Law enforcement agencies sometimes use dog-scent identifications in criminal investigations. They use dogs trained to recognize and match smells to identify substances and people involved in crime. For example, appropriately trained dogs are supposed to recognize a person’s unique smell and then match it to an item or location. This evidence can be introduced at trial and used against a defendant. But as discussed in a recent federal appeals court decision, there can be problems with dog-scent identifications.
Aguilar v. Woodford involved the murder of a man named Guerrero, who had been shot in his car at a stoplight by the driver of a white Volkswagen. (No. 09–55575 (9th Cir. 2013).) Gilbert Aguilar was found guilty of the murder, but at trial he claimed the shooter had been another man, Richard Osuna.
The prosecution based its case against Aguilar on shaky eyewitness testimony and a dog-scent identification linking Aguilar to the white Volkswagen. There was no other physical evidence linking him to the crime, and no evidence of motive. Further, there was substantial evidence linking Osuna to the crime, including statements he made about shooting someone and evidence that witnesses saw him jump into a white Volkswagen and chase the victim’s car. The prosecution refused to conduct an investigation of Osuna and told the police not to pursue it.
In the Aguilar case, police conducted a scent comparison test. They took a sample scent from the vehicle involved in the case, then compared it to a scent taken from Aguilar’s clothing. A trained scent dog, Reilly, was used to determine whether there was a match between the sample scent pads; if there was, he would bark. Reilly did in fact bark, and the prosecution used this scent evidence at trial to link Aguilar to the vehicle involved in the murder. Reilly bolstered the otherwise weak case against Aguilar.
Dogs, just like people (and forensic experts), make mistakes. Dog-scent evidence is prone to error just like many other forms of scientific analysis. And when there’s a specific reason to doubt the accuracy of evidence pointing to a defendant’s guilt, the defense has a right to it.
Two months before Aguilar’s trial, the prosecution revealed in another case that Reilly had made mistaken scent identifications on two prior occasions. The judge in that case didn’t allow the dog-scent evidence to be introduced because it was too flawed. By the time of Aguilar’s trial, Reilly was no longer working as a scent dog.
Aguilar’s attorney didn’t know about the case in which Reilly’s identification was barred from trial. Nor did the lawyer know about Reilly’s previous failed identifications. The federal appeals court found that, had Aguilar’s attorney known these facts and presented them to the trial judge, the judge wouldn’t have admitted Reilly’s identification into evidence. And even if the judge would have still allowed the evidence, Aguilar’s lawyer would have been able to make a persuasive argument to the jury that the identification couldn’t be trusted.
Dog-scent evidence is tricky because the pooch obviously can’t testify. Aguilar’s lawyer couldn’t have asked Reilly about previous mistakes or ever mixing up smells. So, attorneys like Aguilar’s have to rely on experts to explain the procedures and reliability of dog-scent identifications. And it’s that reliability (both in general, and when it comes to a given dog), or lack thereof, that causes judges to frequently reject dog-scent evidence.
While dog-scent information can be used at trial, police and prosecutors need to follow proper procedures. If there are any doubts about the dog’s consistency or ability, then that information has to be turned over to the defense. Otherwise, law enforcement risks a mistrial or overturned conviction.