Down the Rabbit Hole: A Review of Errol Morris’s A Wilderness of Error: The Trials of Jeffrey MacDonald

Posted in: Book Reviews

Errol Morris is to be commended for his willingness to subject the Jeffrey MacDonald case to his customary rigorous examination.  MacDonald was convicted of murdering his wife and two young daughters in February 1970 while he was an Army doctor, and he is still in prison today.  MacDonald consistently has denied committing the crimes, saying that his family was attacked by four “hippie” intruders who assaulted and stabbed MacDonald while killing the rest of his family.

After its investigation, the Army decided not to charge MacDonald with the crimes.  Ultimately, however, the United States Department of Justice indicted MacDonald, and a jury convicted him after a trial in 1979.  Although the conviction was set aside by the United States Court of Appeals for the Fourth Circuit in 1980, that decision was then reversed by the Supreme Court two years later.  Despite a number of subsequent appeals, MacDonald has been in prison since 1982.

The horrific nature of the crimes, and the controversies attendant to the investigations into them as well as the conduct of the federal criminal trial in 1979, have together generated extensive attention and commentary.  Thus, by the time Morris wrote A Wilderness of Error: The Trials of Jeffrey MacDonald, (Penguin Press HC, 2012),the case had been the subject of a number of books, including Joe McGinniss’s infamous Fatal Vision (tainted by the author’s apparent betrayal of his subject) and Janet Malcolm’s The Journalist and the Murderer (marred by the author’s tendentious statements on the relationship between journalists and their subjects).

A Wilderness of Error is an extremely thorough account of the terrible murders that occurred in the MacDonalds’ home in Fort Bragg, North Carolina and the extensive proceedings that followed.  It is a beguiling book that presents forceful arguments to question whether MacDonald received a fair trial.

However, Morris ventures a step too far when he contends that MacDonald did not commit the crimes and therefore was wrongly convicted.  In reaching this conclusion, I have relied on, among other things, Richard Cahn’s review of A Wilderness of Error in the Touro Law Review.  (Cahn, a Long Island attorney, represented the parents of MacDonald’s wife and was involved in legal proceedings leading up to the federal criminal trial.)

The Murders

Part of the enduring fascination with the MacDonald case has to do with the character of the principal players.  The MacDonalds seemed to be, forgive the cliché, an All-American family.  As Morris recounts, Jeffrey MacDonald attended Princeton, left after three years for Northwestern Medical School, entered the Army in 1969, and was assigned to the Special Forces as a group surgeon.  In 1963, when he was a sophomore in college, MacDonald married Colette Stevenson, a sophomore at Skidmore College.  (Jeffrey and Colette had dated in high school.)  The couple’s first daughter, Kimberley, was born in 1964, and their second daughter, Kristen, was born three years later.

The family moved to Fort Bragg, North Carolina, in the summer of 1969, when MacDonald was assigned to an airborne Special Forces unit.  On the evening of February 16, MacDonald put his children to bed—he had worked a 24-hour shift at a local hospital through that morning—while his wife attended an evening class.  At the time, she was pregnant with what would have been the family’s first son.

Around 3:30 a.m. on February 17, MacDonald made a telephone call for help.  The military police arrived within the next 20 minutes or so to find MacDonald injured and his wife and children dead.  MacDonald was taken to the hospital and treated “for multiple bruises, an abrasion, small punctures, two stab wounds (one in his stomach, and one in his chest) and a collapsed lung,” according to Morris.

The Army’s Initial Investigation

As Morris recounts, “the crime scene was horrific” with “three bloody and battered bodies.”  The living room showed some signs of disturbance and the Army investigators found three weapons—a wooden club, a paring knife, and an ice pick—outside the back door of the house.  And, “[o]n the headboard in the master bedroom, the word ‘PIG’ was written in blood, recalling—perhaps reenacting—the Manson family murders committed only months before.”

MacDonald was interviewed by Army investigators less than two months after the murders.  He said that he was attacked by “three men” and that he also “got a glance of this girl with a kind of a light on her face,” perhaps “a flashlight or a candle”; he also saw the “top” of the girl’s boots that he thought were “brown” and described “a floppy hat” worn by the girl with “long, stringy, blond hair.”  (Over time, this testimony has become central to MacDonald’s claim of innocence.)  To this day, MacDonald has insisted that the murders in his house were committed by these four intruders.

The Army investigators thought otherwise, however.  Among other things, they observed that the living room was too neat, given the attacks that McDonald said had occurred and the struggle MacDonald had described.  They believed that MacDonald committed the murders himself, and staged a crime scene to correspond with his account.  As part of this scenario, they postulated that MacDonald stabbed himself a number of times, puncturing his own lung.

The Army presented its charges against MacDonald at a preliminary hearing, over the course of several months, in 1970.  During the hearing, there was testimony by a military policeman responding to the emergency call that he saw a woman with a “wide-brimmed hat” standing on the street, and additional testimony by an area resident about a woman named Helena Stoeckley, who was said to “need[] an alibi” for the night of the murders and who was known to have a blond wig and to wear a “big, old white floppy hat.”

In October, the commanding officer of Fort Bragg dismissed the charges against MacDonald on the grounds of “insufficient evidence.”  Subsequently, MacDonald was discharged from the service and returned to New York.  In 1971, he moved to California, where he worked as an emergency room doctor.

The Criminal Trial in United States District Court

Colette’s parents, Mildred and Freddy Kassab, initially condemned the Army for treating their son-in-law as the prime suspect in the murders.  Over time, however, the Kassabs became convinced that MacDonald was the perpetrator.  In a telephone call with Freddy in November 1970, MacDonald claimed that he and some “Special Forces buddies” found and killed one of the perpetrators.  (Almost a year later, MacDonald admitted to his father-in-law that he had lied about this incident.)

The next month, December 1970, MacDonald appeared on “The Dick Cavett Show” and criticized the Army for its incompetent investigation.  His manner offended his father-in-law, Freddy Kassab; how, Freddy thought, could MacDonald “go on television and almost say nothing about his family and what was done to them[?]”  That same month, Freddy obtained a transcript of the Army hearing.  After reading it, the Kassabs questioned MacDonald’s account of the night of the murders.

Freddy then mounted a campaign to have MacDonald prosecuted for the murder; as part of this campaign, the Kassabs’ attorney prepared a citizen’s criminal complaint against MacDonald and presented it to the Chief Judge of the United States District Court for the Eastern District of North Carolina in April 1974.  This pressure contributed to the government’s decision to convene a grand jury, which indicted MacDonald on three counts of murder in January 1975.

The federal criminal trial did not begin until July 1979.  The government’s case emphasized the physical evidence supporting its theory of MacDonald’s guilt.  The defense, in turn, focused on MacDonald’s character and accomplishments, contending that he did not have a motive to kill his wife and daughters.

The trial lasted more than a month.  A day after closing arguments, the jury convicted MacDonald of two counts of second-degree murder and one count of first-degree murder.  As noted earlier, after initially succeeding on the claim that his right to a speedy trial was violated in the Fourth Circuit, MacDonald has been in prison since the Supreme Court overturned the Fourth Circuit’s decision in 1982.

Did MacDonald Receive a Fair Trial?

Morris is effective in raising questions about whether MacDonald received a fair trial in 1979.  He is critical, and even contemptuous, of Judge Franklin Dupree’s management of the case.  Morris presents statements from a number of interested parties to the effect that the judge was convinced of MacDonald’s guilt and guided the trial accordingly.  Cumulatively, these criticisms have some force.

Much of Morris’s discussion focuses on the trial court judge’s decision not to allow testimony by witnesses who had heard Helena Stoekley talk about being in the MacDonald’s house the night of the murders.  If Stoekley actually was in the house that night, MacDonald’s account of a home invasion was true; or, just as significantly, if the jury believed that Stoekley was in the house that night, then there would be reasonable doubt with respect to MacDonald’s guilt.

To understand this point, context must be provided:  Stoekley was involved in the drug subculture in Fayetteville, North Carolina—as both a user of powerful drugs such as heroin and LSD and a police informant.  On the night of the murders, she was on mescaline.  After that, Stoekley made a number of statements over the years suggesting that she had been present in MacDonald’s house the night of the murders.  The essence of these statements was that Stoekley truly believed that she was in MacDonald’s house that night.  The statements were tantalizing, offering a few pertinent details but not more.

Although she was difficult to track down, Stoekley was found during the trial and interviewed by attorneys for both sides prior to testifying in court.  Stoekley then testified before the jury—and said that she did not participate in the murders.  She also denied having any recollection of discussing subjects related to the murder with the defense attorneys.  Her testimony thus did not support MacDonald’s case, although it did not seem that credible in its entirety.

MacDonald’s attorneys then sought to present the testimony of a number of other persons with knowledge of Stoekley’s prior statements indicating her involvement in the murders.  The trial court judge listened to this testimony outside the jury’s presence and then ruled that it was inadmissible.  In Judge Dupree’s view, the testimony, among other things, lacked “trustworthiness.”

This discretionary decision was within the judge’s province as the trier of fact, and the Fourth Circuit did not disturb that decision.  That decision did, however, prompt one of the Fourth Circuit judges—the Hon. Francis D. Murnaghan—to state:

. . . [T]his case provokes a strong uneasiness in me.  The crimes were base and horrid, and whoever committed them richly deserves severe punishment. As [Fourth Circuit] Judge Bryan has pointed out [in his majority decision], the evidence was sufficient to sustain the findings of guilt beyond a reasonable doubt. Still, the way in which a finding of guilt is reached is, in our enduring system of law, at least as important as the finding of guilt itself. I believe MacDonald would have had a fairer trial if the Stoeckley related testimony had been admitted.

Morris also criticizes the prosecution for its restrictive approach to providing the defense access to the physical evidence in the case.  He cites a research memo by a government law clerk outlining the prosecutor’s duty to disclose exculpatory material under Brady v. Maryland.  The memo, however, is not as sinister as Morris suggests.  After the trial, however, the memo raised eyebrows in light of a number of issues that were raised in the memo with respect to the quality and integrity of the physical evidence used to convict MacDonald.  (It must be noted that before the trial, according to the defense’s forensic expert, MacDonald’s attorneys did not press the government very hard for greater access to the physical evidence, putting the defense at a great disadvantage on appeal.)

Did MacDonald in Fact Commit the Murders?

Since the murders occurred in 1970, investigators, judges, and jurors have been required to decide which story to believe:  the government’s account, which attributes the murders to MacDonald; or MacDonald’s own account, which assigns blame to four drug-seeking intruders.  Twelve jurors found MacDonald guilty, and a number of judges have affirmed that verdict.

The guilty verdict is due, in large part, to the physical evidence supporting the government’s theory of the case and undermining Morris’s contention that MacDonald “was railroaded.”

As Cahn summarizes in his review:

Morris ignores facts from the crime scene that most disinterested observers (or jurors) would likely consider conclusive evidence against the “intruder” theory. Why on earth would a group of murderous strangers take five-year-old Kimberly’s bloody body and move it from the master bedroom (where everyone agrees she was killed) to her own bedroom, carefully tucking her into bed?  How is it consistent with MacDonald’s story (he was beaten and stabbed on the couch in the living room and his pajama top was torn while he was trying to defend himself) that the detached pocket of that pajama was found near Collette’s body on the floor of the master bedroom?  How does MacDonald (or Morris) explain why the pajama top was soaked with Collette’s blood before it was torn, and not afterwards, when MacDonald claims he placed his ripped garment on his wife’s bloody chest in an attempt to keep her from going into shock?  Why are there no marks on the ceiling or walls of the living room, where MacDonald claimed he was clubbed (by a man swinging a wooden club over his head) and stabbed, or, indeed, any signs of a deadly struggle having occurred in the living room?

The jury was persuaded by the government’s case.  Yet we never hear from a juror in Morris’s book.  One juror, Fred Thornhill, told Gene Weingarten of The Washington Post, “We heard the prosecution’s case, which was pretty credible, and then we waited for the defense to blow it out of the water.  They never came close.  They couldn’t refute the physical evidence.”  (Links to Cahn’s review and Weingarten’s magazine article are provided below.)

What almost certainly is MacDonald’s last appeal to set aside his conviction is now pending before a federal district court judge in North Carolina.  Unfortunately for MacDonald, his case is now a zero-sum game.  To prevail on this appeal, MacDonald must essentially show that he is innocent.  But, as has been true for more than four decades, the case for MacDonald’s innocence is problematic—even more problematic than the case for his guilt, even with the errors that may have occurred along the way.  Morris’s book raises provocative questions but the most likely answer still seems to be that MacDonald did murder his wife and daughters.

Richard Cahn’s review of A Wilderness of Error is available at  Gene Weingarten’s article for The Washington Post Magazine is available at

One response to “Down the Rabbit Hole: A Review of Errol Morris’s A Wilderness of Error: The Trials of Jeffrey MacDonald

  1. Dan Fletcher says:

    There is only one reason for MacDonald to move the body of his oldest daughter to her bedroom from the master bedroom and only one reason for him to lie about where she was killed. He was molesting his oldest daughter and was surprised by his wife.