(Story originally created Nov 11, 2014)
NASHVILLE, Tenn. -- An exclusive NewsChannel 5 investigation has uncovered serious questions about the drug dogs that police use to justify searches along Tennessee's interstates.
We discovered those dogs may not be as reliable as some think and, as a result, innocent people could end up being treated like common criminals.
The latest twist in our "Policing for Profit" investigation came after a California police officer and his wife contacted NewsChannel 5 Investigates about their experience in Middle Tennessee -- an experience with officers looking for suspected drug money that they can seize to fund their agency.
"It just happened so fast and I was upset and I was tired," said Lisa Hankins.
Hankins and her husband Ronnie were on their way back home to San Diego in May, after a cross-country trip for a family funeral, when they got stopped in Dickson County by an interstate interdiction agent with the 23rd Judicial District Drug Task Force, supposedly for a traffic violation.
"I got upset when he took my wife out of the car," Ronnie recalled. "There's no reason to take her out of the car."
Lisa refused to give permission for agents to search her car.
"I said I haven't done anything for you to search my car," she remembered.
"And that was not the answer he wanted to hear?" NewsChannel 5 Investigates asked.
"Nope," she said. "I can tell you he was getting upset with me."
So agents brought out a drug dog.
Within seconds, the dog sat on the driver's side. That's how it's supposed to show it's picked up the smell of drugs.
The Hankins no longer had a choice about a search.
As a federal police officer with the Marine Corps, Ronnie immediately had his suspicions.
"He had already made up his mind what he was going to do," Ronnie insisted. "100 percent -- there is no doubt in my mind -- they cued that dog." (Related: I-40 Search Raises New 'Policing For Profit' Questions )
NewsChannel 5 Investigates showed dashcam video from the stop to Auburn veterinary medicine professor Dr. Lawrence Myers. He's a nationally recognized expert on how detection dogs work.
The first thing that Myers noticed is that there's no microphone on the handler, raising all sorts of questions.
"Is the handler giving any audible cues to the dog or just simply saying sit?" he asked.
We followed up, "So in this case we don't hear what he's saying to the dog?"
"Not a clue," he responded.
Myers also noted that the officer keeps his hands to himself on the passenger side, but -- as he rounds the car -- that suddenly changes.
"The officer pointed at the spot where the dog eventually sat before the dog sat," Myers said. "That's actually a cue."
"Intentional?" we asked.
"I don't think so."
Myers said it could be a training issue.
"Dogs take their cues from people. 'I must go investigate: what is he pointing at?' Then having stopped and paused and the whole thing, the officer then releases tension on the leash. The dog goes, hmm, sit."
In fact, the search did not turn up drugs or anything illegal.
Our investigation also uncovered another traffic stop where one interdiction agent from the 23rd led his dog around a vehicle three times without getting an alert.
Instead of accepting the result, he blamed the dog's arthritis.
That's when another officer brought out his dog. After two rounds, he got an alert -- finally giving officers a reason to search.
"One of them is definitely wrong -- which one?" Myers asked.
In fact, a thorough search of the truck found no evidence of drugs or anything illegal.
"The dog doesn't alert, so they assume there must be something wrong with the dog," NewsChannel 5 Investigates noted.
"Right," Myers said, "and when the dog alerts and nothing is found, there must be some reason that it did alert."
A few years back, researchers from the University of California-Davis set up a test inside a church , supposedly to see how well drug and explosive detection teams could find contraband that had been hidden inside.
The dogs hit 225 times.
But there was a catch.
There were no drugs or explosives anywhere, suggesting the dogs can sometimes react to what they sense from their handlers -- instead of what they actually smell.
In fact, a review of reports kept by the 23rd shows that -- out of 24 alerts this year -- in 15 cases, officers did not come up with any contraband that they could seize.
Still, those reports show officers wrote up why they thought they got the alerts.
One car was a "California rental."
In another case, "a friend could have had marijuana the night before."
Or officers spotted what looked like "marijuana shake" -- or residue -- in the carpet.
There's no evidence that officers ever tested what they thought might be residue to confirm their suspicions.
"That's wishful thinking," Myers said about the K-9 reports.
"The dog could be wrong?" we asked.
"Oh, absolutely, dogs are frequently wrong," the professor insisted. "It's just a fact of life. They're an organism."
"But they don't want to put that down in their report," we noted.
Myers answered, "And I don't know why."
Ronnie Hankins insisted "there is no reason that they should have violated our civil rights against search and seizure. They shouldn't have even stopped us. There was no reason to stop us."
He said that -- as a police officer -- he knew his rights, but that still didn't protect him from what he believes was an illegal search.
"If this happened to you, what does that say to you about the constitutional rights of people who pass through this area every day?" we asked.
"There is no telling how they get violated."
As a result of the election of a new district attorney, there's new leadership at the 23rd Drug Task Force, and those officials are reacting to our NewsChannel 5 investigation.
They are putting in place strict new requirements about videotaping such searches.
If officers don't like the results they get from one drug dog, like we saw in that video, they will not be allowed to use a second one.
And any seizures of cash or property must now be approved by the drug task force director.
As for why we don't have audio from the dog handler in the Hankins case, his audio should have been recorded on a camera in his car.
But NewsChannel 5 Investigates asked for video from that camera, and we were told that, for reasons no one could explain, it does not exist.