California has the most populous death row in America with some 517 inmates. Texas leads the nation in executions with 92 since 1976. Nationwide approximately 3,500 condemned men and women are currently waiting to die. According to Bureau of Justice statistics, this country executed 74 inmates in 1997, the most recent year for which data is available. Since the states began killing again in 1977, more than 460 men and women have been put to death. In the past few years Northwestern University journalism professor David Protess and his students have helped liberate three wrongfully convicted condemned men through two class projects in investigative reporting (their work wasn't published but was covered by the press and used by the defense lawyers). Inspired by those successes, MediaFile asked a prominent local defense investigator for tips on digging into death-row cases. Because of concerns about compromising his investigations, he asked to remain anonymous for this interview. Investigator X has been battling capital punishment for two decades on numerous fronts. The longtime Oakland resident has penned essays detailing the arbitrary and racist nature of the death penalty for The New York Times, San Francisco Examiner, Los Angeles Times, and numerous other publications. He has spearheaded activist campaigns against state killing and the prison-industrial complex. He has witnessed one of California's six contemporary executions. And for the past nine years, Investigator X has traveled the nation and the globe digging up exculpatory evidence for the defense of condemned prisoners.

MF: Where does a journalist begin looking into a death row case? What records do you start with?

Mr. X: The best way to get records is not from the institution, it's from the inmate; that is, through the inmate's lawyer. I think there's no moral way to do it without going through the lawyer. The lawyer has that person's life in their hands. Someone must be able to decide what is in the best interests of the condemned person in relation to living and dying, and that person can't be the journalist. Most journalists would scream to hear me say that. We have a strong disagreement on that. The answer is that if you're going to do an investigation like the [Northwestern projects], you're going to have to work with the lawyer. In that case the lawyer will be jumping up and down, thankful to have the help.

If you are going to undertake investigations of the reliability of a particular guilty verdict--well, that's a very difficult thing. But there are obvious steps. First you go back and look at the record and read what it was that convinced a jury that this person is guilty of a crime the person says he or she isn't guilty of. Then you go back and recreate the investigation in the sense that you start interviewing the people who know the most. That's the other codefendants, any witnesses; sometimes it's the police officers who investigated.

You'll have a whole record that's in the trial transcript as a start. If you are working with the lawyer, you'll have access to police reports, to other suspects. You'll recreate the scene from the convict's assertions--"This is why I didn't do it." Whether he's telling the truth or not, he'll have a complete story, and that is the story you're investigating.

MF: How hard is it, years after these crimes occur, to track down the relevant people and gather evidence?

Mr. X: That varies. Sometimes you'll have amazing luck and you'll find people will say things to you that you'd have sworn they'd never be willing to say. You also find people who don't like you and what you stand for and won't talk to you. Or you find people who have a lot to fear in their own lives and ultimately evade you. This happened to me with the witness [in the case of the recently executed Jaturun Siripongs] whom I located at both her residence and her work address in Thailand, and whom I believe might have been a perpetrator of the crime, and almost certainly was at the crime scene. She completely evaded me in Thailand. Even though I knew where she worked, even though I had her address and went to those places repeatedly. I have no power of compulsion. I can't make her come forward. I'm just an investigator; you're just a reporter.

Sometimes you get utter failure, but sometimes those returns, going back and back and back again, finally pay off and someone will talk to you. Something in their own life has changed their heart, and they're ready to talk. I had one case of a man on death row who I believe was unjustifiably convicted. He took part in a robbery, but not the murder--which makes him guilty under California law of one kind of murder [felony murder, accomplice to a robbery-homicide], but not of capital homicide. I wanted to talk to the snitch in the case, the jailhouse informant who testified on behalf of the government that he heard this man say this, and saw him do that, and knew him to be this. He agreed to talk to me briefly but wouldn't talk to me on the record, and evaded and evaded. But finally, once assured that nobody was trying to get back at him for his earlier testimony and that I was simply trying to get him to tell the truth, he testified and signed an affidavit undoing his earlier testimony.

MF: So it's all about pounding the pavement, knocking on doors?

Mr. X: That and luck. Here's a story I think isn't all that uncommon, but to me is shocking. In this case the young man I was working for was tried and convicted of capital murder and sentenced to death. One of my tasks was to interview the jurors--to find out the dynamics, what it was that they were convinced by so that the defense team's case in the next capital trial will focus on other themes, will avoid the things that led them to give death. The other task was to find legal errors committed by the jury so we could challenge their verdict as unfair or legally biased or a product of error.

In this case I found multiple errors, lots and lots of errors by lots and lots of jurors--by more than half of the jury. Some were minor errors, like saying to me, "I knew if they found him guilty I was giving him the death penalty. An eye for an eye." Well, that's not how the law works at all. There's a whole other trial to determine whether the death penalty is the appropriate punishment. But a good deal of the jurors knew the second they found him guilty that they were going to give him death.

The errors ran all the way up to a juror who brought a bible and held a prayer meeting at the deliberations--which is an out-and-out legal violation. There was another juror whose mother-in-law had been raped and murdered, and the murderer was executed. This juror answered "no" to the question during jury selection about whether anyone in her family had been the victim of a violent crime. That's a direct violation.

All these violations are a story in themselves. But the larger story is that not one single thing I found made any difference. The court said, "All this should have been found at such and such a cut-off point, therefore this court doesn't have to pay attention to this evidence. It doesn't exist."

MF: Which courts are more fruitful for journalists to study--trial courts or appeals courts?

Mr. X: The more important errors of justice occur in the trial phase. That is where you find those things that affect a condemned person forever, and they're the hardest to undo. That isn't what habeas corpus looks at. Habeas just binds itself to rules designed to kill your client, the person on death row. That makes for a dodge-ball game of law.

[Habeas corpus is the appeals process that allows a sentenced convict to enter new evidence in a case--such as counsel incompetence or misconduct by jurors, the prosecution, or witnesses. After exhausting state habeas appeals, a death-row inmate can petition the federal courts with a habeas appeal charging that the state convicted the inmate in an unfair, unconstitutional trial. At both the state and federal level, habeas appeals are regulated by a series of strict rules and timelines.]

But the real errors occur in the trial with bad gumshoe work, with noninvestigation, with nonrepresentation that doesn't quite fall to the level the Supreme Court has erected for ineffective counsel--which is a dead attorney! The court has already said a sleeping attorney is effective counsel, a drunk attorney is effective. And in California we've had examples of both sleeping and drunk lawyers.

The case I referred to earlier, where the snitch talked to me, that condemned man was represented by a floridly alcoholic lawyer. The lawyer was in the midst of a bitter divorce and was stalking his wife by bicycle because he'd had his license taken away for DUI. This was all going on during the trial!

Then there's the case of Jaturun Siripongs. The press didn't find it worthy of mention that the lawyer who represented Siripongs was at the same time running for Congress in Orange County. This was simultaneous. How do you run for Congress while at the same time representing in a courthouse a man whose life depends on you? And subsequently he never looked into Siripongs' life in Thailand, and presented nothing to the jury--and didn't even benefit anyway because he lost to Dan Lungren. I might add that Siripongs' case was the attorney's one and only capital trial. How could that not be a story?

A. Clay Thompson is a staff reporter at the San Francisco Bay Guardian, where he covers juvenile and criminal justice