Marco Palacio, a fingerprint examiner with the Orange County Sheriff’s Office has been removed from casework. Performance issues, specifically failing to identify prints of value, mislabeling evidence, and clerical errors were the reasons given by the Orange County Sheriff’s Office.

At this point, the number of cases has grown to over 2,500 where Palacio may have affected the outcome of the case. Once the notice was given, the sheriff’s office scrambled to rework his active cases. Meanwhile, the Orange County Public Defender’s Officer is going through old cases to determine the potential impact.

This revelation has caused many in the criminal justice field to question the reliability of fingerprint evidence. But before these revelations in Orange County, the National Academy of Sciences’ watershed 2009 forensic report brought the flaws in this type of analysis into the limelight  The NAS report criticized many pattern-based forensic disciplines, including fingerprint analysis, for lacking scientific foundation. For example, many examiners testify that the error rate of using fingerprints to identify a person was zero, despite no study supporting such a contention.

Since then, the President’s Council of Advisors on Science and Technology (PCAST) report issued in September 2016 has acknowledged progress made by fingerprint examiners to place this type of analysis on a solid scientific basis. A number of studies have scrutinized the error rate of fingerprints as an identification tool, definitions of a quality latent print, and the effects of cognitive bias on an examiner. The results show that disagreement on definitions of a quality latent print constitutes a source of potential error among examiners. What is a quality, usable print to one examiner may not be to another.  Indeed, this is one of the reasons for Palacio’s removal from casework.

The error rate has turned out to be higher than most would have expected, particularly when examiners previously testified the error rate was zero. The two black-box studies on this point vary widely.  An FBI study places the error rate of false positives at approximately 1 in 306 examinations, while a Miami-Dade study placed this rate closer to 1 in 18. To put these figures in to context, this is the estimated error rate for a competent examiner. Palacio compared 700 to 1,000 fingerprints a month and errors were spotted in 75% of those comparisons.

Time will tell if the full extent of the damage is now known or if more Orange County cases need to be reviewed.