Retrial Of Texas Arson Case Ends With Surprise Guilty Plea Robert Siegel talks with Dave Mann of the Texas Observer about the conclusion of the retrial of Ed Graf on arson charges. On Wednesday, Graf pleaded guilty while the jury was still in deliberations.


Retrial Of Texas Arson Case Ends With Surprise Guilty Plea

Retrial Of Texas Arson Case Ends With Surprise Guilty Plea

  • Download
  • <iframe src="" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript

Robert Siegel talks with Dave Mann of the Texas Observer about the conclusion of the retrial of Ed Graf on arson charges. Graf had been convicted of arson decades ago for setting a fire that killed his two young stepsons. Still, he has always maintained his innocence. A court found that the evidence used in his trial was so questionable that he deserved a retrial. On Wednesday, in a surprise move, Graf pleaded guilty while the jury was still in deliberations, apparently deadlocked on the question of his guilt or innocence.


The story of Ed Graf of Hewitt, Texas, near Waco is a real-life forensics thriller full of plot twists that would be hard to believe even in a fictional TV-crime-lab procedural.

First, Graf was convicted of the arson deaths of his two young stepsons in 1986. He was sentenced to life in prison. And for 28 years, he protested his innocence. His conviction was based on expert testimony about arson that is now considered junk science.

Last year, the Texas Court of Appeals agreed he'd been convicted on the basis of false expert testimony. They ruled and he deserved a new trial. Well, that trial ended this week in stunning fashion. With the jury out deliberating, Ed Graf changed his plea to guilty. Dave Mann of the Texas Observer has followed this story from its earliest days. Welcome.

DAVE MANN: Thank you.

SIEGEL: Graf's first conviction was based at least in part on forensic evidence about arson that came under attack as specious. Now, after a trial I gather without forensic evidence, he pleaded guilty. Does that mean that way back then justice was served albeit very sloppily?

MANN: Not necessarily. I think at first when he pled guilty earlier this week, a lot of people were surprised. And as the details of his plea agreement emerged, it began to make more sense. His plea agreement stipulates that he'll get a six 60-year sentence. And under those stipulations - under some quirks of Texas law - that would make him eligible for parole. And he'll actually be able to get out in few months. So essentially Ed Graf kind of gave up his quest for innocence and pled guilty for the opportunity to get out of prison in a few months.

SIEGEL: He's doing that at a time when the jury had been reported to be deadlocked. Do we have any idea where they were headed?

MANN: Well, the evidence indicates that they were headed for a guilty verdict. They'd initially indicated after their first few hours of deliberations that they were 10 votes for guilty and two votes for not guilty. And as the deliberations went on, the prospect of a hung jury became more likely and the two sides began negotiating a plea agreement. And just as they finished that plea deal, the jury reached a verdict. And in fact, as they were finalizing the plea agreement, the bailiff was entering the courtroom with a note saying the jury had reached a guilty verdict.

SIEGEL: Of course, that verdict was never delivered because of the deal that had happened. This was the very same jury that I gather had sent a question to the judge - how many jurors would have to vote for a verdict to make it unanimous?

MANN: Yeah, in a - in what's been a very emotional and very serious saga and a very emotional trial, that was one of the lighter moments. One of their first questions to the judge was how many votes does it take for us to be unanimous. And he informed them that all 12 had to concur with the opinion.

SIEGEL: This was a trial about two things - on the one hand, the guilt or innocence of Ed Graf and the other the validity of the kind of evidence that was used to convict him the first time - arson evidence. His - his story notwithstanding, there's been something of a revolution in how that evidence is viewed nowadays.

MANN: Yeah, the Ed Graf case is really part of a complete overhaul in how we view fire investigations and fire science. Just about everything that fire investigators once believed were indicators of arson have now been shown to be unreliable. And many of the burn patterns and other indicators that fire investigators used for decades to convict people of arson have shown to essentially be junk science.

We have a much better understanding of how fire behaves now. And arson cases are prosecuted under, you know, what is now probably true science. And that science would lead the experts to believe that the fire in the Ed Graf case was in fact accidental. And that's what some nationally-known experts testified to over the past two weeks.

SIEGEL: Ed Graf's case was championed by the Texas Innocence Project. What did they say about him suddenly pleading guilty?

MANN: Well, I think there's a little bit of confusion because the director of the Texas Innocence Project was quoted in the Waco paper kind of saying well, the most important thing here is the science and not whether each individual person is guilty or innocent. The most important thing is that we reexamine the science and that the proper fire science was offered in the courtroom, no matter what the verdict is.

So I think there was a little bit of hedging there. The reality is that Ed Graf is going to get out of prison because of the quirk in Texas parole law. But I think that the fact he got up in court and admitted guilt definitely rattled some people who for a long time had believed in his innocence.

SIEGEL: Dave Mann, thanks for talking us today.

MANN: My pleasure.

SIEGEL: Dave Mann is the editor of the Texas Observer.

Copyright © 2014 NPR. All rights reserved. Visit our website terms of use and permissions pages at for further information.

NPR transcripts are created on a rush deadline by an NPR contractor. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of NPR’s programming is the audio record.