Every year, the California Board of Parole Hearings decides whether thousands of people serving life sentences are ready to go home. One factor commissioners weigh is whether the applicant has “insight” into the crime that put them in prison. For someone who was wrongfully convicted, that creates an impossible choice: confess to a crime you did not commit, or maintain your innocence and risk being told you lack insight.
Our clients have described this dilemma for years, but it has been difficult to measure. That changed this summer, when a team of mathematicians from the UCLA Mathematics REU’s AI for Justice team, working with The Innocence Center, analyzed a full year of California parole hearing transcripts covering 4,639 decided hearings.
The findings are striking.
“Non-innocence claimants saw a 25% raw grant rate, over double that of innocence claimers.”
Applicants who maintained their innocence were granted parole just 12% of the time. Even after the researchers compared hearings with similar characteristics, including risk assessment scores, disciplinary history, education, and outside support, an 8-point gap remained. According to the report, this is consistent with a penalty attached to the innocence claim itself.
The mechanism appears to be “lack of insight.” Among denied hearings, panels cited lack of insight or minimization of the crime in 87% of cases where the applicant maintained innocence, compared with 61% where they did not.
“Insight, though a reasonable standard, appears to operate differently, and less forgivingly, for applicants who maintain their innocence.”
The Luck of the Draw
Perhaps most troubling is how much the outcome depends on who is sitting across the table. Of the 21 presiding commissioners who heard more than 150 cases in the data, four did not grant parole to a single applicant who claimed innocence. One commissioner granted parole to roughly 31% of applicants who did not claim innocence, and to none who did. Another granted parole at essentially the same rate regardless of what the applicant said about guilt.
“Whether an innocence claim proves fatal to an application may depend in part on which commissioner hears it.”
A person’s freedom should not turn on which name appears on the hearing calendar.
Why This Matters
California law already recognizes this problem. Penal Code section 5011(b) states that the Board shall not require an admission of guilt when setting parole dates. Courts have enforced that rule: in In re Jackson (2011), a California Court of Appeal held that the Board violated section 5011 by relying on an applicant’s refusal to admit guilt. Yet the data suggest that, in practice, maintaining innocence still carries a real cost. More than 30,000 parole-eligible people are serving life sentences in California, and research suggests a meaningful share of incarcerated people report being wrongfully convicted.
What the Courts Have Said
In 2011, in In re Shaputis, the California Supreme Court drew a line that matters enormously for people who maintain their innocence. The court suggested that when a plausible denial of guilt is the only evidence of “lack of insight,” it may not justify denying parole. But it also held that an implausible denial can support a finding that someone remains dangerous, and that deciding what is plausible is largely up to the Board. Reviewing courts must defer to that judgment as long as there is “some evidence,” which the court described as a “modicum,” to support it.
In practice, that means the protection section 5011 offers an innocent person can depend on whether a particular commissioner believes them. Our data suggest commissioners reach very different conclusions.
Justice Goodwin Liu saw the danger. In a concurring opinion, he warned that the Board may not deny parole simply because someone refuses to admit the official version of the crime, and cautioned against allowing lack of insight to become
“a new talisman with the potential to render almost all life inmates unsuitable for parole.”
Nearly fifteen years later, the numbers suggest that for people who maintain their innocence, it may have.
The researchers put it plainly: applicants who may have been wrongfully convicted
“face a process that can require a confession as a condition of release.”
What Comes Next
These are initial findings from a single year of data, and they show associations rather than prove intent. But they put numbers to something our clients have told us for a long time: telling the truth about your innocence can keep you in prison.
This is only the beginning. The Innocence Center intends to gather and analyze parole hearing data every year, so we can track whether this pattern continues, how it shifts across commissioners, and what it looks like as parole practice evolves.
We are grateful to Cameron Golbahary (UCSD), Lowell Justice (Wesleyan), Chi-Hao Wu (UCLA), and Professor Deanna Needell (UCLA) for this important work. The Innocence Center will continue to push for a parole process where no one has to choose between honesty and freedom.
The full report can be viewed and downloaded here.
