Case of a Lifetime
What kept Patsy Kelly Jarrett in jail for a murder she says she didn't commit? According to the author, a Georgetown University law professor who spent decades trying to get her released, it was a misguided devotion to the truth.

By Abbe Smith
Sunday, June 29, 2008; W14

I was in my second year of law school when I met Patsy Kelly Jarrett. It was autumn 1980. We shook hands and sat together in the visiting room of the Bedford Hills Correctional Facility, in Westchester County, N.Y. I was 24; she was turning 29. She didn't seem to mind that I was still in law school and younger than she was. Maybe she had no choice: She needed help, and services provided by the New York University prison law clinic were free. A pretty tomboy with a fresh-scrubbed look and a soft Southern accent, she said her friends called her Kelly, not Patsy. She had come to despise the name Patsy because it summed up too precisely what had happened to her in the summer of 1973 -- her first and last adult summer vacation.

Kelly was serving a life sentence for robbery and murder. At her trial, the prosecution had maintained that she had acted as a getaway driver while her traveling companion, a man named Billy Ronald Kelly, killed a 17-year-old gas station attendant in Sherrill, N.Y., on August 11, 1973. A witness testified that he saw Kelly at the scene behind the wheel of a car that matched the vehicle that she owned, and in which she and Billy Ronald had driven north from her home near High Point, N.C., earlier that summer. At her trial, as she would before and after, Kelly said she was innocent.

Kelly's case, my first, would end up following me for the next 25 years. Kelly was convicted based on nothing more than the memories of that one eyewitness, who saw the driver of the car fleetingly and, when first questioned, couldn't even say whether the person had been a man or a woman. It remains the most haunting miscarriage of justice I have ever encountered.

Yet the case haunts me in another way. At least twice, Kelly was given the chance to plead guilty and dramatically reduce her sentence -- to as little as five or 10 years. She refused; she said she couldn't swear to something she didn't do. She should have.

Patsy Kelly Jarrett and Billy Ronald Kelly met in High Point in the fall of 1972, when they were working at the same textile mill. They were introduced by mutual friends who knew they were both gay. When Billy Ronald, who was about eight years older than Kelly, suggested they visit his friends in Utica, N.Y., the next summer, Kelly didn't think twice. She was headed to New York -- an exciting destination for a small-town girl.

They traveled in Kelly's car, a metallic-blue Plymouth Road Runner with silver-slotted mag wheels and a slightly jacked-up rear. Feisty and fast, it was the perfect car for a working-class tomboy.

Once in Utica, Kelly and Billy Ronald made themselves part of the gay community. There was a gay bar, the Hub, and a bartender named Gerri. Before long, she and Kelly became lovers. Kelly spent most of her days with Gerri and her nights at the Hub. On weekends, she watched Gerri play with a women's softball team. The days blended together. Kelly has always wished she could remember more.

Billy Ronald took up with William Sullivan, a girlish man in his 20s with shoulder-length brown hair. Billy found a job working construction. The arrangement between him and Kelly was that he'd pay for their lodging in exchange for the use of Kelly's car. Kelly would live off her savings.

Their time in Utica came to an end in mid-August 1973, when Kelly ran out of money. She bid a tearful farewell to her summer love and headed back home. Billy Ronald decided to go with her.

On the way, there was one strange incident: At a Kayo gas station in Danville, Va., Kelly awakened, alone, thinking she'd heard a backfire. Billy Ronald came back to the car agitated, his hand cut. Asked what had happened, he told Kelly to mind her own business. Groggy, she went back to sleep. She and Billy Ronald parted ways once they got home.

Some months later, two Virginia State Police officers contacted Kelly. The "backfire" had been a shotgun blast, killing the Kayo's attendant. A witness had gotten her car's plates. Horrified, Kelly identified Billy Ronald and agreed to testify against him. Billy Ronald pleaded guilty and was sentenced to 35 years. The Virginia authorities -- not known for their leniency -- did not believe Kelly to be an accomplice.

Two and a half years after that, there was another knock on Kelly's door. Two New York State Police officers told her she was wanted for the August 11, 1973, murder and robbery of 17-year-old Paul David Hatch.

I'd chosen to work on Kelly's case on my first day at NYU's prison law clinic. The director, Claudia Angelos, said the superintendent of Bedford Hills prison had called personally to ask her to take it on. No one had ever heard of a superintendent doing such a thing.

Under Claudia's guidance, I worked for two years on Kelly's petition for a writ of habeas corpus, through which prisoners appeal their convictions in federal court. I visited Kelly at the prison monthly, sometimes more. I wanted to keep her informed of what we were doing, and I wanted to keep her company.

The second time I met with Kelly, she asked me whether I was gay. I'd had relationships with men and women, so I told her she had good intuition. In the years since, people have asked whether my devotion to Kelly had to do with our shared sexual orientation. But we have always been more different than alike: Kelly is a Southern, small-town, working-class, deeply religious Catholic; I am a Northern, big-city, secular and skeptical Jewish professional. I would have been moved by Kelly whether she was gay or straight.

The crime itself had been brutal. Paul David Hatch, a star athlete who had just graduated from high school, had been working alone at the Sherrill gas station. His body, bound and gagged in a back room, was discovered by his girlfriend shortly after 1 p.m. His throat had been slashed; $279 was missing.

The only evidence placing Kelly at the scene was the testimony of a man named Robert Hyland, a night watchman at a plant up the road, who had driven up to the pumps at approximately 12:50 p.m. With no attendant in view -- these were the full-service days -- Hyland got out of his car and stood there a minute or two. As he was waiting, a blue car pulled up to another pump.

The man who emerged from the station and pumped his gas was someone Hyland had never seen before. There was something peculiar about him; he seemed nervous and had trouble making change. Hyland got a good, close look before driving off with $5 worth of gas.

Two days later, when initially questioned by the police, Hyland gave a detailed description of the presumed killer and helped create a sketch of the man. But he said he hadn't seen the face of the blue car's driver. He remembered "long, black shoulder-length hair," but could not describe age, build, clothing, speech or manner. He couldn't say whether the driver was a man or a woman.

The police investigation rolled on, but the department did not contact Hyland again until December 1975, 28 months after the murder. An officer who had worked on the Virginia case had called to suggest that the New York police take a look at Billy Ronald. An officer showed Hyland two photo spreads. From a set of males, Hyland immediately selected Billy Ronald. From a second group -- all women -- Hyland initially picked two photos, saying they both looked like the person in the car, before settling on Kelly's.

Before the trial, Kelly's court-appointed attorney went to the prosecutor. He pointed to the shaky eyewitness identification, the problem of guilt by association. Unmoved, the prosecutor said his best offer was a plea to robbery, which carried a sentence of five to 15 years. Kelly wouldn't hear of it. She trusted the system: She was sure that if she told the truth, she'd be believed. Everything would be okay.

The trial began in New York on March 14, 1977, more than 3 ½ years after the murder. Kelly and Billy Ronald were tried together, and the evidence against Billy Ronald was overwhelming: His fingerprints had been found on the tape used to bind the body. On the stand, Hyland positively identified Kelly for the first time. When Kelly's lawyer pointed out Hyland's earlier uncertainty, the witness became increasingly adamant.

Multiple studies have shown that a witness's certainty has little to do with the accuracy of identification, but most people don't know this. A common and troubling misperception is that the more confident a witness is, the more accurate the testimony is likely to be. The jury convicted Kelly on March 29, 1977.

I must have gone through the trial transcript 100 times, reading and rereading, indexing, trying to find useful bits to incorporate into the petition. Another clinic student and I visited Billy Ronald at the dreary-looking maximum-security Powhatan Correctional Center in State Farm, Va. He was polite but guarded. He declined to discuss William Sullivan, who we believed had been with him on August 11, 1973. But he was willing to say that he had Kelly's car for the 24-hour period that included the time of the crime, that Kelly was not with him, and that Kelly did not know his whereabouts or actions when he had use of her car.

The process of writing and filing the petition was painstaking and slow. It wasn't finished when I graduated from law school and joined the public defender's office in Philadelphia. I became engrossed in my new life and career. I kept in touch with Kelly, but our contact was sporadic.

To get into federal court, you need plain legal error. Mistaken identification cases are always about factual, not legal, errors. Once convicted, their victims are all but doomed. Claudia used the only argument she had: that the trial judge erred in allowing unreliable identification testimony.

I had been a defender for three years when Claudia called to say Kelly's habeas petition had been granted, overturning her conviction. This was an amazing, against-the-odds victory. Kelly had served 10 years; I couldn't believe that she had prevailed after all this time.

Claudia was pleased but worried about the future. The state will appeal, she said. And the federal appeals court was known to be a much tougher forum. Claudia was right. The state appealed. But before the case was heard, the state made an offer: If Kelly pleaded guilty, she would receive a sentence of time served. This meant that she would be released from prison and could go home.

The question of how hard a lawyer should lean on a client to take a plea is a difficult one. Sometimes effective counseling -- getting through to a client about the reality of his or her situation -- means leaning very, very hard.

Kelly's time behind bars had not been easy. Early on, horrified at the prospect of spending her life incarcerated, she escaped, walking out of a low-security prison. Apprehended, she was transferred to Bedford Hills, where she was not allowed to go anywhere without being searched, often a full-body strip search. Certain jobs were off-limits to her, as was the possibility of living on the Honors Floor, the destination of most well-behaved long-term prisoners. Now, Kelly was doing the hardest kind of time.

Claudia advised Kelly to take the plea, told her that she was lucky to win in the district court, warned her that she would likely lose in the appeals court. She said that if she were in Kelly's shoes, she would take the plea. But, to this day, Claudia blames herself for not exerting enough pressure.

Kelly refused. She was innocent; she wanted vindication. Six months later, the court of appeals reversed and reinstated her conviction. She was now serving a life sentence with no legal remedy.

As the years passed, I went from being a public defender to a law teacher to a combination: a clinical law teacher running a criminal defense clinic at Harvard Law School. I thought of Kelly from time to time, but it was painful to do so. Then, in 1993, I helped organize a conference about women in prison. The keynote speaker was Jean Harris, the former headmistress of McLean's exclusive Madeira School. She had recently been released from Bedford Hills after serving 12 years for the murder of her lover, Herman Tarnower, the author of The Complete Scarsdale Medical Diet. I told her I'd seen her at the prison while visiting Kelly.

"Oh, Kelly," she said immediately. "Everyone knows she's innocent."

At Harris's urging, I contacted Kelly. I apologized for being out of touch, told her what I'd been doing since we had last corresponded, asked about her health and her case. She wrote back saying that a friend -- a former prisoner who wasn't a lawyer but believed in her innocence -- was helping put together a clemency petition.

There was something about Kelly that threatened my usual equilibrium. I knew that if I took up her case again, her innocence would be on my shoulders. I resisted, then caved. Recruiting a couple of law students, I set out to do what I could. There was a strange symmetry to this. I was poring through Kelly's papers again, years later, only this time as the clinical professor, supervising my own students.

We drove to Bedford Hills. It was good to see Kelly again, to hug her, to be in her presence. But it was also sad. She was older, grayer, less lighthearted. When Kelly was first incarcerated, she weighed just over 100 pounds; now she was closer to 200. So much had happened in my life. So little had happened in hers.

Since my days as a public defender, I always have been a proponent of clear-cut, unambiguous professional boundaries. But boundaries are a tricky thing in lawyering. The more the lawyer cares about a client, the trickier they become. With Kelly, I acted in ways I never acted with other clients. I sent her clothes, shoes, books, a Scrabble set, a portable stereo. I intervened in prison matters and family matters. I shared my own life. She signed her letters "love." Soon I signed mine the same way.

It helped that Kelly was respectful. She respected that I had other clients, other demands on my time, that my attention was sometimes directed elsewhere. She respected the family I had built with my partner, Sally, and was thrilled when our son, Joe, was born in 1995. Kelly showed me that sometimes -- not often, but sometimes -- being a good lawyer also means being a good friend, no matter how uncomfortable I am with the idea.

Over the next several months, trying to leave no stone unturned, the students and I contacted Kelly's first lawyer, eyewitness Hyland and the judge who had presided over the original case. We drafted a careful letter to the parents of the young man who had been killed. We tried to find William Sullivan, Billy Ronald's old boyfriend. But the case had occurred more than 20 years before. Most of our letters went unanswered. Billy Ronald responded to our meeting request with one line: "I decline your invitation and wish you luck in your endeavors."

A few years earlier, Sister Antonia, a nun who worked at the prison and who had become Kelly's godmother, had visited Hyland, and he had admitted feeling pressured to identify Kelly at the trial. He agreed to see us, and the students and I took a two-day trip to Upstate New York. What if he said he identified Kelly only because the police told him she was a prime suspect? Could we get a new trial? Get the police to reopen the case?

But Hyland gave us nothing. He was in his early 80s, on oxygen and had apparently had a stroke. He was happy for the company but confused about the case. He stuck to the story he'd told at trial.

Meanwhile, Kelly wrote me letters. Hundreds of them. Every one began by thanking me. In 1993, she wrote: "I'm so happy to have you back in my life, Abbe. I will never forget you for not forgetting me." Some expressed frustration, indignation and sadness -- at times sardonically, such as the words she scrawled at the top of a letter to me on April 24, 1997: "20 years in prison. Wish me a happy anniversary!"

We were essentially back where we'd restarted: Kelly was a convicted murderer with a strong claim of innocence seeking executive clemency from a governor -- Republican George Pataki had since taken office -- who would be loath to grant it. Yet clemency was the only option. We had to move forward.

Our petition, seven inches thick, was denied in July 21, 1998. Under clemency procedures, we had to wait a year before filing another.

During the time Kelly was in jail, the U.S. prison population boomed -- from 1977 to 2004, the number of women prisoners rose by more than 750 percent, according to U.S. Department of Justice statistics. The population at Bedford Hills became younger and more troubled. It was a more dangerous place for a middle-aged woman such as Kelly. Once, a younger inmate who thought Kelly had "disrespected" her viciously attacked Kelly with a metal bucket. Kelly's nose was broken, and she lost sight in both eyes for several days. The attack frightened and angered her.

Kelly's friends did their time and got out, while she remained imprisoned. Whenever I asked her how she tolerated it, how she slept at night, she would answer, "God." I should have been glad that she found comfort in faith. But I had a hard time understanding it.

Every couple of years after the first denial, I filed a new clemency petition, updating, adding, revising, polishing, reworking. Every couple of years, it was denied. I created a file for the one-page form letters, which I optimistically labeled "Correspondence with the Clemency Board." It seemed as if the clemency petitions would be denied no matter how carefully put together. What I needed was access.

I contacted a high-ranking lawyer in the criminal division of the New York State Attorney General's Office, the wife of a college classmate. I felt a little uneasy exploiting that connection, but my friend was now a law professor whose scholarship focused on social justice. How could I not?

The prosecutor listened to what I had to say and agreed to look over the case file. When she called me back, she said her office had no doubt about Kelly's guilt.

No doubt? It must be nice to have no doubt, none whatsoever. Meanwhile, I'm in doubt reading a menu. I'm drenched in doubt buying a pair of shoes. But this prosecutor had no doubt about a one-witness identification case.

It is hard to say exactly what Kelly was feeling during those years. She seemed the same: alternately hopeful and discouraged. Every once in a while, she would throw up her hands and say it wasn't worth it. I ignored her and carried on.

I had my own struggles. Trial work suits me: A case usually goes to trial within a year of the arrest; it's over before you know it. Whatever the result, you move on. Kelly's case threatened to go on forever. I began to think of myself as the most devoted but least adept lawyer on Earth. If there was a category in the Guinness Book of World Records for "ineffectual lawyering," I could be a contender.

Kelly sent a card on every Jewish New Year, birthday, Thanksgiving, Hanukkah, Valentine's Day, Passover and Mother's Day. She sent away for those harder-to-find Jewish holiday cards. She signed them "Shalom." How could I not continue to fight for her? Who else did she have?

Over the years, I contacted countless writers, hoping to get a good newspaper or magazine article. I contacted public relations firms. I wrote about Kelly in law journals. Then, in the spring of 2003, documentarian Ofra Bikel, who was working on a film about guilty pleas for the PBS television show "Frontline," heard about Kelly's case. We talked about including Kelly in the film. I kept calling until Ofra said yes.

Around the same time, people suggested I get in touch with Sara Bennett, then head of the New York Legal Aid Society's fledgling Wrongful Convictions Unit. I called her that summer; she immediately signed on. We both knew that Christmas, only five months away, was clemency time.

With the help of a Legal Aid board member, we were granted a meeting with the governor's counsel. It went well, and we were elated to learn that Kelly would go before the clemency board in November. This was big. Kelly and I had never gotten this far.

Lawyers aren't allowed at clemency hearings, so Kelly would have to speak for herself. Sara and I went into high gear, preparing her in person and by telephone. We conducted practice sessions where we shot hard questions as if we were hostile board members. Kelly had a particularly good answer to the question about where she would live and work if released: at the Mount St. Francis convent in Peekskill, N.Y., where Sister Antonia lived; she would work in the convent infirmary. What could be better for a parole board -- a prisoner who wants to live in a convent!

Kelly later told me she wept through much of the clemency hearing. She said that the board members seemed kind and friendly. A couple had wiped away a few tears themselves, she told me. But on Christmas Eve 2003, we learned that her petition had been denied. It turned out that Gov. Pataki extended executive clemency in only two cases that year -- one of them to an unwitting drug mule who had served 11 years, and the other to Lenny Bruce, who had died in 1966. It felt like a cruel joke.

Under New York state law, parole was now all Kelly had left. But good things had started to happen. Ofra's film, "The Plea," aired in June 2004, and suddenly Kelly had supporters all over the world. Pataki must have gotten mail, too. Soon after, Kelly was told that she was approved for the Honors Floor. She was overjoyed.

In 2005, Sara called me to talk about the parole effort. She wanted to be clear: She would take this on only if she was lead counsel. I was taken aback; Kelly had been my client for years. I had to remind myself that this wasn't about me. Sara was based in New York and knew the institutional players. I needed to step back.

On a Thursday night that April, while I was at my son's baseball practice, I got a call. Kelly's parole petition had been granted. She'd be getting out in June. I was relieved, but not as happy as I thought I'd be. Kelly had served her full minimum sentence: 28 years and 6 months. She was among the longest-serving women in the New York prison system.

After that, the weeks passed quickly. I confess that I struggled with a feeling of anticlimax and envy. It felt like I'd been trying to open a stuck jar for years, using every ounce of determination and strength, and then someone else got it off in a single twist -- as if nothing had preceded her effort. I understood that most people convicted of violent crimes are not released at their first eligibility. But I was tormented by the idea that Kelly might have been paroled even if I had done nothing, if I had never come back into her life.

When I shared my distress with my friend Ilene, saying I wish I could let go of these feelings, she said: "You? You can't even let go of a bad sandwich." She pointed out that Kelly had been my life's work. The end was bound to be difficult.

Kelly has been free for more than three years. Upon her release, she moved into a one-bedroom apartment on the grounds of the Mount St. Francis convent. She works two jobs: caring for sick and elderly nuns in the infirmary and providing in-home nursing care around Peekskill. She has a bank account, credit card, car and cellphone. She volunteers one morning a week at a homeless shelter.

I am so proud of everything Kelly has accomplished. But I hate that she is under the supervision of the parole system. I also hate that she is too frightened to meet people. Her choice of a friend cost her 28 years, and she says she won't make that mistake again. She stays close to the convent.

Since Kelly's release, I've become increasingly aware of our differences. When I saw the film "Brokeback Mountain," a story of forbidden gay love in rural America, I urged her to see it. I thought it might resonate. Kelly saw it with a church friend; her response was tepid.

"Didn't you like it?" I asked.

"Well, it was awfully explicit, and my friend wondered if I was trying to convert her or something," she said.

"Didn't you feel for the two men who weren't allowed to love each other openly?" I asked.

"Well," she said. "I think if you're gay you should just be gay and not get married and cheat on someone."

In more than 25 years of law practice, I have represented a handful of other clients I believed to be innocent and many more who, guilty or not, did not deserve to be so harshly punished. Since Kelly's, I've not had a case like this. I am glad of that. I don't think I could bear another.

Abbe Smith is professor of law and co-director of the Criminal Justice Clinic at Georgetown Law School. She can be reached at smithal@law.georgetown.edu. This essay is adapted from her book Case of a Lifetime, scheduled to be released next month. Copyright 2008 by the author and reprinted by permission of Palgrave Macmillan.