attestation was lower on the formality totem pole: “Because writing and signature have greater evidentiary and cautionary value, they are much harder to dispense with under section 12(2).”92 

Langbein also discovered that section 12(2) had come into play in a large number of “alteration” cases.93 In these matters, the testator had made a valid attested will but then later tried to amend it, often by crossing out existing text and handwriting new terms.94 These changes rarely complied with the Wills Act because they were often unsigned and unwitnessed.95 Nevertheless, courts had liberally applied section 12(2) to honor the testator’s wishes. Langbein applauded this development, noting that these decedents had made the justifiable mistake of assuming that their previous proper execution of the will “had continuing effect.”96

However, Langbein’s view of section 12(2) was not entirely positive. First, he observed that the statute raised tough questions about testamentary intent.97 For instance, in some cases, decedents realized that they were violating the Wills Act. One decedent’s estate planner told him that his handwritten changes to an attested will were insufficient, but the decedent never cured this problem.98 Another decedent, who happened to be a lawyer, thought he was having a heart attack and jotted down his dispositive wishes in a notebook.99 He then recovered, but did not have the writing attested.100 As Langbein explained, courts were struggling with these fact patterns for good reason: The testator’s failure to remedy an instrument that he knew was invalid could be seen as ambivalence towards its instructions.101  

Second, Langbein critiqued section 12(2)’s evidentiary standard.102 On paper, the statute required judges to be “satisfied that there [is] no reasonable doubt that the deceased intended the document to constitute his will.”103 Langbein argued that courts had not taken this stringent test—“the highest standard of proof known to the law”—at face value.104 He suggested changing this rule to conform with what judges actually seemed to be doing: asking whether there is clear and convincing evidence of a decedent’s wishes.105

Despite his concerns, Langbein urged other legal systems to import harmless error. He argued that the rule had a tremendous upside: the ability to honor the wishes of decedents who made small mistakes during the execution process.106 On the opposite side of the ledger, he claimed that the doctrine’s costs were bearable. He conceded that his discovery of more than 40 lawsuits over a short period “in a smallish jurisdiction” raised the specter of a “litigation imbroglio.”107 Nevertheless, he claimed that this apparent spike in filings was “illusory” because South Australia does not allow parties to settle will contests.108 As a result, only three of the decisions under his microscope were genuinely adversarial; in the rest, the court was fulfilling its duty to assess the validity of a purported will.109 Accordingly, he concluded that the time had come to abolish “some of the harshest and most senseless rules that remain in Anglo-American private law.”110

Harmless error eventually migrated to South Australia’s neighbors, including the Australian Capital Territory,111 New South Wales,112 the Northern Territory,113 Tasmania,114 and Western Australia.115 Likewise, it crossed the ocean to Israel,116 Manitoba,117 and Saskatchewan.118

Meanwhile, in 1994, the South Australian Parliament inadvertently made the rules of will execution even more informal. In an attempt to relax section 12(2)’s criminal law-esque “beyond a reasonable doubt” standard of proof,119 lawmakers drafted a clumsy amendment that swung so far to the other side that it permitted courts to enforce documents that merely “express[] testamentary intentions of a deceased person.”120 This opened the door for parties to try to probate a range of inchoate writings. For example, in Estate of TLB, the decedent filled out a “Will Instruction Sheet,” and her solicitor prepared a tentative estate plan based on this document.121 Before the decedent had even seen the lawyer’s handiwork, she committed suicide.122 The court nevertheless admitted the “Will Instruction Sheet” to probate, reasoning that the statute contained “no requirement that the deceased must have intended the document to be his or her will.”123 Accordingly, in 2000, lawmakers tried to close this loophole by requiring an express command that the decedent wanted a particular writing to be legally binding.124

In 2007, Stephanie Lester reexamined the South Australian case law and found that the number of section 12(2) cases had skyrocketed.125 For instance, she unearthed 32 harmless error decisions from 2005 and 2006 alone.126 She also discovered that while Parliament’s 2000 amendments had attempted to clarify that the statute requires a decedent to regard a writing as a will, judges sometimes played fast and loose with this mandate.127 For instance, courts enforced a writing entitled “Proposed Will”128—a draft that a decedent had sent back to her solicitor so he could correct the spelling of a beneficiary’s name129—and a writing with a gaping empty space where the witnesses and the testator were supposed to sign.130

South Australia was not the only country struggling with the harmless error rule. As this Article discusses next, it had also piqued the interest of some American lawmakers.

2.     America

Shortly after Langbein’s article, the harmless error rule began to take root in the United States. Two major law reform projects—the 1990 revisions to the UPC, and the Restatement (Third) of Property: Wills and Donative Transfers—included the doctrine.131 The UPC model harmless error statute, section 2-503, draws heavily from Langbein’s article, and is designed to remedy the two kinds of defects he found most egregious: those related to attestation and handwritten changes to the face of a previously-executed will.132 Thus, section 2-503 applies broadly, retaining (or at least seeming to retain) “writing” as the single obligatory formality:

Although a document or writing added upon a document was not executed in compliance with [the Wills Act], the document or writing is treated as if it had been executed in compliance with that section if the proponent of the document or writing establishes by clear and convincing evidence that the decedent intended the document or writing to constitute . . . the decedent’s will . . . [or] an addition to or an alteration of the will . . . .133

In the early and mid-1990s, Montana, Colorado, Hawaii, Michigan, South Dakota, and Utah adopted section 2-503 verbatim.134

But since then, the pendulum has swung in the opposite direction. Only four states have passed harmless error statutes in the new millennium.135 Moreover, three of these jurisdictions—California, Virginia, and Ohio—have adopted narrow versions of the rule, which I call “partial” harmless error.136 For example, Virginia’s doctrine cannot “excuse compliance with any requirement for a testator’s signature.”137 Ohio goes further by still mandating that the testator sign the will in the presence of two witnesses.138 Colorado, which had initially adopted the UPC’s capacious harmless error standard in 1995, amended the rule in 2001 to “apply only if the document is signed or acknowledged by the decedent as his or her will or . . . the decedent erroneously signed a document intended to be the will of the decedent’s spouse.”139 In fact, one judge in a UPC jurisdiction has emphatically argued that harmless error should not apply to unsigned documents.140

The root of this reluctance is intuitive: Lawmakers are afraid that testamentary informality encourages will contests. Strictly interpreting the Wills Act may sometimes thwart decedents’ intent, but it also serves the salutary purpose of making it abundantly clear whether a document qualifies for probate. Replacing this crystalline regime with a muddy standard could force courts to grapple with an endless array of non-compliant writings. Indeed, law review articles discussing the harmless error rule have observed that it has the potential to mire “the probate system [in] burdensome litigation.”141

Scholars have also criticized the mechanics of the harmless error rule. For instance, Emily Sherwin has challenged the idea that the doctrine’s heightened standard of proof will deter litigation.142 Drawing on economic theories of litigant behavior, Sherwin argues that requiring clear and convincing evidence might actually reduce settlement rates because it makes it harder for the parties to predict their odds of winning on the merits.143

Likewise, Jane Baron contends that harmless error creates uncertainty because it hinges on one of the most malleable concepts in wills law: testamentary intent.144 Baron notes that the doctrine is supposed to remedy “technical, innocuous errors” made by “coldly rational” individuals who see wills as “a means for furthering self-determined ends.”145 But in reality, she argues, property owners do not always conform to that archetype.146 Instead, some of them struggle, equivocate, and change their minds.147 As a result, they leave behind writings (or a series of writings) that reflect the brute fact that they themselves do not know what they want.148 In turn, Baron observes, this prompts well-meaning courts to stretch the boundaries of the harmless error rule.149 Rather than asking whether a decedent intended a document to be her will, they ask whether “the testator had a particular dispositive wish.”150 Because this broad reading of the rule opens the courthouse door to the endless parade of informal documents created by ambivalent decedents, Baron concludes that “the clear and convincing evidence standard has not, and will not, function as a serious limit . . . .”151

Until now, this debate has been hamstrung by a lack of data. There is no American equivalent to Langbein’s shirt-sleeves-rolled-up review of the South Australian experience. Moreover, in the United States, just 9 published152 and 11 unreported opinions153 apply the doctrine. For these reasons, we simply have no idea how the rule impacts the day-to-day operation of an American probate court. Thus, in the next Part, I report the results of a study that begins to fill this void.

III.     Partial Harmless Error: Evidence From California

This Part updates our understanding of the harmless error rule by examining how California’s partial embrace of the doctrine impacted a major county’s probate court. It first details my research methodology. It then presents and analyzes my results.

A.     Methodology and Caveats

I previously assembled a dataset of every testate and intestate administration stemming from deaths that occurred in 2007 in Alameda County, California.154 Alameda County is a diverse urban area that lies across the bay from San Francisco and includes the cities of Albany, Berkeley, Emeryville, Fremont, Hayward, Oakland, and San Leandro. Unfortunately, most of the cases in this sample had concluded by January 1, 2009, when California’s harmless error statute became effective. Thus, for this Article, I expanded the scope of my research.

To do so, I used Alameda County’s judicial website, DomainWeb.155 I used the calendar search function to examine each case that came on the probate court calendar between March 1, 2009 (the date my previous research period ended) and December 31, 2010. To maintain continuity with my original dataset, I omitted any new estate that involved a decedent who passed away before 2007. I collected about ten variables from each relevant case, including: whether the decedent made an attested or holographic will; the total value of the estate; and whether litigation ensued.156 

Together, my original and supplemental research yielded 2,453 estates.157 I will refer to this universe of cases as my “combined dataset.” In addition, I also created a subsample of 1,543 matters that I will call my “harmless error eligible dataset.” Although California’s statute did not kick in until 2009, I treated an estate as “harmless error eligible” if it opened after July 1, 2008. I drew the line there because (1) there is usually at least a two-month lag between the petition for probate and the order appointing the personal representative; and (2) parties have four months after this ruling to challenge it.158 Therefore, as long as a matter began in July 2008, the proponent of a non-compliant instrument would have been able to invoke the harmless error rule.159

California’s probate system has a few idiosyncrasies. For one, unlike many jurisdictions, the state does not have separate tracks for informal and formal matters. Instead, all cases begin the same way. A Probate Examiner evaluates every petition for administration by filling out a “checksheet.”160 The Examiner flags anything unusual about a purported will for the judge. Under this regime, the proponent of an instrument bears the burden of proving that the instrument is valid even if no adverse party appears.161

Figure 2: The Probate Examiner’s Checksheet

Figure 12

California also has adopted a nonprobate mechanism that skews the demographics in my sample. When a married individual passes away, the surviving spouse can collect their share of the estate outside of the normal administrative process by filing a “spousal property petition.”162 Because many spouses leave their assets to their partners, the first spouse to die rarely appears in the probate files. As a result, my dataset is saturated with unmarried decedents.163 In turn, this may increase the ratio of contested cases. The transfer triggered by the first spouse’s death (in which most assets flow to the surviving spouse) is less likely to be contentious than the transfer triggered by the second spouse’s death (where family and friends may expect a cut). By oversampling the second type of transfer, my dataset may exaggerate the incidence of litigation.

Finally, California’s partial harmless error rule is very modest. The legislature grafted the doctrine onto Probate Code section 6110(c), which governs the creation of attested wills:

(1) Except as provided in paragraph (2), the will shall be witnessed by being signed, during the testator’s lifetime, by at least two persons each of whom (A) being present at the same time, witnessed either the signing of the will or the testator’s acknowledgment of the signature or of the will and (B) understand that the instrument they sign is the testator’s will.

(2) If a will was not executed in compliance with paragraph (1), the will shall be treated as if it was executed in compliance with that paragraph if the proponent of the will establishes by clear and convincing evidence that, at the time the testator signed the will, the testator intended the will to constitute the testator’s will.164

This language reveals that harmless error only applies to instruments that do not “compl[y] with paragraph [6110](1)”—in other words, those with attestation defects.165 Thus, California’s version of the doctrine might salvage a would-be will that has one witness or that was signed by two people who were not “present at the same time.”166 But like its counterparts in Colorado, Virginia, and Ohio, it cannot cure problems related to the testator’s signature.167 

Unfortunately, it is not clear whether California’s statute can ratify a purported holograph that fails because the “material provisions” are not in the testator’s handwriting.168 The probate code does not mention holographs in section 6110; instead, it authorizes them in section 6111.169 Arguably, by limiting harmless error to documents that are “not executed in compliance with [section 6110(c)(1)],”170 the legislature exempted writings that violate the discrete requirements of section 6111. This would be consistent with the general idea that California’s harmless error rule merely covers attestation errors (and thus would not encompass the problem of too much typewriting).

But some aspects of the statute cut the other way. Suppose a testator tries to make a holograph, but includes too much typewritten text. This botched “holograph” can nevertheless become a valid attested will if the testator signs or acknowledges it in front of two witnesses.171 Another way of understanding California’s harmless error rule is that it treats clear and convincing evidence that a decedent wanted a writing to be her will as a substitute for proper attestation.172 In turn, because attestation is a kind of holy water that can validate a part-handwritten, part-typewritten “holograph,” harmless error might apply.173

It also seems as though California’s harmless error rule cannot cure a failed alteration to an existing will. Recall that one of the driving forces behind the UPC was to allow testators to handwrite changes on the face of a valid attested will.174 As a result, UPC section 2-503 specifically applies to “an addition to or an alteration of the will.”175 However, the California legislature declined to extend probate code section 6110(c)(2) this far. Although the first draft of the statute was identical to the UPC,176 lawmakers then eliminated the reference to harmless error curing “addition[s]” and “alteration[s].”177 As a Senate Judiciary Committee report explains, this change means that “[u]nlike the UPC . . . the proposed harmless error rule (new 6110(c)(2)) would only apply to the testator’s will and not to any other writing such as . . . an alteration of a will.”178 Because of these carve-outs, California’s partial harmless error rule is actually quite limited.

B.     Results

My harmless error-eligible dataset contains 1,543 cases: 873 testacies and 670 intestacies. Harmless error surfaced five times in these matters. In addition to the dispute over Gerard Caspary’s purported will,179 the rule arose in two uncontested and two contested estates.

Both of the uncontested harmless error cases involved unwitnessed documents. In the first, Estate of Campbell, Billy Gene Campbell typed, dated, and signed a document dividing his property among: his longtime partner, Lana; his son, Daniel; and his grandchildren, Ryan, Mark, and Carlee.180 The writing contained several inside jokes, such as disposing of a specific belt buckle and old-school video game consoles.181 It concluded with a moving passage addressed to Daniel: “Take care of those children. You are truly blessed. Be patient with Ryan, be honest and sincere with Mark and Carlee . . . . Hug and kiss those darlings for me every night.”182

Figure 3: Billy Campbell’s Will

Figure 13

Lana offered the document for probate, and the Examiner requested briefing on “how the [c]ourt can admit a document as a will when it is not witnessed and [its] provisions . . . are not in the handwriting of the testator.”183 Lana responded by invoking section 6110(c)(2) and citing affidavits from Billy’s sister and friends that he had mentioned executing a will that split his estate between Lana and Daniel.184 No adverse party came forward, and the court granted Lana’s petition.185

The second uncontested harmless error case, Estate of Burns, involved markedly different circumstances. In August 2009, Jack Burns hung himself.186 The police discovered three “documents neatly placed on the counter at [his] house.”187 One was typewritten, signed, unattested, and labeled “Will of Jack Ray Burns.”188 It included Burns’s social security number, named his friends Jeff and Lesle Nelson as executors, and gave all of his property to Jeff.189 Another was an informational pamphlet designed to help a lay person serve as personal representative.190 Finally, Burns left a wrenching suicide note entitled “My Last Words,” in which he stated “I have written a Will using Quicken WillMaker Plus 2004. I did not have it witnessed, but I am assuming that something is better than nothing.”191 Both the will and the note were dated January 31, 2009—seven months before Burns took his own life.192

Jeff and Lesle attached these papers to their petition to probate the “[w]ill.”193 They also included a short paragraph that described Burns’s suicide and cited the harmless error rule:

While the [w]ill of Jack Ray Burns was not witnessed, the circumstances and other documents provide clear and convincing evidence that at the time the testator signed the [w]ill, the testator intended the [w]ill to constitute the testator’s [w]ill. Therefore, the [w]ill is valid pursuant to California Probate Code §6110(c)(2).194

Based on this evidence and the lack of objection by Jack’s intestate heirs, the judge admitted the instrument to probate.195

As one would expect, the contested estates were more complicated. The first such case, Estate of Ben-Ali, was a veritable soap opera that ended up as a published appellate opinion.196 Hassan Ben-Ali owned several apartment buildings, including one on Ashby Avenue in Berkeley.197 In the early 1990s, he experienced financial difficulties.198 In 1993, to prevent the IRS from seizing the Ashby property, Hassan transferred title to his son, Taruk.199 “On August 3, 2002, Taruk married . . . Wendelyn Wilburn.”200 Hassan saw Wendelyn as a gold-digger and disapproved of the union.201

In 2004, Taruk died of a drug overdose.202 Hassan found Taruk’s body.203 Rather than reporting the incident, Hassan hid the corpse in a storage area in the Ashby complex.204 Hassan began to forge Taruk’s signature on documents related to the units.205 In 2008, Hassan committed suicide, and the police finally found Taruk’s body.206

Among Hassan’s papers was a typewritten document that appeared to be Taruk’s will.207 It was dated August 16, 2002—just 13 days after Taruk and Wendlyn had married.208 It gave Taruk’s personal property to Wendelyn and the rest of his assets to Hassan.209 The will bore the apparent signatures of Taruk and Wendelyn along with a second witness whose name was so sloppily written that it was illegible.210

Figure 4: The Signature Page of Taruk Ben-Ali’s Purported Will

Figure 14

Taruk’s half-brother D’Artagnan Lloyd sought to probate the 2002 document.211 Taruk’s intestate heirs, including Wendelyn and Taruk’s daughter from a previous relationship, Brittany Desmond, filed a will contest.212 A subsequent bench trial revolved around two issues: (1) the authenticity of Taruk and Wendelyn’s signatures on the purported will; and (2) the impact of the fact that the second witness could not be identified.213 The parties’ submissions also considered whether section 6110(c)(2) could validate the document if the second witness’s signature was insufficient.214 Based on the testimony of handwriting experts, the probate judge ruled that Taruk and Wendelyn’s signatures were genuine.215 In turn, the judge determined that this finding “len[t] credibility to the unknown signature,” thus validating the instrument.216 Given these conclusions, the judge declared that “there [was] no need to determine if [the harmless error rule] ha[d] been satisfied.”217

The court of appeals reversed.218 The panel first determined that the probate judge had misunderstood the relationship between Taruk and Wendelyn’s signatures and the unreadable signature.219 As the justices explained, there was no authority for the proposition that “the genuineness of the decedent’s signature and of the signature of only one subscribing witness [is] . . . sufficient to establish the genuineness of another witness’s signature.”220 The court then turned its attention to harmless error.221 It held that no reasonable person could have found that there was clear and convincing evidence that Taruk had intended the 2002 writing to be his will.222 The court ticked off several factors that suggested that Hassan had created the document, including: (1) his antagonism toward Wendelyn (whom the instrument effectively disinherited, despite the fact that she and Taruk had just married when he executed it); (2) Hassan’s shocking, protracted cover-up of Taruk’s death; (3) Hassan’s track record of signing Taruk’s name when it served his self-interest; and (4) the absence of any proof that Taruk had mentioned making a will.223

The second contested harmless-error matter, Estate of Delakovias, was also rather wild. In 1993, Anthony Delakovias, who was unmarried and childless, executed a trust that named his cousin Stacie Delakovias as successor­­–trustee and divided his property among some relatives and a longtime friend.224 Anthony also executed a “pour over” will that gave to Stacie as trustee anything that he had failed to transfer to the trust.225 He died on July 26, 2009, with a $135,000 bank account and a house that was worth $130,000 but was encumbered by a $330,000 mortgage.226 Because Anthony had not retitled these assets in the name of the trust, he still owned them in his individual capacity.227 Thus, Stacie filed a petition for probate.228

As Stacie was marshalling Anthony’s estate, she discovered a handwritten document dated January 31, 2009, and entitled “Amendment to Trust Agreement.”229 This document stated that Anthony wanted his house to pass to his roommate of 14 years, Toni Fanfa.230 Apparently, Fanfa had a checkered past which included arrests for drug possession and identity theft.231 Stacie filed a petition for instructions, asking the court to decide whether the amendment was valid.232

Days before the hearing, Fanfa alleged that she had uncovered another writing.233 This typewritten piece of paper was designated “Amendment to my Will,” and was nestled inside a guitar book that Anthony had loaned to a friend.234 It stated “I want [Fanfa] to keep the residence that we shared for the past 14 years and I want my estate to pay off that mortgage.”235 It also declared that Fanfa should receive Anthony’s “accounts [and] monies.”236 It was dated July 2, 2009—three weeks before Anthony passed away—and bore the signatures of Anthony and just one witness.237

Figure 5: Anthony Nicholas Delakovias’s Purported Codicil

Figure 15

Fanfa filed a brief asking the probate court to validate the purported codicil under section 6110(c)(2).238 She bolstered her argument by citing the witness’ declaration that Anthony said that he wanted “Fanfa to be taken care of and . . . had little contact with his own family.”239 Stacie filed a reply in which she questioned Fanfa’s credibility.240 The court set a trial date, but the matter then settled.241

IV.     Policy Implications

This Part discusses how my research informs the debate about the harmless error rule. It first addresses the claim that harmless error spawns litigation. It then examines the relationship between partial harmless error and testamentary intent.

A.     Litigation Rates

Recall that the most common objection to the harmless error rule is that it will open the litigation floodgates. My data tells a different story.

To frame this discussion, I should note that conflict is endemic to Alameda County probate court. As Table 1 reveals, my combined dataset contains 308 cases in which a party filed a request for relief, which means that litigation occurred in 12.5% of all estates.242 Because some estates featured multiple causes of action, there were 368 separate claims. Seventy-four of these were will contests.243

Table 1: Litigation (Combined Datasets)

Table 6

The five harmless error claims were a mere drop in this ocean. As noted above, 1,543 cases, including 873 testacies, opened after July 1, 2009. Accordingly, the doctrine arose in 0.32% of all eligible estates and 0.57% of all eligible testacies.245

In fact, that statistic exaggerates the potential of California’s harmless error rule to generate lawsuits. Neither Estate of Campbell nor Estate of Burns involved adversarial proceedings.246 To be sure, both petitioners needed to file pleadings addressing section 6110(c)(2). But the proponent in Estate of Campbell met her burden with a two-page brief and three short declarations,247 and Estate of Burns’s personal representative actually baked her harmless error allegations into a short paragraph in her petition for probate.248 Even Estate of Ben-Ali and Estate of Delakovias were not “pure” harmless error cases.249 Those matters had already degenerated into litigation when a party invoked section 6110(c)(2). Harmless error thus added a claim to an existing lawsuit, rather than creating a lawsuit out of whole cloth. Indeed, harmless error produced a will contest where none would have otherwise existed just once, in Estate of Caspary.250

There are several possible explanations for this paucity of claims. For one, our expectations have been colored by the 41 harmless error cases that Langbein found in South Australia.251 Although the populations of that country in the 1970s and Alameda County in the 2000s are almost identical—1.4 million252—this proves to be quite misleading. Langbein’s research spanned a decade, whereas my research only covers about four years. Even more importantly, Langbein culled his data from a system that handled about 5,000 estates every year.253 Conversely, thanks to the vociferous contemporary appetite for probate avoidance, Alameda County’s courts only process between 550 and 650 testacies and intestacies annually.254 Given this massive difference in scale, it is not surprising that I found fewer harmless error cases.  

Furthermore, unlike South Australia, California honors holographic wills.255 If a decedent in South Australia handwrites her dispositive wishes and does not have the instrument attested, the result is a section 12(2) case.256 In sharp contrast, a probate court in California can process the same instrument as a matter of course. Because 126 of the wills (9%) in my combined dataset were holographs, this divergence in the substantive law also decreased the raw number of harmless error cases in my sample.

Finally, even massive legal changes can have little immediate impact. Recall that South Australia passed section 12(2) in 1975, but the first harmless error case was not handed down until 1978.257 Likewise, my research, which covers the period immediately after California adopted the law, may have captured the proverbial calm before the storm. Some lawyers may not have known about the rule, or incorrectly assumed that it was not retroactive.258 Indeed, there are a handful of cases in my harmless error-eligible dataset where proponents of a defective instrument inexplicably fail to invoke the doctrine.259

Nevertheless, there are also signs that California’s partial harmless error rule may never generate much litigation. For one, non-compliance with the Wills Act is relatively rare. As Table 2 reveals, the vast majority of will contests arose from allegations of undue influence and incapacity, not from a decedent’s failure to observe the formalities. Indeed, only 18 challenges were grounded on a lack of due execution. Moreover, as Table 3 demonstrates, two-thirds of these cases did not involve attestation errors and thus fell outside the scope of section 6110(c)(2).

Table 2: Causes of Action in Will Contests (Combined Datasets)

Table 7

Table 3: Allegedly Improper Formalities (Combined Datasets)

Table 8

For example, harmless error is irrelevant in what I discovered to be the most common species of will contest: disputes over whether a purported holograph was executed with testamentary intent.262 The Alameda County files are littered with signed, handwritten documents in which a decedent discusses her posthumous wishes.263 Are these scraps of paper, letters, and diary entries supposed to be full-fledged estate plans, blueprints for future wills, or stream-of-consciousness notes? Harmless error cannot help with this question. A party who seeks to admit a casual, handwritten instrument to probate must prove by a preponderance of the evidence that it was supposed to be a will.264 Conversely, harmless error only applies if there is clear and convincing proof that the decedent wanted a non-compliant writing to be a will.265 The doctrine thus asks precisely the same question as traditional law, but imposes an even more onerous burden.266

In addition, California’s partial harmless error rule limits lawsuits by excluding defects related to the signature prong of the Wills Act. As noted, California is one of four partial-harmless-error jurisdictions that insists that a will be signed.267 My review of the Alameda County files suggests that this immutable formality weeds out a fair number of will contests. Although only three adversarial matters featured the issue, there were several unsigned, will-like documents lurking in undisputed estates.268 In fact, this may be the tip of the proverbial iceberg. Because an unsigned writing is per se invalid, potential beneficiaries have no incentive to lodge such an instrument with the probate court. Accordingly, there may be dozens more that were never filed in the record.

Finally, I would be remiss if I did not mention that the specific Wills Act glitches that harmless error was designed to remedy were practically non-existent. Recall that the reform was inspired, in part, by outrage over cases in which courts nullified a purported will due to the fact that the witnesses were not “present at the same time”269 when the testator signed or acknowledged the will.270 This problem arose just once, in an uncontested matter.271 The Examiners spotted the fact that there was a discrepancy between the dates on the will’s signature block, but the decedent’s lawyer submitted an affidavit that clarified that this was a typographical error.272

Likewise, although one of harmless error’s animating purposes is to validate holographic amendments to the face of a previously executed will,273 no litigated case involved this issue. This may be because California is quite liberal in permitting alterations. It has long enforced handwritten changes to attested wills, provided that the changes are signed or initialed.274 I unearthed several such examples, which sailed through probate without contest.275 Also, two attested wills in my harmless error-eligible data bore handwritten but unsigned additions and deletions.276 One was Elizabeth Secrease’s will, which is pictured below. Section 6110(c)(2) does not apply to these revisions for two reasons: (1) it does not govern alterations; and (2) it cannot apply to language that the testator did not sign. And indeed, both cases proceeded along the same trajectory: After the Examiner flagged the matter, the court enforced the wills without the alterations.277

Figure 6: Two Pages from Elizabeth Secrease’s Will

Figure 6.1

Figure 6.2

In sum, my review of the Alameda County files did not bear out the prediction that harmless error would vastly increase the number of will contests. At least in my research, claims that qualify for section 6110(c)(2) are exceedingly rare.278 

B.     Testamentary Intent

This section briefly discusses the relationship between harmless error and testamentary intent. In doing so, it argues first that, in some cases, the doctrine does a much better job than strict compliance in honoring a testator’s wishes. Second, this virtue comes with a price tag: In two kinds of matters, harmless error creates tricky questions about what a decedent was trying to accomplish. Third, partial harmless error statutes like section 6110(c)(2) reduce the number of these tough cases by retaining specific Wills Act elements.

Even in the small pool of light cast by my research, one can see how harmless error can be superior to the rigors of traditional law. Before section 6110(c)(2) became effective, two estates featured eerily similar facts. A decedent typed his dispositive wishes, signed the page, and then had it notarized.279 In the pre-harmless-error era, the outcome was simple: The court denied probate for failure to obtain a second witness.280

These unfortunate outcomes provide a vivid counterpoint to the two uncontested harmless error matters in my sample. The first, Estate of Campbell—in which Billy Campbell typed a half-page love letter to his family—is a veritable poster child for the harmless error rule.281 Although the document was breezy and light-hearted, its personal flourishes established the effort that Campbell poured into it (fulfilling the ritual function) and its authenticity (satisfying the evidentiary and protective objectives).282 The second, Estate of Burns, in which Jack Burns committed suicide, arose from starkly different facts.283 Nevertheless, Burns’s suicide note referenced his “will,” leaving little doubt that he took testation seriously and actually drafted the instrument in question. These cases reinforce harmless error’s core insight: sometimes, the circumstances can be just as important as meticulous adherence to the Wills Act.284

But this holistic approach also has a dark side. As Langbein, Lester, and Baron have observed in both South Australia and America, harmless error can spawn thorny questions about a decedent’s intent.285 In particular, two types of cases recur. In one, a decedent does not execute a will, but leaves behind a document—often, notes or answers to a questionnaire—that is the first step along the path to executing a will. It is unclear what the outcome should be. As a matter of brute statutory interpretation, harmless error should not apply. Even in partial harmless error states, the test is whether there is “clear and convincing evidence that . . . the testator intended [the writing] to constitute the testator’s will.”286 An instrument made in anticipation of making a future will does not meet this benchmark.

Consider again Estate of Caspary.287 The page that Caspary typed on May 25, 2005 expressed his dispositive wishes.288 But Caspary almost certainly did not regard it as his will. After all, if he thought that it was effective, why did he meet with his lawyer the next day to discuss executing a formal testamentary instrument?289 Then again, according to Caspary’s friends, he was a sloppy genius: both brilliant and “quite disorganized regarding his documents and personal records.”290 He might have intended to sign the writing his lawyer had prepared, but never got his act together.

There is also a second category of problematic harmless error matters. As Langbein observed in his seminal article, some decedents know that they have violated the Wills Act.291 There are flickers of this issue in my research. Even in Estate of Burns, which I praised above, Jack Burns recognized that his Quicken-prepared “[w]ill” was deficient.292 As his suicide note admitted, he “did not have [the will] witnessed,” but was “assuming that something is better than nothing.”293 Similarly, in Estate of Delakovias, the sole witness to Anthony Delakovias’s purported codicil testified that “he also told [Anthony] to have it witnessed by another person or notary for it to be valid.”294 It is hard to know what to do with this information. In some situations, the decedent’s failure to follow up and correct the mistake might be seen as evidence that she had second thoughts about the way the instrument distributed her property.  

Although these issues are beyond the scope of this Article, I do want to emphasize that they tip the scales toward partial harmless error’s policy of retaining certain will-creation formalities as indispensable. Even if one finds Estate of Caspary, Estate of Burns, and Estate of Delakovias problematic, at least they featured wills that the testator had signed. The testator’s signature is a badge of authenticity, separates drafts from final versions, and reinforces the solemnity of testation. Predicating harmless error on a writing bearing this imprimatur, as California does, alleviates some of the creeping doubt that flows from non-compliance with the Wills Act. A signature is potent evidence that a decedent assented to the terms of a will-like document—an anchor in the murky waters of testamentary intent.

V.     Conclusion

In one large California county, the harmless error rule did not lead to a surge in litigation. Of course, more research is necessary to examine whether my results are generalizable, or whether more disputes emerge as the statute settles into its new role. In addition, my data suggest that partial harmless error can be a risk-free baby step for states that want to liberate their “citizens [from] the injustice of the traditional law,”295 but are wary about veering too far to the other side.

  1. [1]. Janet Gilmore, Medieval History Professor, Devoted to Students, Dead at 79, UCBerkeleyNews (Apr. 9, 2008),; see also Paula S. Fass et al., In Memoriam: Gerard Ernest Caspary, Univ. of Cal., http:// (last visited
    Mar. 19, 2018).

  2. [2]. See Gilmore, supra note 1.

  3. [3]. See id.

  4. [4]. See id.

  5. [5]. See Fass et al., supra note 1; see also Gilmore, supra note 1.

  6. [6]. See Fass et al., supra note 1.

  7. [7]. See id.

  8. [8]. See Gilmore, supra note 1; see also Fass et al., supra note 1.

  9. [9]. See Fass et al., supra note 1.

  10. [10]. See id.

  11. [11]. See Opposition to Petition for Prob. of Will at 2, Estate of Caspary, No. RP08396884 (Cal. Super. Ct. Mar. 27, 2009).

  12. [12]. See Petition for Prob. of Will & for Letters Testamentary at 8, Estate of Caspary, No. RP08396884 (Cal. Super. Ct. Jan. 28, 2009).

  13. [13]. Id.

  14. [14]. Id. (emphasis added).

  15. [15]. Id.

  16. [16]. Declaration of Michael Ferguson at exhibit A, Estate of Caspary, No. RP08396884 (Cal. Super. Ct. Apr. 1, 2009).

  17. [17]. See id.

  18. [18]. See id.

  19. [19]. See id. at 1–2.

  20. [20]. See id.

  21. [21]. See Fass et al., supra note 1.

  22. [22]. See John H. Langbein, Substantial Compliance with the Wills Act, 88 Harv. L. Rev. 489, 489 (1975).

  23. [23]. See, e.g., Cal. Prob. Code § 6110 (West 2017).

  24. [24]. In re Sage, 107 A. 445, 445 (N.J. 1919).

  25. [25]. Petition for Letters of Administration at 1, Estate of Caspary, No. RP08396884 (Cal. Super. Ct. July 7, 2008); see also Alameda County Public Administrator, Alameda Cty. Soc. Servs. Agency,
    public_administrator.cfm (last visited Mar. 19, 2018) (providing general information about the Public Administrator’s services).

  26. [26]. See Cal. Prob. Code § 6110(c).

  27. [27]. Bruce H. Mann, Formalities and Formalism in the Uniform Probate Code, 142 U. Pa. L. Rev. 1033, 1035 (1994).

  28. [28]. See Cal. Prob. Code § 6110(c)(2).

  29. [29]. See id.

  30. [30]. Petition for Prob. of Will & for Letters Testamentary at 1, Estate of Caspary, No. RP08396884 (Cal. Super. Ct. Jan. 28, 2009).

  31. [31]. See Declaration of Leonard Johnson in Support of Petition for Prob. of Will of Gerard E. Caspary at 1–2, Estate of Caspary, No. RP08396884 (Cal. Super. Ct. Jan. 28, 2009); Declaration of Paula S. Fass in Support of Petition for Prob. of Will of Gerard E. Caspary at 1–2, Estate of Caspary, No. RP08396884 (Cal. Super. Ct. Mar. 12, 2009); Declaration of John E. Lesch (a.k.a. Jack Lesch) in Support of Petition for Prob. of Will of Gerard E. Caspary at 1–2, Estate of Caspary, No. RP08396884 (Cal. Super. Ct. Mar. 12, 2009).

  32. [32]. See Receipt on Final Distrib. at 1, Estate of Caspary, No. RP08-396884 (Cal. Super. Ct. Dec. 31, 2015).

  33. [33]. Wills Act 1936 (SA) pt II s 12(2) (Austl.). In fact, Israel may have adopted a similar rule even earlier, in 1965. Samuel Flaks, Excusing Harmless Error in Will Execution: The Israeli Experience, 3 Est. Plan. & Community Prop. L.J. 27, 28–29 (2010). However, because the country’s Supreme Court interpreted it narrowly and the legislature has amended it several times, its original contours are slightly unclear. See id. at 48–53.

  34. [34]. See generally John H. Langbein, Excusing Harmless Errors in the Execution of Wills: A Report on Australia’s Tranquil Revolution in Probate Law, 87 Colum. L. Rev. 1 (1987) (exploring Australia’s transition to relaxing the requirement of strict adherence to execution laws for a will to be valid).

  35. [35]. See infra notes 111–18 and accompanying text.

  36. [36]. See Unif. Probate Code §§ 2-503, 2-805 (Unif. Law. Comm’n 2010); Restatement (Third) of Prop.: Wills and Donative Transfers § 12.1 (Am. Law. Inst. 2003).

  37. [37]. See Cal. Prob. Code § 6110(c)(2) (West 2017); Colo. Rev. Stat. § 15-11-503 (2017); Haw. Rev. Stat. § 560:2-503 (2006); Mich. Comp. Laws § 700.2503 (2002); Mont. Code Ann. § 72-2-523 (West 2017); N.J. Stat. Ann. § 3B:3-3 (West 2017); Ohio Rev. Code Ann. § 2107.24 (LexisNexis 2016); S.D. Codified Laws § 29A-2-503 (West 2018); Utah Code Ann. § 75-2-503 (LexisNexis 2017); Va. Code Ann. § 64.2-404 (2017).

  38. [38]. See infra text accompanying notes 134–40.

  39. [39]. Mark Glover, In Defense of the Harmless Error Rule’s Clear and Convincing Evidence Standard: A Response to Professor Baron, 73 Wash. & Lee L. Rev. Online 289, 296 (2016).

  40. [40]. See Langbein, supra note 34, at 51. As Langbein noted, Israel had followed a version of the harmless error rule for more than a decade, and a well-known judge and expert on probate law had expressed the view that the doctrine “actually prevents a great deal of unnecessary litigation.” Id. at 50.

  41. [41]. See id. at 13–15. As I mention infra notes 107–09 and accompanying text, this number was exaggerated because South Australian courts at the time did not allow the parties to settle will contests.

  42. [42]. See Stephanie Lester, Admitting Defective Wills to Probate, Twenty Years Later: New Evidence for the Adoption of the Harmless Error Rule, 42 Real Prop., Prob. & Tr. J. 577, 584–86 (2007).

  43. [43]. Jane B. Baron, Irresolute Testators, Clear and Convincing Wills Law, 73 Wash. & Lee L.
    Rev. 3, 21 (2016).

  44. [44]. See Statute of Frauds 1677, 29 Car. 2 c 3, § 5 (Eng.).

  45. [45]. See Wills Act 1837, 1 Vict. c. 26 (Eng.).

  46. [46]. Id. § 9.

  47. [47]. See, e.g., Jesse Dukeminier et al., Wills, Trusts, and Estates 225–27 (8th ed. 2009).

  48. [48]. In re Harty’s Estate, 148 N.Y.S. 1052, 1053 (Sur. Ct. 1914).

  49. [49]. See, e.g., In re Groffman, (1968) 1 W.L.R. 733, 739 (Eng.).

  50. [50]. In re Lynch’s Estate, 161 P.2d 24, 26 (Cal. Dist. Ct. App. 1945).

  51. [51]. Id. at 27.

  52. [52]. Id.

  53. [53]. Id. (quoting 1 William Herbert Page, A Treatise on the Law of Wills § 344, at 640 (3d ed. 1941)).

  54. [54]. Langbein, supra note 22. But see Peter T. Wendel, Wills Act Compliance and the Harmless Error Approach: Flawed Narrative Equals Flawed Analysis?, 95 Or. L. Rev. 337, 342 (2017) (arguing that the conventional account of Wills Act formalism is exaggerated).

  55. [55]. See, e.g., Stephen Clowney, In Their Own Hand: An Analysis of Holographic Wills and Homemade Willmaking, 43 Real Prop., Tr. & Est. L.J. 27, 37–38 & 37 n.42 (2008) (“At mid-century, only 19 states allowed holographic wills. Today, 27 jurisdictions authorize holographic wills.” (citation omitted)); see also E.G.L., Comment, Holographic Wills and Their Dating, 28 Yale L.J. 72, 72 (1918) (tracing the adoption of holographs to the influence of the Roman Empire and Napoleonic Code).

  56. [56]. See, e.g., Cal. Prob. Code § 6111(a) (West 2017).

  57. [57]. See, e.g., In re Kimmel’s Estate, 123 A. 405, 406 (Pa. 1924) (“The letter now being considered was all in the handwriting of decedent ... .”).

  58. [58]. See John Marshall Gest, Some Jolly Testators, 8 Temp. L.Q. 297, 301 (1934).

  59. [59]. See Geoff Ellwand, An Analysis of Canada’s Most Famous Holograph Will: How a Saskatchewan Farmer Scratched His Way into Legal History, 77 Sask. L. Rev. 1, 1 (2014).

  60. [60]. See, e.g., Maris v. Adams, 166 S.W. 475, 478 (Tex. Civ. App. 1914) (holding that a bequest was not a holographic will because it was not completely handwritten).

  61. [61]. In re Thorn’s Estate, 192 P. 19, 19 (Cal. 1920).

  62. [62]. Id.

  63. [63]. See id. at 20, 22.

  64. [64]. See generally Lon L. Fuller, Consideration and Form, 41 Colum. L. Rev. 799, 799–804 (1941) (analyzing the purposes of legal formalities with a focus on contract’s consideration doctrine); Ashbel G. Gulliver & Catherine J. Tilson, Classification of Gratuitous Transfers, 51 Yale L.J. 1, 5–13, (1941) (examining formalities in will execution).

  65. [65]. Gulliver & Tilson, supra note 64, at 4; see Fuller, supra note 64, at 800.

  66. [66]. Gulliver & Tilson, supra note 64, at 6.

  67. [67]. Id. at 5–6, 9 n.25; Fuller, supra note 64, at 800.

  68. [68]. Fuller, supra note 64, at 800; Gulliver & Tilson, supra note 64, at 5.

  69. [69]. See Fuller, supra note 64, at 801–03.

  70. [70]. Id. at 801.

  71. [71]. Langbein, supra note 22, at 494.

  72. [72]. See Gulliver & Tilson, supra note 64, at 9–13.

  73. [73]. See id. at 9–10.

  74. [74]. See, e.g., Langbein, supra note 22, at 496–97; James Lindgren, Abolishing the Attestation Requirement for Wills, 68 N.C. L. Rev. 541, 554–55 (1990); Bruce H. Mann, Self-Proving Affidavits and Formalism in Wills Adjudication, 63 Wash. U. L.Q. 39, 49 (1985).

  75. [75]. See Langbein, supra note 22, at 515–16.

  76. [76]. See generally Norman F. Dacey, How to Avoid Probate! (5th ed. 1993).

  77. [77]. See, e.g., John H. Langbein, The Nonprobate Revolution and the Future of the Law of Succession, 97 Harv. L. Rev. 1108, 1109–15 (1984) (discussing will substitutes including trusts and life insurance).

  78. [78]. See supra notes 74–76 and accompanying text.

  79. [79]. See generally Law Reform Comm. of S. Austl., Twenty-Eighth Report of the Law Reform Committee of South Australia to the Attorney-General: Relating to the Reform of the Law on Intestacy and Wills (1974) (making several recommendations about improving the country’s inheritance laws).

  80. [80]. See id. at 5–10.

  81. [81]. Id. at 10–11.

  82. [82]. Wills Act Amendment Act 1975 (SA) s 9 (Austl.) (repealing section 12(2) of the Wills Act and inserting the quoted language).

  83. [83]. See, e.g., Leigh A. Shipp, Comment, Equitable Remedies for Nonconforming Wills: New Choices for Probate Courts in the United States, 79 Tul. L. Rev. 723, 724 (2005).

  84. [84]. See generally Langbein, supra note 34 (discussing the South Australian experiment of excusing harmless errors in the execution of wills). Langbein began with a primer on the South Australian succession process. See id. at 9–12. The county’s default mode of estate administration is “common form,” in which the probate Registrar handles routine estates. See id. at 12. If the registrar encounters a difficult question, she refers the matter to the court of general jurisdiction, which is called the Supreme Court. Id. at 12–13. In those chambers, the judge conducts a desk trial, where the parties submit affidavits rather than testify. Id. at 12. Conversely, contested matters are heard in “solemn form,” which involves the full panoply of live witnesses and cross-examination. See id. at 12–13. The Supreme Court can either write an opinion for publication (“reported cases”) or issue a bare judgment (“unreported cases”). See id. at 13–14. Because section 12(2), as originally drafted, expressly conferred “its ... power upon the Supreme Court, not the registrar of probates,” all harmless error cases were adjudicated at that level. Id. at 13.

  85. [85]. Id. at 9, 13–14, 15 n.54.

  86. [86]. See id. at 16.

  87. [87]. Id. at 15. Twenty-one had resulted in published opinions. Id. at 13.

  88. [88]. See id. at 16–23.

  89. [89]. See id. at 16–18.

  90. [90]. See id. at 18–23.

  91. [91]. See id. at 22–29.

  92. [92]. Id. at 22.

  93. [93]. See id. at 29–33.

  94. [94]. Id. at 29.

  95. [95]. See id. at 2, 29.

  96. [96]. Id. at 31 (emphasis omitted).

  97. [97]. See id. at 18–21.

  98. [98]. Id. at 32–33.

  99. [99]. Id. at 20–21.

  100. [100]. Id.

  101. [101]. Id. at 32–33.

  102. [102]. See id. at 34–37.

  103. [103]. Wills Act 1936 (SA) s 12(2), as modified by Wills Act Amendment Act 1975 (SA) s 9 (Austl.) (emphasis added).

  104. [104]. Langbein, supra note 34, at 34 (quoting In re Estate of Kelly (1983) 34 SASR 370, 384 (Austl.)).

  105. [105]. Id. at 35–37.

  106. [106]. See id. at 51–54.

  107. [107]. Id. at 37.

  108. [108]. Id. at 37–38.

  109. [109]. See id.

  110. [110]. Id. at 54.

  111. [111]. See Wills Act 1968 (ACT) pt 11A (Austl.).

  112. [112]. See Wills, Probate and Administration Act 1898 (NSW) pt 18A (Austl.).

  113. [113]. See Wills Act 2015 (NT) s 12 (Austl.).

  114. [114]. See Wills Act 1992 (Tas) pt 2 div 7 s 26 (Austl.).

  115. [115]. See Wills Act 1970 (WA) ss 32–33 (Austl.). The Australian province of Queensland also passed a palliative doctrine known as substantial compliance. See Succession Act 1981 (Qld) pt II s 9(a) (Austl.). However, Langbein discovered that courts “ha[d] interpreted the measure so narrowly as to render it nearly useless.” Langbein, supra note 34, at 41. But see Lester, supra note 42, at 598–99 (finding that Queensland judges had become slightly more lenient over the years).

  116. [116]. See Langbein, supra note 34, at 48–51.

  117. [117]. See Wills Act, C.C.S.M. 2014, c W150 (Can.).

  118. [118]. See Wills Act, 1996, S.S. 1996, c W-14.1, s 7, amended by S.S. 2001, c 51 (Can.).

  119. [119]. See Lester, supra note 42, at 583–84. The 1994 amendments to section 12(2) also eliminated the mandate that only the Supreme Court could apply the statute. In turn, this allowed the Probate Registrar to dispose of straightforward harmless error matters, instead of automatically transmitting them to a trial judge. See id.

  120. [120]. Wills (Miscellaneous) Amendment Act 1994 (SA) s 7 (Austl.).

  121. [121]. In re Estate of TLB (2005) 94 SASR 450, 452–53 (Austl.).

  122. [122]. See id. at 451.

  123. [123]. See id. at 457.

  124. [124]. Statutes Amendment and Repeal (Attorney-General’s Portfolio) Act 2000 (SA) pt 15 (Austl.) (amending Wills Act 1936 (SA) s 12(2) (Austl.)).

  125. [125]. Lester, supra note 42, at 582–84.

  126. [126]. Id. at 584 n.29, 607. Overall, Lester examined 60 South Australian cases from two sources. First, she read the reported and unreported section 12(2) decisions that the Supreme Court had handed down between January 1986 and March 2007. Id. at 584. Second, she pulled every harmless error case that the Registrar had resolved in 2005 and 2006. Id. She noted that her original research plan—to review all Registrar rulings since 1986—had been foiled by the skyrocketing number of section 12(2) applications. Id. For instance, she found 32 such petitions in the span of just two years. Id. at 584 n.29. Lester also augmented her research with 61 cases from the Supreme Court of New South Wales. Id. at 585–86.

  127. [127]. See id. at 589–90 (stating that judges “felt more comfortable probating unattested documents with somewhat questionable testator intent” once they felt familiar with the amendments).

  128. [128]. The court explained “that (1) proposed can mean intended, and (2) the document was found with other ‘important documents’ like the decedent’s school picture.” Id. at 589.

  129. [129]. Id. at 591.

  130. [130]. Id.

  131. [131]. Unif. Probate Code §§ 2-503, 2-805 (Unif. Law Comm’n 2010); Restatement (Third) of Prop.: Wills & Donative Transfers § 12.1 (Am. Law Inst. 2003).

  132. [132]. Unif. Probate Code § 2-503. As Langbein would later explain, the “prototypical” harmless error case is one in which a witness “steps out of the room to powder her nose before the other has completed signing.” John H. Langbein, Major Reforms of the Property Restatement and the Uniform Probate Code: Reformation, Harmless Error, and Nonprobate Transfers, 38 ACTEC L.J. 1, 9 (2012).

  133. [133]. Unif. Probate Code § 2-503.

  134. [134]. Colo. Rev. Stat. § 15-11-503 (2017); Haw. Rev. Stat. § 560:2-503 (2006); Mich. Comp. Laws § 700.2503 (2002); Mont. Code Ann. § 72-2-523 (West 2017); S.D. Codified Laws § 29A-2-503 (West 2018); Utah Code Ann. § 75-2-503 (West 2017); Unif. Probate Code
    § 2-503 (Unif. Law Comm’n 2010).

  135. [135]. Cal. Prob. Code § 6110(c)(2) (West 2017); N.J. Stat. Ann. § 3B:3-3 (West 2017); Ohio Rev. Code Ann. § 2107.24 (LexisNexis 2016); Va. Code Ann. § 64.2-404 (2017).

  136. [136]. I discuss California’s harmless error statute in depth infra notes 154–78 and accompanying text.

  137. [137]. Va. Code Ann. § 64.2-404(B)(i) (2017). The statute also carves out a narrow exception for “circumstances where two persons mistakenly sign each other’s will, or a person signs the self-proving certificate to a will instead of signing the will itself.” Id.

  138. [138]. See Ohio Rev. Code Ann. § 2107.24(A)(3) (LexisNexis 2016). Ohio’s Wills Act is strict: It requires testamentary dispositions to be (1) “signed at the end by the testator” and
    (2) “subscribed in the conscious presence of the testator, by two or more competent witnesses, who saw the testator subscribe, or heard the testator acknowledge the testator’s signature.”
    Id. § 2107.03 (emphasis added). The state’s harmless error statute creates a little more wiggle room by permitting the judge to admit a document if there is compelling evidence that “[t]he decedent signed the document” (presumably anywhere) “within the range of any of the witnesses’ senses, excluding the sense of sight or sound that is sensed by telephonic, electronic, or other distant communication.” Id. § 2107.24(A)(3) (emphasis added).

  139. [139]. 2001 Colo. Sess. Laws 887. In 2001, Colorado amended its statute to clarify that harmless error “shall apply only if the document is signed or acknowledged by the decedent as his or her will or if it is established by clear and convincing evidence that the decedent erroneously signed a document intended to be the will of the decedent’s spouse.” Colo. Rev. Stat. § 15-11-503(2) (2017).

  140. [140]. See In re Estate of Ehrlich, 47 A.3d 12, 23 (N.J. Super. Ct. App. Div. 2012) (Skillman, J., dissenting). But see In re Estate of Attia, 895 N.W.2d 564, 568 (Mich. Ct. App. 2016) (“[A] will does not need to be signed in order to be admitted to probate under [the UPC’s harmless error rule] ... .”); In re Prob. of Will & Codicil of Macool, 3 A.3d 1258, 1265 (N.J. Super. Ct. App. Div. 2010) (opining that the UPC’s harmless error statute validates unsigned writings as long as the proponent “prove[s] by clear and convincing evidence, that: (1) the decedent actually reviewed the document in question; and (2) thereafter gave his or her final assent to it”); cf. In re Estate of Vendola, No. P-119-13, 2014 WL 4744651, at *1 (N.J. Super. Ct. App. Div. Sept. 4, 2014) (refusing to admit to probate a draft that the decedent had neither “seen, read, [n]or signed”).

  141. [141]. See, e.g., C. Douglas Miller, Will Formality, Judicial Formalism, and Legislative Reform: An Examination of the New Uniform Probate Code “Harmless Error” Rule and the Movement Toward Amorphism, 43 Fla. L. Rev. 599, 705 (1991); see also Baron, supra note 43; Lloyd Bonfield, Reforming the Requirements for Due Execution of Wills: Some Guidance from the Past, 70 Tul. L. Rev. 1893, 1915 (1996); Adam J. Hirsch, Formalizing Gratuitous and Contractual Transfers: A Situational Theory, 91 Wash. U. L. Rev. 797, 829 (2014); Daniel B. Kelly, Toward Economic Analysis of the Uniform Probate Code, 45 U. Mich. J.L. Reform 855, 881 (2012); Kelly A. Hardin, Note, An Analysis of the Virginia Wills Act Formalities and the Need for a Dispensing Power Statute in Virginia, 50 Wash.
    & Lee L. Rev. 1145, 1181–83 (1993).

  142. [142]. Emily Sherwin, Clear and Convincing Evidence of Testamentary Intent: The Search for a Compromise Between Formality and Adjudicative Justice, 34 Conn. L. Rev. 453, 470–73 (2002).

  143. [143]. Id. at 472–73.

  144. [144]. See generally Baron, supra note 43 (exploring the difficulties courts face in applying the “clear and convincing” standard to testator intent and harmless errors). For lucid explanations of why testamentary intent is such a confusing concept, see generally Mark Glover, A Taxonomy of Testamentary Intent, 23 Geo. Mason L. Rev. 569 (2016) (explaining that courts apply several subtly different versions of the concept); Katheleen R. Guzman, Intents and Purposes, 60 U. Kan. L. Rev. 305, 306 (2011) (calling testamentary intent “an extraordinarily elusive concept”); and James Lindgren, The Fall of Formalism, 55 Alb. L. Rev. 1009, 1017 (1992) (“[T]estamentary intent is not well understood or defined.”).

  145. [145]. Baron, supra note 43, at 7.

  146. [146]. Id.

  147. [147]. Id. at 73.

  148. [148]. See id. at 29.

  149. [149]. See id. at 38.

  150. [150]. Id.

  151. [151]. Id. at 8.

  152. [152]. See Estate of Ben-Ali, 157 Cal. Rptr. 3d 353, 361 (Cal. Ct. App. 2013); Estate of Stoker, 122 Cal. Rptr. 3d 529, 534 (Cal. Ct. App. 2011); In re Estate of Wiltfong, 148 P.3d 465, 467–68 (Colo. App. 2006); In re Estate of Sky Dancer, 13 P.3d 1231, 1233–34 (Colo. App. 2000); In re Estate of Attia, 895 N.W.2d 564, 568 (Mich. Ct. App. 2016); In re Estate of Hall, 51 P.3d 1134, 1136 (Mont. 2002); In re Estate of Ehrlich, 47 A.3d 12, 18 (N.J. Super. Ct. App. Div. 2012); In re Prob. of the Will & Codicil of Macool, 3 A.3d 1258, 1266 (N.J. Super. Ct. App. Div. 2010); In re Estate of Hand, 73 N.E.3d 880, 884 (Ohio Ct. App. 2016).

  153. [153]. Doherty v. Doherty, Nos. C073408, C073423, 2014 WL 7399373, at *9 (Cal. Ct. App. Dec. 29, 2014); Estate of Lara, No. H039060, 2014 WL 2108962, at *6 (Cal. Ct. App. May 20, 2014); Estate of Reese, No. D060749, 2013 WL 64379, at *7 (Cal. Ct. App. Jan. 7, 2013); Estate of Richards, No. B226261, 2011 WL 6062018, at *5 (Cal. Ct. App. Dec. 21, 2011); Estate of Patterson, No. D055461, 2010 WL 1820221, at *5 (Cal. Ct. App. May 7, 2010); In re Estate of Southworth, No. 297460, 2011 WL 2623381, at *3 (Mich. Ct. App. July 5, 2011) (per curiam); In re Estate of Windham, No. 287937, 2010 WL 293064, at *1 (Mich. Ct. App. Jan. 26, 2010) (per curiam); In re Estate of Smoke, No. 273114, 2007 WL 4415499, at *2 (Mich. Ct. App. Dec. 18, 2007) (per curiam); In re Estate of Berg, No. 268584, 2006 WL 2482895, at *23 (Mich. Ct. App. Aug. 29, 2006) (per curiam); In re Estate of Vendola, No. P-119-13, 2014 WL 4744651,
    at *2 (N.J. Super. Ct. App. Div. Sept. 4, 2014) (per curiam).

  154. [154]. David Horton, Tomorrow’s Inheritance: The Frontiers of Estate Planning Formalism, 58 B.C. L. Rev. 539, 581–82 (2017). See generally David Horton, In Partial Defense of Probate: Evidence from Alameda County, California, 103 Geo. L.J. 605 (2015) (using this dataset to study the probate process); David Horton & Andrea Cann Chandrasekher, Probate Lending, 126 Yale L.J. 102 (2016) (using the dataset to analyze companies that pay heirs and beneficiaries cash advances in return for shares of their inheritance); David Horton, Wills Law on the Ground, 62 UCLA L. Rev. 1094 (2015) (using this dataset to examine substantive wills doctrine).

  155. [155]. DomainWeb, Superior Court of Cal.: Cty. of Alameda, http://www.alameda.courts. (last visited Mar. 27, 2018).

  156. [156]. When I first began using DomainWeb for research in 2014, the website made all filings in testate and intestate probate administrations available for free. Unfortunately, it soon began limiting access to the right half of all documents and charging users about $1 per page to download the entire page. How This Site Works, Superior Court of Cal.: Cty. of Alameda, (last visited Mar. 27, 2018). Thus, I was not able to gather as much information as I did in my initial Alameda County projects.

  157. [157]. The 1,411 testacies distributed an average estate value of $691,899 in an average of 485 days. The 1,042 intestacies transmitted a mean of $482,258 in a mean of 637 days.

  158. [158]. See, e.g., Cal. Prob. Code §§ 8226, 8270 (West 2017).

  159. [159]. In fact, as I mentioned in the Introduction, the first harmless error case arose from an estate that was filed on July 7, 2008. See Petition for Prob. of Will & for Letters Testamentary at 1, Estate of Caspary, No. RP08396884 (Cal. Super. Ct. Jan. 28, 2009); see also supra text accompanying notes 1–32.

  160. [160]. The Probate Examiner is the rough equivalent of the South Australian Registrar.

  161. [161]. Indeed, my dataset contains several examples of the court refusing to enforce a purported will although no contestant challenged it. See, e.g., Petition for Prob.: Alameda Cty. Prob. Exam’r Checksheet at 1, Estate of Carroll, No. RP09491388 (Cal. Super. Ct. Mar. 22, 2010) (noting that a purported codicil had been only signed by one witness); see also First Codicil to the Pour Over Will of Michael J. Carroll at 1–2, Estate of Carroll, No. RP09491388 (Cal. Super. Ct. Dec. 29, 2009). Likewise, the Examiners’ checksheet frequently prompted the parties to address potential defects with supplemental briefs. See, e.g., Petition for Prob.: Alameda Cty. Prob. Exam’r Checksheet at 2, Estate of Latimer, No. RP10538225 (Cal. Super. Ct. Mar. 7, 2011) (mentioning that the proponent of a codicil had filed a memorandum of points and authorities to address the issue that a potential holographic codicil had been initialed by the testator, but not signed).

  162. [162]. See Cal. Prob. Code § 13500 (West 2017).

  163. [163]. California’s status as a community property state also makes it unique. All of the other American jurisdictions that have adopted the harmless error rule are common law property states. In an even starker difference, South Australia protects surviving spouses through its family maintenance regime. See, e.g., Langbein, supra note 34, at 11–12.

  164. [164]. Cal. Prob. Code § 6110(c) (emphasis added).

  165. [165]. Id.

  166. [166]. Id. §§ 6110–11.

  167. [167]. See supra notes 136–39 and accompanying text.

  168. [168]. Cal. Prob. Code § 6111(a); see also Peter T. Wendel, California Probate Code Section 6110(C)(2): How Big Is the Hole in the Dike?, 41 Sw. L. Rev. 387, 399–406 (2012) (looking to plain language and legislative history to interpret the scope of section 6110).

  169. [169]. Cal. Prob. Code § 6111.

  170. [170]. Id. § 6110 (c)(2).

  171. [171]. Id. § 6110(c).

  172. [172]. Moreover, as a matter of brute grammar, an unattested, Frankensteinian, handwritten/typewritten “holograph” satisfies the textual hook for applying the rule: It “was not executed in compliance with” section 6110(c)(1). Id. § 6110(c)(2) (emphasis added).

  173. [173]. California’s first reported appellate decision involving section 6110(c)(2) almost fell into this doctrinal quicksand. In Estate of Stoker, Wayne Stoker asked a third party to handwrite his dispositive wishes on a piece of paper. Stoker signed the document in front of two witnesses, who did not also subscribe it. Estate of Stoker, 122 Cal. Rptr. 3d 529, 532 (Cal. Ct. App. 2011). Given this abundant evidence that Stoker wanted the instrument to be his will, the contestants were limited to the bright-line argument that the harmless error statute does not govern “cases involving handwritten documents.” Id. at 534. The court rejected this theory, explaining that the rule governs “wills that are ‘in writing’ and signed by the testator.” Id. (quoting Cal. Prob. Code § 6110(a)–(b)(1)); however, in this case the relevant “document is a written will signed by decedent.” Id. But the court stopped short of extending California’s harmless error rule to holographs, observing only that “handwritten non-holographic wills are not excluded from the scope of this statute.” Id. (emphasis added).

  174. [174]. See supra notes 132–33 and accompanying text.

  175. [175]. Unif. Probate Code § 2-503 (Unif. Law Comm’n 2010).

  176. [176]. Cal. Legis. Assemb. B. 2248, 2007–08 Leg., Reg. Sess. (Cal. 2008), http://www.

  177. [177]. S. Judiciary Comm., Bill Analysis, Assemb. B 2248, 2007–2008 Leg., Reg. Sess.
    (Cal. 2008),

  178. [178]. Id.

  179. [179]. See supra text accompanying notes 1–32.

  180. [180]. See Will of Billy G. Campbell at 1, Estate of Campbell, HP10543410 (Cal. Super. Ct.
    Oct. 25, 2010).

  181. [181]. Id.

  182. [182]. Id.

  183. [183]. Petition for Prob.: Alameda Cty. Prob. Exam’r Checksheet at 1, Estate of Campbell, No. HP10543410 (Cal. Super. Ct. Nov. 23, 2010).

  184. [184]. See Declaration by Petitioner in Response to Prob. Exam’r’s Notes at 2–3, Estate of Campbell, No. HP10543410 (Cal. Super. Ct. Nov. 29, 2010); see also Declaration of Edwin A. Weatherton in Support of Petition for Prob. of Will at 1–2, Estate of Campbell, No. HP10543410 (Cal. Super. Ct. Nov. 3, 2010); Declaration of Elizabeth L. Nay in Support of Petition for Prob. of Will at 1–2, Estate of Campbell, No. HP10543410 (Cal. Super. Ct. Nov. 3, 2010); Declaration of Sarah A. Hayse in Support of Petition for Prob. of Will at 1–2, Estate of Campbell, No. HP10543410 (Cal. Super. Ct. Nov. 3, 2010).

  185. [185]. See Order for Prob. at 1, Estate of Campbell, No. HP10543410 (Cal. Super. Ct. Dec. 1, 2010).

  186. [186]. See Ex Parte Petition for Letters of Special Admin. at attachment 3e(2) p. 1, Estate of Burns, No. FP09469690 (Cal. Super. Ct. Aug. 20, 2009).

  187. [187]. Id.

  188. [188]. Id. at attachment 3e(2) p. 3.

  189. [189]. See id. at attachment 3e(2) pp. 8–10.

  190. [190]. See id.

  191. [191]. Id. at attachment 3e(2) p. 24.

  192. [192]. Id. at attachment 3e(2) pp. 3, 26.

  193. [193]. Id. at attachment 3e(2) pp. 1–26.

  194. [194]. Id. at attachment 3e(2) p. 1.

  195. [195]. See Order for Prob. at 1, Estate of Burns, No. FP09469690 (Cal. Super. Ct. Sept. 18, 2009).

  196. [196]. See generally Estate of Ben-Ali, 157 Cal. Rptr. 3d 353 (Cal. Ct. App. 2013).

  197. [197]. Id. at 354–55.

  198. [198]. Id. at 354.

  199. [199]. Id. at 355.

  200. [200]. Id.

  201. [201]. See id.

  202. [202]. Id.

  203. [203]. Id.

  204. [204]. Id.

  205. [205]. Id.

  206. [206]. Id.

  207. [207]. Id.

  208. [208]. Id.

  209. [209]. Id. at 356.

  210. [210]. Id.

  211. [211]. Id. at 356. For reasons that are not apparent from the record, Ivan Golde, who was named as the alternate executor, also filed a petition for probate. Id.

  212. [212]. See id.

  213. [213]. See id. at 1–2.

  214. [214]. See Petitioner Wendelyn Evyonne Wilburn’s Closing Argument at 3–4, Estate of Ben-Ali, No. HP09432597 (Cal. Super. Ct. May 16, 2011); Post Trial Brief of Brittany Desmond at 4, Estate of Ben-Ali, No. HP09432597 (Cal. Super. Ct. May 13, 2011).

  215. [215]. See Final Statement of Decision & Judgment at 3–5, Estate of Ben-Ali, No. HP09432597 (Cal. Super. Ct. June 27, 2011), rev’d, 157 Cal. Rptr. 3d 353 (Cal. Ct. App. 2013).

  216. [216]. Id. at 5.

  217. [217]. Id. at 6.

  218. [218]. Estate of Ben-Ali, 157 Cal. Rptr. 3d 353, 361–62 (Cal. Ct. App. 2013).

  219. [219]. Id. at 358–61.

  220. [220]. Id. at 360.

  221. [221]. See id. at 361–62.

  222. [222]. Id.

  223. [223]. Id.

  224. [224]. Petition to Determine Persons Entitled to Distribution of Estate at 1–3, Estate of Delakovias, No. HP09481108 (Cal. Super. Ct. May 9, 2011).

  225. [225]. Id. at 1–2.

  226. [226]. Id. at 2, 6.

  227. [227]. Id. at 2.

  228. [228]. Id.

  229. [229]. Id. at 4.

  230. [230]. Id.

  231. [231]. See id. at 7–8.

  232. [232]. See id. at 4–5.

  233. [233]. Id. at 5.

  234. [234]. Id.

  235. [235]. Original of Amendment to Will at attachment, Estate of Delakovias, No. HP09481108 (Cal. Super. Ct. Aug. 26, 2010).

  236. [236]. Id.

  237. [237]. Id.

  238. [238]. See Opposition to Petition to Determine Persons Entitled to Distribution of Estate at
    3–4, Estate of Delakovias, No. HP09481108 (Cal. Super. Ct. May 12, 2011).

  239. [239]. Id. at 3.

  240. [240]. See Response to Toni Fanfa’s Opposition to Petition to Determine Persons Entitled to Distribution of Estate at 5, In re Estate of Delakovias, No. HP09481108 (Cal. Super. Ct. June 17, 2011).

  241. [241]. See Notice of Settlement at 1–2, Estate of Delakovias, No. HP09481108 (Cal. Super. Ct. Oct. 11, 2011).

  242. [242]. I defined probate “litigation” narrowly to mean a petition that prompted an objection from an adverse party. I excluded quasi-adversarial matters in which a party needs to convince the judge to grant a particular remedy. For more on this topic, see Horton & Chandrasekher, supra note 154, at 157 n.290.

  243. [243]. See infra Table 1.

  244. [244]. An “850 petition” seeks to reclaim title to property held by someone else. See Cal. Prob. Code § 850 (West 2017). It can be filed by a personal representative who seeks to reclaim property on behalf of the estate, and it can be filed against an estate by “any interested person.” Id. § 850(a)(2).

  245. [245]. The ratio of will contests declined from 33 out of 910 cases (3.6%) before July 1, 2008 to 41 in 1,543 (2.6%) afterwards (although this change was not statistically significant). This drop-off in claims is consistent with the overall trend. Before July 1, 2008, there were 134 contested matters (14.7%); following that date, there were 174 disputed estates (11.3%). This is a statistically significant difference (p < 0.05).

  246. [246]. See supra notes 180–95 and accompanying text.

  247. [247]. See Declaration by Petitioner in Response to Prob. Exam’r’s Notes at 2–3, Estate of Campbell, No. HP10543410 (Cal. Super. Ct. Nov. 29, 2010); see also Declaration of Edwin A. Weatherton in Support of Petition for Prob. of Will at 1–2, Estate of Campbell, No. HP10543410 (Cal. Super. Ct. Nov. 3, 2010); Declaration of Elizabeth L. Nay in Support of Petition for Prob. of Will at 1–2, Estate of Campbell, No. HP10543410 (Cal. Super. Ct. Nov. 3, 2010); Declaration of Sarah A. Hayse in Support of Petition for Prob. of Will at 1–2, Estate of Campbell, No. HP10543410 (Cal. Super. Ct. Nov. 3, 2010).

  248. [248]. See Ex Parte Petition for Letters of Special Admin. at attachment 3e(2) p. 1, Estate of Burns, No. FP09469690 (Cal. Super. Ct. Aug. 20, 2009).

  249. [249]. See supra text accompanying notes 196–241.

  250. [250]. See supra text accompanying notes 1–32.

  251. [251]. See supra text accompanying notes 84–87.

  252. [252]. Compare Population Size and Growth, Austl. Bureau of Statistics,
    au/ausstats/abs@.nsf/Lookup/by%20Subject/1301.0~2012~Main%20Features~Population%20size%20and%20growth~47 (last updated Nov. 11, 2015) (between 1.2 million and
    1.3 million people in the 1970s), with Alameda County, Bay Area Census, http:// (last visited Mar. 29, 2018) (about 1.4 million people at the 2000 census).

  253. [253]. See Langbein, supra note 34, at 37 n.174 (“Grants of representation, mostly of testate estates, numbered 5368 for 1980, 5565 for 1981, 5537 for 1982, 5334 for 1983, and 5103 for 1984.”).

  254. [254]. My dataset consists of 556 estates that opened in 2007, 656 estates that opened in 2008, 630 estates that opened in 2009, and 611 estates that opened in 2010.

  255. [255]. See supra notes 168–69 and accompanying text.

  256. [256]. See Langbein, supra note 34, at 3 (explaining that Australian states do not recognize holographic wills).

  257. [257]. See supra note 85–86 and accompanying text.

  258. [258]. It was not until 2011 that a California appellate court held that harmless error applied retrospectively. See Estate of Stoker, 122 Cal. Rptr. 3d 529, 534–35 (Cal. Ct. App. 2011); cf. In re Estate of Smith, 651 N.W.2d 153, 157–58 (Mich. Ct. App. 2002) (holding that UPC’s version of the rule is retroactive).

  259. [259]. See, e.g., Petition for Prob.: Alameda Cty. Prob. Exam’r Checklist at 1, Estate of Carroll, No. RP09491388 (Cal. Super. Ct. Mar. 22, 2010) (noting that a purported codicil had been only signed by one witness); see also First Codicil to the Pour Over Will of Michael J. Carroll at 2, No. RP09491388 (Cal. Super. Ct. Dec. 29, 2009). Compare Contest & Grounds of Objection to Prob. of Purported Will & to Petition to Administer Estate at 2, Estate of Sabatini, No. RP08–410643 (Cal. Super. Ct. Dec. 2, 2008) (noting that a purported will had not been attested), with Verified Answer of Richard A. Sabatini & Russell R. Sabatini to Contest & Objection to Prob. of Purported Will & to Petition to Administer Estate at 1–3, No. RP08410643 (Cal. Super. Ct. Jan. 28, 2009) (defending this document but not invoking harmless error).

  260. [260]. Two of these cases were not eligible for harmless error because they preceded the state’s adoption of the rule. See Last Will & Testament of John B. Tillman at 2, Estate of Tillman, No. RP08387282 (Cal. Super. Ct. May 13, 2008); Original Will at 2, Estate of Ali, No. RP08377894 (Cal. Super. Ct. Mar. 21, 2008).

  261. [261]. Although both of these matters were filed after July 1, 2009, only one (Caspary, which I mentioned in the Introduction), involved a harmless error claim. Compare Contest & Grounds of Objection to Prob. of Purported Will & to Petition to Administer Estate at 2, Estate of Sabatini, No. RP08410643 (Cal. Super. Ct. Dec. 2, 2008) (noting that a purported will had not been attested), with Verified Answer of Richard A. Sabatini & Russell R. Sabatini to Contest & Objection to Prob. of Purported Will & to Petition to Administer Estate at 1–3, No. RP08410643 (Cal. Super. Ct. Jan. 28, 2009) (defending this document but not invoking harmless error).

  262. [262]. In addition, probate code section 6111 requires the “material provisions” of a holographic will to be in the testator’s handwriting. Cal. Prob. Code § 6111 (West 2017). As noted above, California’s harmless error rule may not apply to a document that flouts this command. See supra text accompanying notes 164–70. Perhaps for this reason, I did not find a single litigated case that featured this issue. For the rare example of a typewritten document that was (apparently) intended to be a holograph, see Holographic Last Will & Testement [sic] of Alex Ozeroff, dated June 19, 1991, Estate of Ozeroff, No. RP07311595 (Cal. Super. Ct. Apr. 19, 2007).

  263. [263]. See, e.g., Last Will & Testament at 1, Estate of Albach, No. HP07332024 (Cal. Super. Ct. Aug. 28, 2007) (handwritten document that may be notes for later attested will); Petition for Prob. of Will & for Letters Testamentary at attachment 4e(2), Estate of Sahlin, No. HP08367279 (Cal. Super. Ct. Jan. 23, 2008) (multiple will-like documents); Petition for Prob. of Will & for Letters of Admin. with Will Annexed at attachment 4e(2), Estate of Wit, No. RP07362504 (Cal. Super. Ct. Jan. 3, 2008) (possible will written in diary).

  264. [264]. See In re Estate of Bernard, 23 Cal. Rptr. 828, 830 (Cal. Dist. Ct. App. 1962).

  265. [265]. See In re Estate of Johnson, 60 P.3d 1014, 1016–17 (Mont. 2002).

  266. [266]. See id.

  267. [267]. See supra notes 135–39 and accompanying text.

  268. [268]. See, e.g., Petition for Prob.: Alameda Cty. Prob. Exam’r Checksheet at 1, Estate of Skye, No. HP09483736 (Cal. Super. Ct. Jan. 12, 2009); Memorandum of Points & Auths. in Support of Successor Tr. Anthony Hampton’s Motion for Judgment on the Pleadings at 2, Estate of Ferguson, No. RP08387690 (Cal. Super. Ct. Aug. 5, 2008) (mentioning a dispute over an unsigned will); Supplemental Petition for Prob. of Will & for Letters Testamentary at attachment 4e(2), Estate of Ferguson, No. RP08387690 (Cal. Super. Ct. May 16, 2008) (attaching unsigned and unattested will); Purported Will of Eli A. Ramirez, Estate of Eli A. Ramirez, No. FP08368656 (Cal. Super. Ct. Apr. 14, 2008) (featuring handwritten but unsigned pages); Last Will & Testament of Helen G. Mitchell at 3–4, Estate of Mitchell, No. RP07332427 (Cal. Super. Ct. Jun. 25, 2007) (involving an unsigned and unattested will).

  269. [269]. Cal. Prob. Code § 6110(c)(1)(A) (West 2017).

  270. [270]. See supra note 132 and accompanying text.

  271. [271]. Petition for Prob.: Alameda Cty. Prob. Exam’r Checksheet at 1, Estate of Lou, No. RP10540631 (Cal. Super. Ct. Oct. 28, 2010).

  272. [272]. See Declaration of Sara R. Diamond Regarding Publication of Notice and a Discrepancy in Date of Will at 1–2, No. RP 10540631 (Cal. Super. Ct. Nov. 8, 2010).

  273. [273]. Unif. Probate Code § 2-503 (Unif. Law Comm’n 2010).

  274. [274]. See Estate of Nielson, 165 Cal. Rptr. 319, 323 (Cal. Ct. App. 1980).

  275. [275]. See, e.g., Last Will & Testament of Robert G. Lawton at 1, Estate of Lawton, No. RP09436180 (Cal. Super. Ct. Feb. 17, 2009); Will of Laura Rachael Allan at 3, Estate of Allan, No. RP08392542 (Cal. Super. Ct. June 13, 2008); Last Will & Testament at 1, Estate of Aure, No. RP08390904 (Cal. Super. Ct. June 4, 2008); Last Will & Testament of A. Danette Comer at 1–5, Estate of Comer, No. RP08380494 (Cal. Super. Ct. Apr. 7, 2008).

  276. [276]. See Last Will & Testament of Elizabeth Blakes Secrease at 1–3, Estate of Secrease, No. RP10537319 (Cal. Super. Ct. Sept. 20, 2010); Last Will & Testament of Lorraine Alice Gomes Fleming at 1, Estate of Fleming, No. FP08403243 (Cal. Super. Ct. Aug. 11, 2008).

  277. [277]. See Order for Prob. at 1, Estate of Secrease, No. RP10537319 (Cal. Super. Ct. Oct. 25, 2010); Order for Prob. at 1, Estate of Fleming, No. FP08403243 (Cal. Super. Ct. Jan. 30, 2009).

  278. [278]. Of course, this cuts two ways. First, it suggests that harmless error may not generate litigation. Second, it implies that, to some degree, the statute is a solution in search of a problem.

  279. [279]. See Petition for Prob.: Alameda Cty. Prob. Exam’r Checksheet at 1, Estate of Tillman, No. RP08387282 (Cal. Super. Ct. Jun. 24, 2008); Declaration of Notary Public Regarding Testator’s Signature at 1–2, Estate of Schwind, No. RP07362658 (Cal. Super. Ct. Dec. 21, 2007).

  280. [280]. Although the UPC allows testators to use notaries in lieu of witnesses, California has not adopted this rule. See Unif. Probate Code § 2-502(a)(3)(B) (Unif. Law Comm’n 2010).

  281. [281]. See supra text accompanying notes 180–82.

  282. [282]. For a case with roughly similar facts that comes out the other way, see Estate of Hand, 73 N.E.3d 880, 884–88 (Ohio Ct. App. 2016) (refusing to enforce a “love letter” that concludes with a will-like passage on the grounds that it was not intended to be a will).

  283. [283]. See supra text accompanying notes 186–95.

  284. [284]. Ironically, the three completely unattested documents seem much more likely to be legitimate than the two writings that had one witness’s signature.

  285. [285]. See supra text accompanying notes 97–101, 125–30, 144–51.

  286. [286]. Cal. Prob. Code § 6110(c)(2) (West 2017) (emphasis added).

  287. [287]. See supra notes 1–32 and accompanying text.

  288. [288]. See supra notes 12–15 and accompanying text.

  289. [289]. See supra notes 11–16 and accompanying text.

  290. [290]. See Declaration of Paula S. Fass in Support of Petition for Prob. of Will of Gerard E. Caspary at 1–2, Estate of Caspary, No. RP08396884 (Cal. Super. Ct. Mar. 12, 2009).

  291. [291]. See supra text accompanying notes 97–101.

  292. [292]. See Petition for Letters of Special Admin. at attachment 3e(2) p. 24, Estate of Burns, No. FP09469690 (Cal. Super. Ct. Aug. 20, 2009).

  293. [293]. Id.

  294. [294]. Declaration of Stacie Delakovias in Response to Toni Fanfa’s Opposition to Petition to Determine Persons Entitled to Distribution of Estate at 2, Estate of Delakovias, No. HP09481108 (Cal. Super. Ct. June 17, 2011).

  295. [295]. Langbein, supra note 34, at 54.


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

Thanks to John H. Langbein for helpful comments, and to Thomas P. Gallanis and the editors of the Iowa Law Review for inviting me to participate in this symposium.