Doris Duke's Estate and the Weight of a Testator's Choice
- 10 minutes ago
- 3 min read
Doris Duke died in October 1993 leaving an estate worth more than a billion dollars, and the years that followed produced one of the most closely watched fiduciary disputes in modern New York practice. What makes it useful as a planning example is not its size but a single decision the will made. Duke named, as the one individual to serve as executor of that fortune, the man who had worked as her butler. Nearly everything after that traces back to the choice, and to a question every client faces: who is equipped to carry the plan out?
Most Estate Tales begin with a magazine account. This one begins with the record itself. In Matter of Duke, decided January 11, 1996, the New York Court of Appeals reviewed the summary removal of Duke's preliminary co-executors and reversed it. The opinion rewards reading in the original: a court's own account of a contested estate is disciplined in a way press coverage is not. It separates what was alleged from what was proved, and describes what a court must have before displacing the person a testator chose. Cornell Law School's Legal Information Institute publishes it free.
Duke, heiress to the tobacco fortune built by her father, James Buchanan Duke, died at her Beverly Hills home at 80, and her will directed the bulk of her estate to charity. It also named Bernard Lafferty — described in the opinion as her assistant and confidant, and her butler since 1986 — as the lone individual co-executor, with substantial personal benefits. Lafferty selected United States Trust Company as corporate co-executor. Duke died in California, but the estate ran through Surrogate's Court, New York County, where preliminary letters issued on November 1, 1993.
Within little more than a year, interested parties moved to remove the co-executors. The allegations were serious and wide-ranging: commingling of estate and personal assets, waste of estate property, improvidence, and substance abuse, with separate objections aimed at the bank's oversight and at a loan it had made. The Surrogate appointed an investigator, Richard Kuh, who interviewed more than fifty people. On May 22, 1995, relying on his report, she removed both co-executors without a hearing. A divided Appellate Division affirmed.
The Court of Appeals reversed, on procedural rather than exculpatory grounds. The report contained no sworn statements and its sources were largely undisclosed, so it gave the Surrogate no record for her own findings, and the co-executors had filed affidavits disputing the central facts. Removal of a fiduciary, the Court observed, is “a judicial nullification of the testator's choice,” available only where statutory grounds are clearly established; summary removal on untested hearsay was an abuse of discretion. The Court took no position on whether the allegations were true; it remitted the case for proceedings that could answer that. They never happened. A settlement in May 1996 had Lafferty resign while keeping reduced benefits, and he died that November at 51, never charged with a crime.
For estate planners, the most important point is not the scale of the fortune. It is the distance between trusting someone and equipping them. Duke appears to have trusted Lafferty entirely, and nothing in the record establishes she was wrong to. But a fiduciary's work is technical: marshaling assets, filing accountings, managing conflicts, keeping estate money rigorously apart from personal money, and withstanding scrutiny from beneficiaries, co-fiduciaries and — where charity is involved — the Attorney General. Loyalty does not supply those skills, and naming someone who lacks them exposes that person as much as the estate. The case cuts both ways: fiduciaries are owed real procedural protection even when accusations are grave, and securing it took years of litigation the estate paid for.
Few families face a billion-dollar accounting, but the problem is ordinary. A parent names the loyal child, the nearby neighbor, the longtime bookkeeper, or the caregiver who was there at the end, and the relatives who were not there read that choice as evidence of something. Suspicion follows proximity. A fiduciary who commingles funds even innocently, who cannot produce records, or who benefits personally under the instrument he administers will spend years explaining himself. The estate pays for that, and so does every relationship inside the family.
A well-built plan anticipates this. It names a fiduciary equipped for the work, and a successor if that person cannot serve. Where it helps, it pairs someone who knows the family with a professional who knows the mechanics, and says who decides when they disagree. It addresses compensation and any bequest to the fiduciary openly, so the arrangement reads as documented intention rather than discovery. It leaves records and instructions clear enough that following them requires no guessing. None of that guarantees peace, but if the plan is questioned there is something solid to answer with.
Estate Planning Lesson: Choose a fiduciary for what the job will actually demand, not only for how much you trust the person.
