Part 2 – Manufacturing Guilt: Wrongful Convictions

I believe [the accused] did something very, very wrong in that room. But I also believe that nobody has asked me to play God. I’ve been asked to apply the law. Justice belongs to God; men only have the law. Justice is perfect, but the law can only be careful. […]

If we as a jury wanted to understand why the burden of proof fell on the prosecution, and fell with such gravity, we needed only to reflect on what we had discovered directly about the real power of the state and its agents: there was, in a deep way, no recourse. Yes, there were appeal courts, constitutional protections, citizen juries like us. But in the end – in the end, there was, simply, the final power of the state. There was always this. This was a power even more terrifying, in a way, than a man with a knife in a closed room. That sort of raw, physical power, for all its horrors, can never extend indefinitely in all directions. If you were to run outside, people would object, would (in principle) come to your aid. In the room, you know this, even if you cannot actually escape. But there is nowhere to run from the state: more or less everyone outside the room (the courtroom, the prison) has already accepted the legitimacy of what the state chooses to do to you. In fact, all those people out there, they actually constitute the state itself. If you run out to them, they will help catch you. There is nowhere to go.

A Trial by Jury: Burnett, D. Graham: 9780375727511: Books –, at pp. 139 and 163.

(Part 1


In this excellent book (Duped: Why Innocent People Confess – and Why We Believe Their Confessions: Kassin Ph.D, Saul: 9781633888081: Books –, social psychologist Saul Kassin deals extensively with the psychology of false confessions. He discusses many cases of false confessions, such as the Central Park Five, in which all five innocent youth confessed falsely.

Discusses the Kitty Genovese story (1964 Queens, NY) in extensive detail (chapter 2). The case is known popularly for the “bystander effect,” given that many people ignored a vicious rape/murder they could hear – although much of that has been questioned (see also). Winston Moseley was arrested, confessed (knowing extensive detail that no one else knew), and was eventually convicted (surprisingly not executed) and died of old age in jail decades later. He also confessed to a Kralik and Johnson murder. He was never prosecuted for the Johnson murder, while someone else was wrongfully convicted for the Kralik murder. Police refused to take Moseley’s confession seriously – particularly since they already had a confession from young Alvin Mitchell (after a 50-hour interrogation) for the Kralik murder (leading to the eventual wrongful conviction of Mitchell). One of the reasons for the Mitchell wrongful conviction was the trial judge (at the second trial – the first resulted in a hung jury) refusing to allow the victim’s dying statement (that she did not recognize her assailant, which would have excluded Mitchell but not Moseley) on technical hearsay grounds – gotta love evidence law!

One shocking quote about the refusal of the prosecutor to acknowledge its error (p. 140); a theme prominent in the following book, as well:

In a 2001 rape case in West Virginia, DNA testing not only excluded confessor Joseph Buffey but also matched a convicted sex offender. Yet rather than agree to vacate Buffey’s conviction, the prosecutor’s office proposed the new theory that Buffey and the sex offender were accomplices. This theory flatly contradicted the confession extracted from Buffey, which made no mention of an accomplice, and it contradicted the victim’s own lucid account of a harrowing and protracted ordeal involving a single perpetrator. (p. 140)

Another example cited is Juan Rivera, who was found guilty at a retrial – despite DNA evidence excluding him (p. 266).

Another terrible example is Billy Wayne Cope.

Another great quote (p. ix):

I’ve seen judges unwilling to serve as the gatekeepers they get paid to be by failing to exclude from evidence confessions that were quite clearly coerced, not voluntary.

I’ve seen prosecutors bend logic until it breaks, refusing to concede innocence despite DNA results that excluded the confessor, without doubt, and identified the actual perpetrator.

I’ve read contemporaneous newspaper accounts during trial that accepted the government’s public relations spin on the case hook, line and sinker.

Convicting the Innocent

In this book (Convicting the Innocent: Where Criminal Prosecutions Go Wrong: Garrett, Brandon L.: 9780674066113: Books –, the author, Duke law professor Brandon L. Garrett, digs deep into many U.S. wrongful convictions (250 known DNA exonerations, obtaining trial transcripts of some 88% of them) and finds some appalling results. The most common exoneration was for rape (68% – for which the conviction rate apparently isn’t high enough), and another 21% convicted of both murder and rape (p. 5). A full 16% confessed to crimes they did not commit, with all but two of those confessing to details they could not have known without police “coaching” (p. 6). Eyewitness misidentification comprised 76% of the cases. “Invalid forensics” were found in 93 (of 153) analysts called by the prosecution. Informants were present in 21% of the cases. 17 were sentenced to death. More than 90% of the exonerees for whom transcripts were obtained asserted their innocence at trial. The majority of convictions went through multiple rounds of appeals untouched (p. 11). It took an average of 15 years for each exoneration. Judges and prosecutors sometimes opposed DNA testing. Most appallingly, judges (and prosecutors) sometimes initially refused to exonerate these people even after DNA proved their innocence (p. 11).

In one memorable case (David Gray), the prosecutor couldn’t remember when testifying whether he may have said something like “it won’t hurt you” when inducing a jailhouse informant to perjure himself (p. 122).

I got a kick out of a judge’s comments, during the sentencing of Michael Evans and Paul Terry, that the defendants walked with a swagger that suggested no remorse. Their lawyer bravely retorted that it is also clearly indicative of innocence (p. 162).

In the Rolando Cruz case, a few jurors later told the press that “half of the jurors had their minds made up before the trial even started” (p. 173).

At Arvin McGee’s third trial, a potential rogue juror (i.e. not prepared to convict) was extensively pressured by the other jurors to join the consensus for a guilty verdict. The defence’s mistrial application on this basis was subsequently denied. It took another 14 years for his DNA exoneration (p. 173).

In the case of Kennedy Brewer, the state refused to conduct DNA testing, lying that the sample was insufficient for testing. Many years later, post-conviction (for a brutal murder-rape of his own toddler daughter), Brewer managed to get the state to test the samples. He was excluded, so his conviction was vacated. However, the state kept him in limbo for the next 8 years, deciding whether or not to re-try him (p. 180).

Curtis McCarty had his conviction reversed twice (14 years apart) on appeal grounds, based on issues with the forensic evidence (the second time on fresh evidence about the misconduct of the expert, somehow discovered by his lawyers). He served 21 years in prison (p. 181).

In 10% of the cases that had written appeal decisions, judges had called the state’s case against the innocent and wrongfully accused “overwhelming” (p. 202).

In the late 1980s, Earl Washington Jr. had come “within nine days of his execution” (p. 219). Then, after 9 years on death row (in 1993), DNA testing practically excluded him – but not conclusively. As a result, his death sentence was commuted, but he remained in jail to serve a life sentence. He was only exonerated and released in 2001, after – get this – his lawyers discovered that the state had withheld a 1993 DNA test excluding him. A later independent audit sharply criticized the Virginia crime laboratory for producing false results and succumbing to political pressure to not exonerate Washington (p. 221).

In at least 49% of the cases, the exonerees had to obtain a judge’s order to get the DNA testing – suggesting the state improperly opposed the application (p. 227). For at least 18 exonerees, judges initially denied their applications for DNA testing (p. 228). For example, in the Bruce Godschalk case, the judge refused to order testing, seemingly convinced (wishful thinking, I’d suggest) by the strength of the confession that there would be nothing to see. Same with Byron Halsey. In other cases, it was denied on technical, procedural grounds. I’m guessing the legislation in the U.S. is far more favourable than in Canada, by the way, which is even more reluctant to entertain the possibility of innocence.

Shockingly, judges “refused to set free at least twelve exonerees despite at least preliminary DNA test results excluding them” (p 230).

A particularly cruel example is Roy Criner, where Texas Appeals Court Judge Sharon Keller fabricated creative excuses to justify not letting him go, even though DNA evidence excluded him (p. 230).

Radley Balko

See also Radley Balko very recently here:

  • You might think that if DNA testing could establish the innocence of a death row prisoner, a state would want to allow it, both so it didn’t execute an innocent man, but also so it could pursue real justice for the killer’s victims. You’d be wrong, at least in Texas. And the Supreme Court will decide this term whether Texas can get away with it.

  • Meanwhile, the Texas Court of Criminal Appeals has refused to overturn a conviction and death sentence, despite the fact that both the trial judge and the DA for the district where the prisoner was tried both say he was wrongly convicted.

Or Radley Balko on Charlie Vaughn (The maddening irrelevance of Charlie Vaughn’s innocence (, emphasis added:

Arkansas provides no real way for prisoners like Vaughn to get back into state court, even when they have persuasive evidence of their innocence. This ought to be where the federal courts would intervene. But because of the Antiterrorism and Effective Death Penalty Act (AEDPA), a law signed five years after Vaughn’s 1991 conviction, the federal courts won’t even consider the evidence of his innocence. AEDPA gives prisoners like Vaughn one shot at federal review. The two other defendants never took their shot. Vaughn took his with that handwritten letter in 1995. So now he’s out of luck.

Believe it or not, this wrongful conviction inmate (self-represented, hand-written) appeal was dismissed in 1996 after a “careful and thorough review of the record” (chilling, probably false, and reminds me of similar language in R v NBM, 2021 ABCA 14: “Overall, when viewed with the benefit of the entire record, NM’s arguments in support of the first ground of appeal represent nothing more than a thinly veiled effort to convince the panel to reconsider the evidence and make different credibility assessments contrary to the standard of review and the role of this Court,” at para. 16; the word “record” is cited some 10 times there). He simply gets “one shot” and since he lost in 1995, he’s out of luck. Evidence of innocence be damned.

Manufacturing Convictions

In this 1998 book (Manufacturing guilt: Wrongful convictions in Canada: Anderson, Barrie: 9781895686937: Books –, sociologists Barrie Anderson and Dawn Anderson go through several known Canadian wrongful convictions. They raise the very interesting point of a judge’s innate conflict of interest (p. 15):

Judges therefore have a tremendous responsibility to maintain justice. However, they also have the very important task of maintaining public confidence in the judicial system and emphasizing its legitimacy. To this end, lawyer Alastair Logan (1995) asserts, the judiciary will act to preserve the reputations of police officers, prosecutors, expert witnesses or others acting on behalf of the Crown when their reputation or the system is called into question. The possibility therefore exists that a judge may unintentionally or maliciously conduct a trial or instruct the jury in a way that is prejudicial to the accused, if he or she perceives that to do otherwise would somehow jeopardize the integrity of the judicial system.

Further, commenting on prosecutorial misconduct (p. 16):

Unfortunately, the desire to win has caused many lawyers to engage in questionable, even unscrupulous, tactics, which are frequently condoned by the legal profession as a whole. Lawyers learn quickly what works well in the courtroom and what does not. What works are techniques that may distort the truth, confuse the jury and make apparent liars out of honest witnesses. The legal system is highly structured and those within it share a culture that emphasizes winning cases rather than doing justice. For too many lawyers the courtroom has become a place of winning cases rather than a forum for discovering truth and serving justice. When the need to win takes precedent over truth, the seeds of wrongful conviction have been sown in the fertile soil of legal indifference, personal greed and public apathy.

The first case dealt with in depth is Donald Marshall Jr., which I dealt with in Part 1 (the last is Steven Truscott, which I discuss extensively here). The next is David Milgaard. In that 1970 Saskatchewan case, the police pressured witnesses to perjure themselves and fabricate testimony about observing Milgaard commit a crime – similar to Marshall (and Truscott). Despite a key witness later recanting his testimony and the presence of a viable suspect (Larry Fisher, convicted in 2000), the federal government refused to admit its mistake for years.

In 1992, the Supreme Court of Canada insisted Milgaard had “a fair trial” yet granted him a new trial due to the new evidence. Eventually, Larry Fisher was charged (in 1997). The state eventually conceded its mistake and paid $10 million compensation in 1999. Milgaard was a mere 16 years old at the time of arrest/conviction (slightly older than Truscott was) He died earlier this year.

The 1992 SCC decision is troubling (paras. 3-4, emphasis added):

It is appropriate to begin by stating that in our view David Milgaard had the benefit of a fair trial in January 1970. We have not been presented with any probative evidence that the police acted improperly in the investigation of the robbery, sexual assault and murder of Gail Miller or in their interviews with any of the witnesses. Nor has evidence been presented that there was inadequate disclosure in accordance with the practice prevailing at the time. Milgaard was represented by able and experienced counsel. No error in law or procedure has been established. At the conclusion of the trial, there was ample evidence upon which the jury, which had been properly instructed, could return a verdict of guilty.

However, fresh evidence has been presented to us. Ronald Wilson, a key witness at the trial, has recanted part of his testimony. Additional evidence has been presented with respect to Milgaard’s alleged motel room confession [to friends]. More importantly, there was evidence led as to sexual assaults committed by Larry Fisher which came to light in October 1970, when Fisher made a confession.

Myopically, the SCC tried to make Ronald Wilson the “fall guy” for this wrongful conviction. They went so far as to charge him with contempt of court for his previous perjured testimony (note that contempt and perjury are almost never used against perjuring Crown witnesses, strangely enough). They later vacated the contempt citation without explanation. Would it have been that difficult for the SCC to concede that the state engaged in egregious misconduct from start to finish?

Milgaard Inquiry (2008) – available here. Incidentally, Milgaard always viewed the Inquiry’s findings as a whitewash – likely correctly.

Like Truscott, Milgaard did not testify. It is unclear whether his lawyers were also blamed for the wrongful conviction on this ground. Hopefully not.

Note that Milgaard’s 1971 appeal was dismissed (R. v. Milgaard, 1971 CanLII 792 (SK CA), with leave to appeal later predictably dismissed by the SCC. No apology forthcoming, I presume.

In 1980, Larry Fisher’s wife told police she thought her husband is guilty of this crime. Ten years later, Milgaard’s appeal counsel (the legendary Hersh Wolch) found out about this through an anonymous tip (timeline). Note the Inquiry begrudgingly admits “that the decision by Saskatoon Police not to follow up on the Linda Fisher report was not reasonable” (p. 13).

Wilbert Coffin is the next case described. It is an interesting case from rural Quebec in 1953. Three American hunters were murdered, and the government – under pressure to pin the murders on someone – settled on Coffin. Despite an incredibly weak Crown case, he was convicted in 1954, and hanged in 1956 (following his predictably unsuccessful appeals). A good illustration of how the prosecutor can manufacture a conviction (and subsequent execution) out of whole cloth. It later appeared that police suppressed proof of his innocence – a letter from one of the hunters dated after Coffin had left – p. 69. It also appears his defence lawyer (Raymond Maher) dropped the ball; promising the jury that he would call 100 witnesses to prove his client’s innocence, he called none.

Guy Morin is another disturbing example of a 1984 wrongful conviction. He was fingered by police because he appeared weird to them. The nature of the crime (a rape-murder) is the same as both Truscott and Milgaard, as well as Sophonow’s. The victim was Morin’s 9-year-old neighbour. The police claimed that an unrecorded part of their interview with Morin made him a key suspect, in their minds (p. 75). As with Truscott and Marshall, police refused to take other viable suspects seriously (note: Calvin Hoover, eventually matched as the perpetrator through DNA, is not even one of the 4 other viable suspects mentioned).

Notably, there was no time for Morin to have done the deed if the victim’s mom got home when she originally said she did. Magically, her timeline then changed by 20-30 minutes – reminding me of similarities with Truscott. Still, the timeline required that he did “all this driving, raping, killing, cleaning and shopping” in 75 minutes – a strong argument of the defence at trial (p. 82).

Eventually, the police extracted what they considered a “confession” from Morin – over a few days, he told an undercover officer that “no one would ever know the real relationship he had” with the victim (p. 81). AHA!

Also, two jailhouse informants claimed that Morin confessed to them. Of course, they were given an incentive by police to testify.

Clay Ruby, the lawyer at the first trial, insisted on a backup position that Morin was insane. The jury ultimately acquitted. Very unfortunately, the crown appealed successfully (R. v. Morin, 1988 CanLII 8 (SCC) – a slim 4-2 majority upholding the majority of the ONCA; note that in the US the State is never allowed to appeal an acquittal). A subsequent appeal based on delay was denied at ONCA, as well.

On the retrial, the judge (James Donnelly – the name rings a bell; his father, Frank Donnelly was likely Truscott’s defence lawyer) was openly and extremely hostile to defence counsel (a team led by the excellent and legendary Jack Pinkofsky). In the interim, the victim’s body was exhumed, and revealed much misconduct on the part of the forensic team (most of the results were exculpatory). There were substantial unsuccessful attempts to have the charges dismissed on the basis of police/Crown misconduct (withheld disclosure). Nevertheless, he was convicted the second time around. Courtroom observers “detected the warmth [Judge] Donnelly displayed towards the prosecution team and the repressed hostility he directed at Pinkofsky” (p. 88 – note his inauthentic obituary stating he never allowed bullying, etc.). Donnelly “encouraged the jury to endorse the most ludicrous testimony of some Crown witnesses while rejecting important defence evidence” (p. 88). In a glaring misstep identical to one taken by the trial judge in Truscott, the judge provided speculation to fill in a major gap in the Crown’s evidentiary case.

In 1993, Morin was surprisingly granted bail pending appeal – only the second person in Canada to do so after a conviction of first-degree murder (p. 89). Thankfully, less than 2 years later – right before the appeal was to be heard – DNA evidence excluded him.

See also Kirk Makin’s book Redrum the Innocent and Real Justice: Guilty of Being Weird: The story of Guy Paul Morin: Faryon, Cynthia J.: 9781459400924: Books –

In this context (following the then-recent Milgaard, Morin and Marshall exonerations), the SCC came out with R. v. Stinchcombe, 1995 CanLII 130 (SCC) – theoretically improving the State’s disclosure obligations to accused. Whether or not it affected anything in practice is a different story.

Later Kaufman Commission (1996) – available here.

Thomas Sophonow was wrongly convicted of murder/SA in 1981 Winnipeg. A teenage female employee was murdered, and Sophonow’s description was similar to the man police were lookin for. The witness they relied on was John Doerkeson, who appeared to know a lot about the incident (surprisingly never appeared to be a suspect). Did I mention Doerkeson positively “identified” Sophonow eventually? Sophonow had a strong alibi, and police largely refused to “investigate” this alibi defence. The Crown refused to accept it, in any event. At the first trial, there was a mistrial (hung jury). At the second trial, the Crown witnesses’ evidence surprisingly improved significantly. The Crown also produced multiple lying jailhouse snitches. The judge significantly hampered the defence with caustic comments about its (truthful) theory and witnesses (p. 101). The wrongful conviction was thus manufactured by the Crown and with a significant “assist” to the sworn-to-be-impartial judge.

Defence counsel (the famed Greg Brodsky) won a retrial on appeal with little trouble, thankfully. On the third trial, the Crown Stuart Whitley engaged in the ridiculous tactic of trying on “the glove” himself and claiming that it should fit Sophonow if it fit him, as they were of similar size and build (it had previously not fit Sophonow). The judge (Benjamin Hewak) shockingly agreed with the Crown over defence objection. A further defence witness was sufficiently intimidated by police prior to trial to refuse to testify. After several days of deliberation, the Crown successfully applied to excuse a single juror – clearly the only one harbouring doubt about guilt. A conviction was entered minutes later.

The Court of Appeal properly overturned the conviction, emphatic that Sophonow should not be tried for a fourth time. The Crown unethically appealed the decision to the SCC – thankfully without success.

Sophonow sat in jail for 4 years for a crime he did not commit. The Manitoba government refused to provide any compensation – insisting that he needed to prove his innocence.

Fifteen years later (in 2000), Sophonow was cleared of the accusations and awarded $2.6 million, with an Inquiry struck. He still has no idea what information the state had, as they somehow continue to refuse to share it with him.

3 thoughts on “Part 2 – Manufacturing Guilt: Wrongful Convictions”

Leave a Reply