Being poor, Thibodeaux could not afford his own lawyer and was assigned a public defence attorney by the courts. His attorney happened to be a former detective who had retrained as a lawyer, and this was his first murder case. At the time of the trial he was, unbeknown to Thibodeaux, applying for a transfer to the same district attorney's office that was prosecuting his client.
"I was willing to overlook the fact that he was an ex-detective," Thibodeaux says now. "But if I had known my lawyer was filing to be transferred to the DA's office I would have asked to have him removed from my case."
The trial lasted just three days. Over the course of it the prosecution tried to explain away the lack of any evidence of sexual intercourse or rape on Crystal's body by speculating that "semen-destroying maggots" had been at work.
Thibodeaux's lawyer, for his part, did not even refer to the confession. "You will read the entire trial transcript and he never utters the word 'confession', as though if he didn't mention it, it would go away," Kaplan says.
The jury was out for just 45 minutes before they delivered a guilty verdict. The next day the same jury sentenced Thibodeaux to death for murder and aggravated rape, even though no rape – indeed no sexual contact of any sort – had taken place.
Essentially he was convicted and sentenced to death because he was poor and could not afford a proper defence, and the state-provided one was simply not up to the task. But even when other lawyers got involved and did their job properly, the deck was so stacked against him that it still took 12 years for the state to accept his innocence.
The article quotes a defence lawyer as estimating that between 2% and 4% of death row inmates are innocent. Even if you believe the death penalty is morally acceptable (and I don't), that sort of error rate shouldn't be. But it raises a wider question about the US justice system: if they're finding this sort of error rate in death penalty cases, where presumably people are paying attention (even if only after the fact), what's the error rate in more ordinary cases, which face far less scrutiny? How many innocent people are in American jails, serving time and being brutalised for crimes they did not commit?