When the U.S. Supreme Court hears arguments in Harris v. Florida (Docket No. 11-817) this fall, it is likely to determine whether training and certification are sufficient to establish a drug dog’s reliability for a search of a vehicle. The Florida Supreme Court determined that the prosecution had to produce the dog’s field records and had the burden of establishing probable cause.
In a recent California case, the tracking and alert of an uncertified tracking dog that would probably not have met the reliability standard required of a detection dog in Florida was admitted to establish the identity of the perpetrator in a murder. Although the U.S. Supreme Court will not consider a tracking case, its pronouncements regarding training, certification, field records, and the evidentiary burden to establish reliability are likely to affect future decisions in cases such as the one we analyze here.
A Murder in Southern California
Carlos C. was shot and killed as he rode on his skateboard in Ontario, California, the evening of November 6, 2008. The shooter was described as a black man wearing a hooded sweatshirt and khaki shorts. Officer Louis Mena was the first to arrive, finding Carlos alive but not responsive. He was pronounced dead in a hospital emergency room a short time later. He had been shot twice in the back.
One witness, Jair Monares, who had encountered Kevin Tyron Stanford shortly before the shooting, described Stanford pulling a gun on him and shooting it in the air. Monares picked Stanford out of a lineup and testified that he had seen Stanford wearing a hoodie and shorts. Another witness, Maricela Sierra, heard someone yell “Black Crips” and went out from the laundry room of her apartment where she saw a black man wearing a hoodie and khaki shorts chasing someone riding a bicycle.
A third witness, Raul Ramirez, was riding his bicycle when he heard gunshots and then encountered the defendant, whom he identified at trial. Stanford apparently said to Ramirez, “What the hell are you looking at fuckin’ Mexican. Go back to Mexico, fuckin’ wetback.” Ramirez rode off but Stanford chased him. The bicycle’s chain derailed and Ramirez fell against the curb. Stanford took the bicycle and fled. Maricela Sierra, having caught up with Ramirez, helped him describe events to the 911 operator.
A fourth witness, Jessica Villalobos, who happened to be Jair Monares’s girlfriend, saw a black man from the second floor apartment where she lived, then later heard gunshots after which she saw a body on the sidewalk and a skateboard rolling down the street. She saw the defendant pick up the bicycle and head towards a church. She called Monares, who left his house and saw Stanford for a second time, though this time Stanford did not look at or speak to him.
Ava Stone was returning to the apartment which she shared with a number of people at about 7:30 when she was stopped by a police officer, who told her that a suspect in a murder had been seen entering her apartment. The officer asked her to have all males in the apartment over 18 years old come outside. The defendant was in a bathroom and ignored her request, so she and another man told the police. The defendant came out after taking a shower and changing his clothes.
Monares and Ramirez were brought by the police to the area. Monares recognized and identified the defendant as the person he had twice seen as events unfolded. While Ramirez recognized the defendant, he was too afraid to make a formal identification at that time. The Ontario police also brought a bloodhound, which followed a path from the place where Ramirez’s bicycle had been abandoned to the defendant as he stood in a group of about six black men.
K-9 Willow and Officer Rivera
Willow was a purebred bloodhound that had been acquired by the Pomona Police Department two years before the incident. Willow’s handler was Officer Cesar Rivera, who had been a K-9 handler for 14 years. Before Willow, Rivera had handled two Belgian Malinois, both trained in patrol work and narcotics detection. Rivera had in fact written the proposal for the police department to acquire a bloodhound. Further details concerning Rivera and Willow were described by the court in summarizing Rivera’s testimony:
“[The] Pomona Police Department obtained Willow from a foundation in Florida that has a training facility in Georgia. Willow had been trained to track for almost a year by a trainer at the foundation and had also been used in the field by law enforcement agencies before Pomona Police Department acquired her. In her work for the other law enforcement agencies, Willow had done approximately six or seven finds, i.e., where she actually tracked and located someone. Rivera trained with Willow at the Georgia facility for a week before bringing her back to Pomona. Rivera described the training, which involved the use of a training track and keeping a log of the dog's performance. After he brought Willow back to Pomona from Georgia, Rivera spent two to four hours a day for two months training her to do hard surface tracking before putting her in the field. Rivera kept logs and only put Willow out into the field when ‘she was very reliable where she actually stayed on the track and located the decoy.’”
It is clear the dog received training prior to being acquired and brought to California. The prior training and use of the dog by other law enforcement agencies should have been explored as it might have indicated problems which caused the other agencies to return the dog to the foundation.
Curiously, Rivera apparently testified that there is no test or certification for bloodhounds. This is incorrect since a number of organizations have testing and certification programs for tracking dogs. See, e.g., the National Police Bloodhound Association. The intent may have been to state that the police department itself had no formal program. Rivera acknowledged that Willow had bad days—he said about 5% of the time—when she did not perform well. Rivera provided Willow’s training logs to the defense.
Rivera described how Willow works:
“Rivera described for the court how he gets Willow to track, a process that involves taking her to where the subject was last seen, having her sniff an object that contains the subject's scent, and then telling her to ‘go find him.’ Willow then sniffs around, finds the direction of travel, and ‘once she finds the direction of travel, then she's on the go, she's tracking.’ Rivera trains Willow every day he works with her, which is about 16 hours each month, and once a year he goes back to Florida to meet with other handlers and exchange information.”
It was not clear what was in the records Rivera kept of these daily training sessions. If they involved tracking of volunteers, more information about the team’s skills could have been obtained from them.
Willow at the Crime Scene
Rivera took Willow to Ramirez’s bicycle, which had been left near a church. Rivera had Willow sniff Ramirez so as to eliminate him as a tracking subject, a procedure called “the missing member,” which is described in Police and Military Dogs, Chapter 5: Scenting to Items or Locations with Odors of Multiple Individuals. (We note that one expert described the procedure as “not wholly reliable”.)
The court described Willow’s actions in the investigation as follows:
“Willow sniffed various parts of the bike, and then proceeded on her own track. Willow followed her nose through a hole in the brick wall between the alley and the church, down the alley to Parkside, through Stone's apartment complex, and straight to six or seven men who were lined up in the middle of the street, shoulder to shoulder. After sniffing each of the men, Willow sat down in front of defendant, which Rivera explained means she had matched the scent she picked up from the bike with defendant. Officer Rivera walked Willow away from the area briefly while the men moved to stand in different positions. Rivera then returned with Willow and gave her the command to go find him. Willow again sniffed each person and alerted on defendant by sitting down behind him.”
It is to be noted that Willow had not followed Stanford's path exactly, since he had gone into the building where he took a shower and changed his clothes.
Optimal scent lineup design often specifies that different foils should be used in separate lineups, since a situation such as the one described here, occurring moments after the initial identification, is likely to produce the same result, perhaps because of some attractiveness of the person alerted to that may be unrelated to the scent of the trail.
No mention was made of research that has indicated that dogs may be more accurate when following a trail than when identifying an individual at the end of the trail. In an experiment conducted near Phoenix, dogs followed the correct trail from a test site where an improvised explosive device was detonated to the area where the person who “set” the device was standing, but in several trials the dogs then identified the wrong individual. (See Police and Military Dogs, Chapter 5: Is Trailing More Successful than Identification?; also see prior blog regarding the research.)
Court Rejects Application of Standard for Admission of Scientific Evidence
The defense attacked the introduction of the canine testimony by asserting that it did not meet requisite scientific standards under California v. Kelly, 17 Cal.3d 24 (1976) (California adheres to the somewhat more difficult but now superseded federal standard enunciated in Frye v. U.S., 293 F. 1013 (D.C.Cir. 1923), rather than the looser standard adopted by the U.S. Supreme Court in 1993 in Daubert v. Dow Pharmaceuticals, 509 U.S. 579, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993).) The court rejected the argument, as California courts have before, saying that dog tracking evidence did not involve a scientific technique subject to Kelly. The California appellate court acknowledged that scientific validity might have been required had the police used a scent transfer unit to place the scent given the dog on a pad. Although not a lineup, the court states to “the process by which Willow identified defendant involved tracking or trailing,” meaning that the selection of Stanford in the group fits within a procedure sometimes referred to as a “station identification.”
We have previously described a tracking case where tracking law was strained to cover a station identification by a California Court, essentially tracking law to a poorly constructed scent lineup. The court here perhaps assumed that an identification based on all the scents emanating from an individual at the end of the track would be more accurate than an identification based on matching specific scents left on objects that a criminal may have held. Yet even in highly controlled scent lineups, experienced dogs can pick a foil over the target. The risks are even greater with a lineup of people as opposed to a scent-matching procedure. Generally speaking, European scent lineup technology has long rejected the use of lineups of individuals. As noted in Police and Military Dogs, one Dutch researcher writing as far back as 1936 stated:
“It is still more dangerous to let a dog choose a person out of a row of people on the score of the odour of an object offered to the dog. Even if the trainer knows nothing about the test, and has himself no suspicion of anybody, there still remains the possibility, by no means a light one, that the dog may respond to the faintest movement of one of those persons.”
In other words, cueing inducing an alert could come from behavior exhibited by the handler, the suspect, or another person present and visible to the dog.
Further, the description of the tracking does not specify the length of the track involved, which has been found to correlate with the accuracy of an identification at the end of the track. (See Chapter 4: Length of the Trail, noting that the longer the trail, the more likely the dog would not follow a decoy who had been walking with the target when the two people separated and left separate trails.)
Foundation for Admission of Tracking Evidence Deemed Satisfied
The court here cited the necessary elements for tracking or trailing evidence to be admissible:
(1) the dog's handler was qualified by training and experience to use the dog;
(2) the dog was adequately trained in tracking humans;
(3) the dog has been found to be reliable in tracking humans;
(4) the dog was placed on the track where circumstances indicated the guilty party to have been; and
(5) the trail had not become stale or contaminated.
Since the court determined that this foundation had been established, “the trial court did not abuse its discretion by admitting evidence that Willow tracked defendant from the bicycle to where defendant stood in a group of men.” It is not clear whether Rivera knew who Stanford was. Although it might be assumed based on the court’s description of when he and the dog arrived that no one had identified Stanford to him, or described Stanford in sufficient detail for him to make the identification himself, it is not clear that this issue was investigated in cross-examination or otherwise. If Rivera had knowledge that could have allowed him to deduce who Stanford was in the group, the risk of cueing by the handler is significant.
Reliability of the Canine Team
Defense counsel attacked Willow’s reliability:
“[D]efendant first cites the purported absence of a log documenting Willow's success rate. In defendant's view, the ‘hundreds of pages of reports’ Officer Rivera produced did not constitute a ‘log’ and therefore did not establish Willow's reliability in tracking humans. Defendant concedes that Officer Rivera testified to Willow's reliability, i.e., she is on the mark 95 percent of the time. If, as defendant contends, that figure was a “guestimate” on the part of the officer, that is a circumstance that goes to the weight the jury should give the tracking evidence, but it does not affect the admissibility of the evidence.”
It is not clear if the hundreds of reports could have been distilled by defense counsel into intelligible evidence, or if counsel even attempted to do so. It may be supposed that this was either impossible or not done since an accurate success rate would have obviated the need for a “guestimate” and allowed inquiry into questions such as how many trails went cold and how often the dog alerted when encountering an individual at the end of the trail. If the materials could not answer such questions, it must be questioned whether reliability was in fact established.
The defense also argued that the tracking evidence was inadmissible because Rivera had no formal training. This too the court rejected:
“Although defendant takes issue with the extent of that training, he does not identify the specific deficiencies or recount the additional training Officer Rivera should have had in order for the tracking evidence to be admissible. Defendant also claims that the absence of licensing or certification affects the validity of Officer Rivera's testimony regarding Willow's reliability because Willow's performance has not been independently evaluated. Defendant's claim assumes that if a licensing organization existed, it would require independent assessment. The claimed deficiencies, if any, go to the weight of the evidence and not whether it is relevant and therefore admissible.”
Again, the implication is that independent testing is unavailable, which is not the case. Also, it seems as though the burden of establishing the dog’s reliability did not lie with the prosecution, but rather that the defense had to establish that the dog was not reliable. (In Harris, which the U.S. Supreme Court will consider, the State of Florida asserts in its appeal that the court placed “an excessive evidentiary burden” on the prosecution, by requiring the production of a dog’s field activity reports. Curiously, the State of Florida’s brief in Harris notes that “California courts do not require evidence of a dog’s success rate in the field to establish probable cause.”)
The defense apparently argued that Willow had been used to find lost children and that this was not the same as tracking a criminal. More needs to be known about the team’s work to judge the level of her tracking skills in situations such as was presented during the investigation here.
The defense was grasping at straws in arguing that there was no other evidence that the defendant had gone through the brick wall. The prosecution was required to demonstrate that the dog was placed on the track where evidence indicates the target had been—here, the place where the bicycle was abandoned. It is not necessary to show that every place along the trail followed could be established by other evidence than the path the dog took. That the dog went to the group of men rather than to the apartment where the defendant had been for some time may indicate that the defendant had created a cross-track somewhere outside the apartment, which the dog followed as having the stronger or most recent scent, but as with a hunting dog, a tracking dog may follow the stag after it crosses over its initial trail. In a station identification, the tracking dog usually resumes tracking at the station because the scent it has been given is again detected and the dog follows the scent to an individual and alerts as an identification.
Having rejected other challenges, the court affirmed the murder conviction.
Science will out in the end, though it may take a long time. We are not arguing that a Frye or Daubert hearing should be held to admit tracking evidence, but we are saying that the foundation should be established before the jury hears about the tracking and that scientific research can inform the analysis of the foundational requirements. It is not clear from the court’s description if the dog was adequately trained or if it is a reliable tracker. Thus, at least two out of the five foundational elements are in serious question. Also, the dog’s reliability in scent identification was not even alluded to and the dog’s alert, though generally admissible under American law, should have been acknowledged as perhaps due to other factors than a matching of scents.
When the U.S. Supreme Court hears Harris this fall, it may consider why training and testing records for some drug dogs demonstrate much higher accuracy than field records. The mantra that too many handlers repeat about residual odor cannot hide the fact that U.S. Supreme Court precedent has given narcotics canine handlers the power to support searches of luggage with minimal and cars without reasonable suspicion, providing a means of acting on a hunch. Similarly, poorly designed and implemented scent identification procedures provide an identification that can be introduced with minimal corroboration. If that identification occurs under a procedure that does not have a high probability of producing an accurate result, as has happened with almost all judicially described American scent lineups, then a court which approves such a procedure provides a fallback for getting a conviction when other evidence is lacking or disappears before it can be presented to the jury.
This case amounts to a sort of station identification, though it did not take place in a police station. It would not have been acceptable evidence in the Netherlands, where even much more rigorously controlled and scientifically supportable scent identifications are no longer deemed acceptable evidence. (For the history of this, see Ruud Haak, K9 Fraud!) Nevertheless, with requirements of corroboration, and cautions to juries about limited value, such identifications have a long tradition of being admissible in American courts.
There was a good deal of eyewitness testimony in the prosecution of Kevin Tyron Stanford and the canine identification, even if improperly admitted, was only one of a number of identifications and could probably be termed harmless error. Therefore, though not without serious flaws, and certainly not elegant, the decision probably reached the correct result.
Regardless of how the Supreme Court rules in Harris, its thinking regarding training, certification, the keeping of field records, and the placement of evidentiary burdens in canine narcotics detection cases, will only be dictum for a court considering facts like those presented in this California case. Nevertheless, the high Court’s reasoning will provide grist for judicial mills across the country and a decision such as California v. Stanford may not be made under precisely the same logic in the future.
California v. Stanford, 2012 WL 1365744 (Ct. App. 2012)
This blog was written by John Ensminger and L.E. Papet.