Law360 Canada (August 31, 2026, 8:43 AM EDT) --

John L. Hill

Francis Clifford Smith was sentenced to death in Connecticut in 1950. He was 25 years old.

He was granted a last meal eight times. Each time, his execution was stayed. Smith died in prison on June 25, 2026, at the age of 101, having spent more than 70 years in prison and, by some accounts, becoming the longest-serving prisoner in the United States.

Smith’s extraordinary story is not simply about capital punishment. It is a cautionary tale about the fallibility of criminal convictions and about what happens when doubts about guilt arise after the machinery of justice has already delivered its verdict.

Smith was convicted of first-degree murder for the killing of Grover Hart, a nightwatchman at a Connecticut yacht club. Smith had been arrested with another man in connection with the killing. The other man accepted a plea agreement and testified against Smith.

That evidence ultimately sent Smith to death row, but the conviction never sat well. Over the years, witnesses recanted their statements. Another inmate later confessed to the murder. And Major Leo Carroll, a police officer who had interrogated Smith after his arrest, eventually told the state Board of Pardons that he was not even sure Smith had been present at the time of the murder.

Those doubts repeatedly interrupted the state’s attempts to carry out his execution. In 1954, Smith’s death sentence was commuted to life imprisonment.

He remained behind bars for more than seven decades, though his record was not without complications. He escaped from prison in 1967 and was recaptured 12 days later. He was temporarily released for Christmas in 1974. In 1975, he was granted parole but returned to custody after committing petty offences that violated the conditions of his release. Thereafter, Smith reportedly resisted efforts to restore him to parole until 2020, when he was finally released under supervision to a care facility for elderly people within the justice system.

There was an incongruous gentleness to the final decades of a life that had begun with a violent accusation. At Osborn prison, Smith became known as the “Birdman of Osborn” for feeding birds outside the prison. According to a Connecticut corrections official, Smith would stuff bread into his clothing to feed them. Guards, aware of what he was doing, generally looked the other way.

The image is difficult to reconcile with the young man sentenced to death in 1950, but that may be precisely the point.

Canada abolished the death penalty for murder in 1976. The last executions in Canada occurred in 1962.

So why should Canadian lawyers care about a 101-year-old American prisoner whose death sentence was commuted more than seven decades ago?

Because the most important lesson of Smith’s case has little to do with capital punishment; it concerns wrongful conviction.

Canadian courts have long recognized that the criminal justice system is not infallible. The problem is that the law’s mechanisms for correcting a wrongful conviction are inevitably more complex than those for producing a conviction in the first place.

Once a verdict is entered, the presumption of innocence disappears. Rules, deadlines and standards of review govern appeals. Fresh evidence must satisfy legal tests. As the years pass, witnesses die, memories fade and physical evidence can disappear.

Smith’s case illustrates the nightmare scenario: evidence supporting the conviction gradually unravels. At the same time, the legal system must determine whether the unravelling is sufficient to justify overturning a judgment that has already stood for years.

Canada has had its own painful experience with this phenomenon. The names are familiar: David Milgaard, Donald Marshall Jr., Guy Paul Morin, Thomas Sophonow and others whose convictions were later found to be unreliable or wrongful. Their cases demonstrate that wrongful conviction is not an American problem or a relic of an earlier era. It is a problem inherent in any system in which human beings investigate crimes, assess evidence and decide questions of guilt.

And the Smith case presents another uncomfortable question: how much uncertainty is enough?

A criminal trial does not require absolute certainty. It requires proof beyond a reasonable doubt. But that standard applies only at trial. Once a conviction has been entered, the legal system does not simply retry the case whenever someone raises a new question about the evidence. That distinction is necessary for finality, but it can also create profound difficulty when the original verdict was wrong.

One particularly interesting feature of Smith’s case is the later recantation of witness statements. Recantation evidence poses a notorious problem for appellate courts. A witness may have lied initially. A witness may be telling the truth later. Or the witness may have lied both times. The mere fact that a witness changes their story does not establish that the original evidence was false.

But a recantation should not be dismissed automatically simply because it is inconvenient to the verdict. That is where the courts face one of the most difficult tasks in criminal law: determining whether new information genuinely undermines the reliability of the conviction or merely adds another piece of contradictory evidence.

Smith’s case became particularly troubling because the doubts did not stem from a single source. They reportedly included recanting witnesses, a subsequent confession by another prisoner, and a police officer’s own doubts about whether Smith had even been present. The cumulative effect is considerably more significant than any single piece of evidence considered in isolation.

An important institutional difference separates Smith’s experience from the modern Canadian system. Canada now has a dedicated mechanism for investigating claims of wrongful conviction. The federal Miscarriage of Justice Review Commission was established under Bill C-40 to replace the previous ministerial review process.

That development reflects a simple recognition: ordinary appeals are not always enough.

A person who has exhausted the normal appellate process may still possess a credible claim that a miscarriage of justice occurred. A specialized review mechanism provides another avenue for investigating such claims.

But the Smith story remains relevant because no procedural mechanism can eliminate human error.

Investigators can make mistakes. Witnesses can be mistaken or dishonest. Police can become committed to a theory of the case. Defence lawyers can miss evidence. Prosecutors can misunderstand the significance of evidence. Experts can be wrong. Juries can reach the wrong conclusion. And judges, despite their experience and training, are not immune from error.

The most striking feature of Smith’s story may ultimately be the passage of time. Smith was 25 when he was convicted. He was 101 when he died. Between those two ages lies virtually an entire human lifetime.

Even if one sets aside the question of whether Smith was actually innocent, a question his case never definitively resolved, the possibility that an innocent man spent seven decades imprisoned should trouble anyone who works in criminal justice.

Capital punishment makes the stakes particularly obvious. If the state executes an innocent person, the error cannot be corrected. But imprisonment also has consequences that cannot be restored, as I have learned practising prison law. Seventy years taken from a person’s life cannot be returned by an acquittal, a pardon or a financial settlement.

That is why the law’s commitment to finality must always be balanced against its commitment to accuracy.

The story of Francis Clifford Smith offers Canadian lawyers a useful reminder of that balance. A conviction is a legal conclusion. It is not a guarantee of factual truth. And sometimes, the most important question a justice system can ask is not whether a verdict was properly entered, but whether it is still safe to believe it.

Smith was repeatedly prepared for execution and repeatedly spared. In the end, he was not put to death. But the eight last meals remain a grim symbol of something larger: the extraordinary consequences that can follow when a criminal justice system becomes certain of a person’s guilt before the evidence has earned that certainty.

For Canadian criminal lawyers, that is a lesson worth remembering, even without the death penalty.

John L. Hill practised and taught prison law until his retirement. He holds a JD from Queen’s and an LLM in constitutional law from Osgoode Hall. He is also the author of Pine Box Parole: Terry Fitzsimmons and the Quest to End Solitary Confinement (Durvile & UpRoute Books) and The Rest of the (True Crime) Story (AOS Publishing). Contact him at johnlornehill@hotmail.com.

The opinions expressed are those of the author(s) and do not necessarily reflect the views of the author’s firm, its clients, Law360 Canada, LexisNexis Canada or any of its or their respective affiliates. This article is for general information purposes and is not intended to be and should not be taken as legal advice.

Interested in writing for us? To learn more about how you can add your voice to Law360 Canada, contact Analysis Editor Peter Carter at peter.carter@lexisnexis.ca or call 647-776-6740.