Another scandal about forensics

The FBI overstated forensic hair matches in nearly all trials up till 2000. 26 of their 28 examiners overstated forensic matches in ways that favoured prosecutors in more than 95 percent of the 268 trials reviewed so far. 32 defendants were sentenced to death, of whom 14 were executed or died in prison.

In the District of Columbia, the only jurisdiction where defenders and prosecutors have re-investigated all FBI hair convictions, three of seven defendants whose trials included flawed FBI testimony have been exonerated through DNA testing since 2009, and courts have cleared two more. All five served 20 to 30 years in prison for rape or murder. The FBI examiners in question also taught 500 to 1,000 state and local crime lab analysts to testify in the same ways.

Systematically flawed forensic evidence should be familiar enough to readers of this blog. In four previous posts here I’ve described problems with the curfew tags that are used to monitor the movements of parolees and terrorism suspects in the UK. We have also written extensively on the unreliability of card payment evidence, particularly in banking disputes. However, payment evidence can also be relevant to serious criminal trials, of which the most shocking cases are probably those described here and here. Hundreds, perhaps thousands, of men were arrested after being wrongly suspected of buying indecent images of children, when in fact they were victims of credit card fraud. Having been an expert witness in one of those cases, I wrote to the former DPP Kier Starmer on his appointment asking him to open a formal inquiry into the police failure to understand credit card fraud, and to review cases as appropriate. My letter was ignored.

The Washington Post article argues cogently that the USA lacks, and needs, a mechanism to deal with systematic failures of the justice system, particularly when these are related to its inability to cope with technology. The same holds here too. In addition to the hundreds of men wrongly arrested for child porn offences in Operation Ore, there have been over two hundred prosecutions for curfew tag tampering, no doubt with evidence similar to that offered in cases where we secured acquittals. There have been scandals in the past over DNA and fingerprints, as I describe in my book. How many more scandals are waiting to break? And as everything goes online, digital evidence will play an ever larger role, leading to more systematic failures in future. How should we try to forestall them?

4 thoughts on “Another scandal about forensics

  1. There are two issues at play here: bias and/or reliability of the method used. Regarding the latter it is somewhat surprising that the Wahington Post does not mention the recently established US Forensic Science Standards Board(FSSB) that is intended to address such inherent systemic failures – which may be grounds for some hope. Unfortunately, at the first FSSB public meetings in February in Orlando (FL, US) I was concerned by the fact that the Digital Evidence OSAC (Organization of Scientific Area Committees; digital evidence being my area of interest) seemed to primarily adopt existing practices, without any clear consideration of what it would mean to reframe the discipline from a scientific perspective. (I publicly expressed my concern at the meeting.)

    The establishment of the FSSB is based on recommendations put forward by the National Academies of Science (NAS) in the 2009 NAS report. Chapter 5 of that report deals with shortcomings in many forensic disciplines. It is relatively positive about the state of digital forensics science; however, digital forensic science was born as a distinct discipline while the report was being written. At this stage the concerns about other forensic disciplines in the NAS resport apply more and more to what is done under the rubric of digital forensics. So, Ross, I share your concern expressed at the end of your post. A few of us are working on (tentative) ideas to address your final question, but our work is currently far from mainstream.

  2. Intellectual honesty – integrity, in other words – is rapidly becoming endangered in the West. It is being driven to the edge of extinction by the only thing that far, far too many people think has any value: money. If you are a professional, one of the few ways to accumulate money is to become famous or to get promoted in your organization. And that is done, all too often, by producing the results that make your bosses smile. Of course systematic bias, such as occurred in the FBI labs, may be unconscious; but it is no less pernicious for that – indeed, maybe it is even worse. The only solution is to reward professionals for doing things right, not for delivering the results their bosses like.

Leave a Reply to Ross Anderson Cancel reply

Your email address will not be published.