Tuesday, October 21, 2014

Deputies Stand by as Restaurant Excludes a Service Dog; Sheriff's Department May Be Liable

People with service animals still encounter resistance to bringing their animals into places of public accommodation, such as restaurants.  They rightfully expect to be able to do this without difficulties but, if an issue arises, they hope that the police, if summoned, will at least be able to explain some of the rudiments of disability law to the employees of the business involved. 

Nevertheless, a number of cases suggest that some law enforcement agencies remain ignorant of the rights of individuals with disabilities who use service animals, and that this ignorance has too often been forgiven by courts.  A recent decision from the federal district court for the Western District of Louisiana, however, suggests that courts may not continue to shield officers who unthinkingly help a business bar or eject an individual with a service animal. 

Before getting to the specifics of the recent decision and three prior decisions that were completely unsuccessful in this regard, in the manner of a law school class let us pose a series of hypotheticals.  The following basic facts apply to each of the hypotheticals: 

You have a service dog—say, a seizure-alert dog that generally is able to alert to an oncoming seizure several minutes before you have one.  The dog is also trained to stay beside you and provide something for you to hold onto so that you can sit or lie down without injuring yourself during the first minutes of a seizure.  The dog has a vest but you do not always put the vest on the dog.  You do, however, keep some documents regarding the dog’s function with you at all times.  

Also assume that you are a member of a neighborhood watch team and one night a week patrol the neighborhood with your dog and another member of the team.  The neighborhood watch team has a meeting once a month.  Formerly the team met in houses of the members but it has become too large and a decision is made to move the meeting to a restaurant.  A diner nearby is chosen because it is convenient, can be walked to, and agrees to host the meeting and provide donuts for $50 if the meeting is held on a Monday night after 8 p.m., when business is slow in any case.  You have only been in the diner twice, but neither time did you have your dog with you. 

You take your dog to the neighborhood watch meeting.  The dog is not wearing a vest.  Outside the diner, the manager, who is greeting members of the watch team as they arrive, sees your dog and says that you cannot bring it into the restaurant.  You explain the dog is a service dog and pull the documentation regarding the dog’s function from a briefcase you are holding.  The manager refuses to look at the papers.  Several other members of the team take your side but the manager insists that the diner only admits guide dogs. 

Now, consider the following four hypothetical variations on what happens next:
  1. Because you are a member of the neighborhood watch, you have the local police station on speed dial and call the station to request assistance because your rights are being violated.  In five minutes a patrol car arrives.  The officer knows you and explains to the manager that your dog should be allowed the same access as would be given to a guide dog.  The manager calls up the owner of the diner, who backs up the manager’s decision to exclude the dog.  The manager remains adamant that you cannot bring the dog in.  The officer explains to you that he cannot force the manager to admit the dog, but that you have rights and can file a complaint in the matter.  You go home.
  2. The facts are the same except that the officer who responds to your call is not known to you and talks to the manager but does not attempt to convince him that he should admit both you and your dog to the diner.  The officer tells you that he can do nothing for you, that the diner has the right to refuse service to anyone, and that you must leave.  If you attempt to enter the restaurant the officer says he will have to arrest you and impound your dog.
  3. The facts are the same except that when you get to the diner and begin arguing with the manager, you do not need to call the police station because there is a police car in the lot and the officer in the car is eating a sandwich.  Although he sees your interaction with the manager, he says he will talk to you after he finishes his lunch break.  He makes a call to his girlfriend and talks to her for nearly twenty minutes before getting out of his car and listening to you.  You do not know this officer but he clearly does not know anything about the law of service dogs.  He says he will call his chief, as he has been instructed to do when situations arise with which he is not familiar.  By the time the chief calls him back you have been waiting for nearly an hour and the meeting has gone on without you. The manager locked the door so when the officer knocks the manager has to unlock it. The officer explains to the manager that his chief has told him that you are entitled to bring the dog into the meeting, but the manager still refuses.  The meeting is now almost over and you go home.
  4. The facts are the same except when you get to the diner there are two policemen outside the restaurant talking to members of the watch.  One of the officers is going to speak at the meeting and the other has accompanied him for moral support.  They have come as a courtesy to the neighborhood watch group.  When the manager tells you that you cannot bring your dog into the meeting, the officers see what is going on and you ask for their assistance.  They tell you that they cannot help you and that your problem is a federal matter on which they have no jurisdiction.  They go into the diner without speaking to the manager.  You and your dog walk home.
There seems to me to be enough of a moral failing on the part of the police in the last three hypotheticals that some level of liability should attach.  Nevertheless, decisions from which the hypotheticals were adapted (with a degree of poetic license) indicate that only the fourth situation might lead to any liability or even a serious reprimand from a court after the presentation of the case. 

Let us review the decisions from which I drew these hypotheticals. 

Pizzeria Ejects Patron with Service Dog

Pona v. Cecil Whittaker’s, Inc., 155 F.3d 1034 (8th Cir.1998) 

The employees of Cecil Whittaker’s Pizzeria asked Marilyn Pona, who suffered from degenerative spine and joint disease, to leave a pizzeria because she had a dog with her.  Pona alleged that the officers responding to the scene refused to explain the law regarding service dogs to the restaurant, but did inform her of her remedies, and asked her to leave the premises. 

Pona filed claims under the Americans with Disabilities Act, 42 U.S.C. 1983 (providing a civil action for a deprivation of rights), and the Missouri Human Rights Act.  She asserted that the police refusal to assist her in gaining access to the restaurant was due to a formal policy (Special Order 86-S-31) which directed police officers to take no enforcement action with regard to the Missouri statute (209.150) that gave her the right to be accompanied by a service dog in a place of public accommodation.

Cecil Whittacker’s Inc. moved for summary judgment on the basis that as the franchisor the pizzeria, it could not be liable under the ADA because it did not own, lease, or operate a place of public accommodation (as required to come under 42 U.S.C. 12182(a)).  The district court agreed.  The manager of the franchise had called Donald Glenn, president of CW, the franchisor, who told the manager that he “wouldn’t have any animals in [his] restaurant” because it “doesn’t look good for the franchise.”  Glenn denied making such a statement, but for summary judgment purposes it was assumed that he did.  The court said that such a statement by Glenn would not alter the fact that the franchisor had no control over the franchise in this regard and was at most giving advice. 

There were three judges and three opinions in this Eighth Circuit case.  All three judges agreed that the franchisor could not be held liable because of its lack of control of the franchise, even if an executive had given some advice to the franchisee’s management.  Two of the judges found that the City of St. Louis Board of Police Commissioners did not have a policy that effectively withheld enforcement services from disabled citizens.  There were some policy documents, but there were complex questions as to the authority under which they were issued and whether they applied to the particular situation involving Pona and the restaurant. Judge R.S. Arnold, however, disagreed with his colleagues and said there was in fact a policy that effectively withheld law enforcement services from disabled citizens. 

Judge M.S. Arnold held that because the ADA claims failed, a 1983 violation for deprivation of federal rights based on such claims must fail.  Also, an ADA violation is not actionable under 1983 because it must be presumed that the enforcement provisions of the ADA are the exclusive mechanism for enforcing the ADA.  He did state, however, that “I intimate no view on the question of whether Title III violations can ever form the basis for a cause of action under § 1983.” Judge Panner said more generally that ADA Title II and III claims are not cognizable as 1983 claims.  Judge R.S. Arnold did not mention 1983. 

Pona also argued that the St. Louis police officers violated the Missouri Human Rights Act, which makes it unlawful to deny anyone accommodations provided by any place of public accommodation on the grounds of handicap.  Judge M.S. Arnold doubted “the facts would support a finding that Ms. Pona was denied service at the pizzeria ‘on the grounds of … handicap.’ It was her dog, not Ms. Pona herself, to which the pizzeria raised objection.”  Judge Panner disagreed, saying that denial of services on account of a service dog was denial of services on account of a disability, but without further explanation agreed that the state law claim should be dismissed.  This could be because the franchisor was not in sufficient control to be liable or because no federal claim survived. 

Would anyone seriously argue that telling a person in a wheelchair not to bring it in a restaurant, though the person needing the wheelchair could not come in without it, was not a denial based on a disability? 

Popeye’s Manager Refuses Service to Customer with Hypoglycemia Alert Dog

Gipson v. Popeye’s Chicken & Biscuits, 942 F.Supp.2d 1303 (N.D. Ga. 2013)

Taylor Gipson has a dog named Bear that can detect his blood sugar level gets too high or too low.  On May 12, 2012, Bear alerted Gipson in a manner that indicated to him that his blood sugar was low and he entered a Popeye’s restaurant to order food, which would have corrected this condition.  He took a table near the back door to wait for his food while Bear lay quietly on the floor next to him. 

Shanika Parks, manager of the restaurant, came to Gipson’s table and asked if Bear was a seeing-eye dog.  Gipson replied that Bear was not and Parks demanded that he leave the restaurant with the dog.  Gipson tried to tell Parks that he was entitled to keep a service dog, even though it was not a guide dog, with him under federal law.  According to the Georgia federal district court (employing some passages from Gipson’s complaint):

“Parks ‘became agitated’ when Plaintiff stated he had a right under the Americans with Disabilities Act to remain in the restaurant with Bear… Her ‘agitation escalated’ when Plaintiff would not leave…. Parks stated that Plaintiff was ‘costing her customers’ and ‘demanded’ that Plaintiff and Bear ‘get out of her restaurant.’ … Despite Parks’s ‘escalating hostility and agitation,’ Plaintiff ‘calmly’ explained to Parks why he was legally entitled to have Bear in the restaurant.”

Parks then threatened to call the Cobb County Police Department to remove Gipson and Bear.  Gipson again refused to leave and explained that he had a right to have Bear in the restaurant.  Parks made good on her threat and called the police, as did Gipson, who assumed that the police would explain his rights to Parks. 

Before the Cobb County Police arrived at the Popeye’s restaurant, another customer approached Parks and told her that Gipson had a right to have a service dog in the restaurant.  The customer even apologized to Gipson for Parks’ behavior. It is at least a small comfort that the public is becoming aware of the rights of users of service animals, apparently somewhat more than some employees of national businesses and some police officers. Popeye’s restaurants, according to its answer to the complaint, have sound policies with regard to service animals, even if some employees do not know about them.

Anyone who has worked in disability law, or civil rights law, would assume on the facts that the police, by this time in the history of American law, would arrive and tell the manager that for nearly a generation service dogs have included many types of dogs besides guide dogs, as recently happened at a restaurant near Boston. Continuing with the court’s narrative of events:

“When the police arrived, Parks quickly went outside to meet the officer and loudly demanded that Plaintiff and Bear be ordered off the property or removed by force…. Plaintiff’s mother arrived on the scene to hear Parks’s conversation with the police…. Plaintiff also then spoke with Officer Fuller and explained why he believed under the Americans with Disabilities Act that he and Bear had a right to be in the restaurant…. Plaintiff’s mother offered Officer Fuller a card which explained Plaintiff’s rights, but Officer Fuller declined to look at it… Officer Fuller stopped Plaintiff from speaking and said he ‘knew the law.’  Officer Fuller spoke with Parks again and she again stated that Plaintiff had to leave the property…. Officer Fuller explained to Plaintiff that because the restaurant was ‘private property,’ Plaintiff and Bear were trespassing and had to leave immediately…. Although Plaintiff again tried to explain his rights, Plaintiff was eventually ‘forced to comply’ with Parks’s demand that he leave the restaurant.”

Narrative Portion of Incident Report in Popeye's Dispute
Officer Fuller’s incident report reflects that Parks was aware of Gipson’s contention that the dog was a service dog, but also that she believed that several customers walked inside the business and then left because of the presence of the dog. The fact that a service animal might be disturbing to another guest is not a reason for barring access to a place of public accommodation unless the dog is out of control (which probably would indicate it is not a service animal).  The incident report states that Bear was wearing a service dog vest.  The vest does not prove that Bear was a service dog under federal law (as the Department of Justice has indicated in regulatory releases), but the ability of the dog to alert to hypo- or hyperglycemia does suggest that the dog very likely qualified as a service dog. 

Cobb County argued that Gipson was not denied any service by Officer Fuller, who was responding to the calls both from Parks and Gipson.  Officer Fuller informed Gipson that because the restaurant was private property and because the manager wanted Gipson to leave, he would need to leave.  The County argued that Fuller’s actions did not deny “services, programs, or activities” of a public entity under Title II of the ADA, which applies to state and local governments.  The court cited Bledsoe v. Palm Beach County Soil & Water Conservation District, 133 F.3d 816 (11th Cir. 1998) as providing that discrimination by a public entity need not be limited to “services, programs, or activities” of the entity but to all discrimination by the entity.  The district court then analyzed whether there had been discrimination in Officer Fuller’s handing of the matter:

“[T]he court focuses on whether Plaintiff has alleged sufficient facts to show that the exclusion, denial of benefit, or discrimination was by reason of Plaintiff's disability. The court finds that Plaintiff has not. Based on the facts alleged in the complaint, Officer Fuller responded to the scene and listened to the position of both sides. The restaurant manager informed Officer Fuller that she wanted Plaintiff to leave because Plaintiff's service dog was scaring away other patrons. Plaintiff and his mother told Officer Fuller that Plaintiff had the right to be accompanied in the restaurant by his service dog. Officer Fuller determined that the restaurant was private property and the restaurant manager could ask Plaintiff and his service dog to leave. Officer Fuller, therefore, did provide services to both Plaintiff and the restaurant manager. Plaintiff disagrees with the outcome of those services, but there is no doubt that Officer Fuller responded to the scene and attempted to resolve the conflict. There can be no expectation that the police will always resolve a conflict in one's favor and Plaintiff has alleged no facts which would show that Officer Fuller's determination that the restaurant manager could ask Plaintiff and his service dog to leave the private property was one he reached on the basis of Plaintiff's disability.”

The court then makes a highly questionable analogy:

“If the court were to determine that Plaintiff was denied services based on his disability because Officer Fuller did not convince the restaurant manager that Plaintiff and his service dog could remain in the restaurant, the police would become responsible for sorting out civil liabilities. While one might argue that whether a service dog is permitted in a restaurant is a fairly straight-forward question (and one that the court will need to address with respect to Popeye's liability in this civil action), another patron might challenge the degree of slope of a handicapped ramp into the restaurant, a much more difficult question to resolve on the scene. There can be no expectation that police officers are equipped to address that type of situation when responding to a disturbance call. County police officers are not civil lawyers. Plaintiff has not alleged that the officer violated his constitutional rights. Most significantly, Plaintiff is not left without a remedy; he has sued Popeye's Restaurant to enforce his rights under the Americans with Disabilities Act.”

Listening to a store manager without remarking that her position likely violates a customer’s rights, then telling the customer with a service dog to leave, is to be distinguished from a situation where an officer attempts to inform the store manager that her actions could lead to liability for both herself and her employer, and cautioning the manager that the officer expects he may be called to testify if the customer files suit.  Explaining remedies to a person with a wheelchair might be the only appropriate action where the slope of a ramp is too great, because nothing else would be possible at the moment.  Here, more of a dialogue initiated by the officer appears to have been both possible and desirable. 

Courthouse Staff Delays Admission of Witness with Service Dog

Sears v. Bradley County Government, 821 F.Supp.2d 987 (ED Tenn., 2011) 

Brenda Sears has a seizure disorder and uses a service dog named O’Neal.  She went with the dog to the Bradley County Criminal Justice Center on August 11, 2009, to testify for a friend’s son.  Sergeant Brown, operating the metal detector and security check at the courthouse, informed Sears that dogs were not allowed in the courtroom, to which she replied that the dog was a service animal.  It took 20 minutes before Brown spoke with a court officer and showed him papers that had been given him by Sears’ husband regarding service animals.  The papers were passed onto the judge in the case, who said he would look at them after lunch.  The judge eventually said Sears could enter the courtroom with the dog. 

During the lunch hour that the judge took, Sears stated that she was required to remain standing because there was no chair for her to use and that she was not allowed to use a restroom in the courthouse, forcing her to use one in a Taco Bell nearby. 

Sears filed claims under the ADA, Title II, 42 U.S.C. 1983, and various state law torts.  The Tennessee federal district court dismissed the ADA complaint against Brown in his individual capacity, finding no individual liability under Title II of the ADA.  Brown was granted summary judgment as to the complaint against him in his official capacity as, according to the court, the facts did “support an inference Sergeant Brown intentionally discriminated against Plaintiff on account of her disability.”  This, of course, raises the question already mentioned as to whether refusing to allow an individual with a disability to use a dog can amount to discrimination against a person because of a disability. The court said Brown was entitled to qualified immunity on the 1983 claim. 

As to the county, “[f]ailure to supervise is not a viable theory for recovery of compensatory damages in a Title II ADA claim, since such failure is necessarily not directed at a particular disabled individual.”  Failure to train Sergeant Brown was also “not a specific act of intentional discrimination against the Plaintiff herself….”  The county said that “Sergeant Brown's conduct towards Plaintiff was not motivated by discriminatory intent but his genuine bewilderment at how to handle service animals. Moreover, Sergeant Brown's efforts to seek approval from Judge Randolph, including his forwarding of the papers provided by Plaintiff's husband to Judge Randolph's court officer, indicate he was not attempting to discriminate against Plaintiff.”  The claims against the county also failed as the court did not believe Sears would be able to establish “an unconstitutional policy or custom, nor actionable failure to train or supervise its officers….”  

As to the 1983 claim, the county argued that it was entitled to summary judgment “because, setting aside the question of whether Plaintiff even suffered a constitutionally-cognizable injury, Plaintiff cannot show the County had an official unconstitutional policy or custom which was responsible for the injury, or that the County was ‘deliberately indifferent’ to the rights of Plaintiff and similarly-situated individuals."  The court agreed, and noted that after the incident, “Bradley County has adopted a policy and held a training session addressing the needs of disabled individuals with service animals.” This subsequent effort to educate the County’s law enforcement personnel turned out to be significant in the case discussed next (where no such effort was made despite a previous incident). 

In other words, delay and indifference do not amount to constitutional violations, at least when the subsequent admission of the service dog can be said to preclude any proof that there was intentional discrimination against someone with a disability using a service dog. 

(For an earlier case on similar facts, which was discussed in the opinion in Sears, see Valder v. City of Grand Forks, 213 F.R.D. 491 (2003). Also, as my legal colleagues will be quick to point out if I don’t, the preamble to Department of Justice final regulations regarding access to Title II facilities specifically mentions courthouses.  75 Fed. Reg. 56192 (September 15, 2010).)

Former Police Officer with Service Dog Barred from Neighborhood Watch Meeting

Albright v. Sheriff’s Department of Rapides Parish, No. 12-2117, 2014 US Dist LEXIS 132946 (2014)

David Albright has two dogs that, in a complaint filed in a Louisiana federal district court, he says are service dogs that alert to episodes of cataplexy and narcolepsy, which cause seizures and instantaneous sleep spells.  Albright says that the dogs alert five minutes in advance to impending episodes, which can occur three to four times a day.

On August 8, 2011, Albright went to the Sieper Junction Café in Rapides Parish to attend a Neighborhood Watch meeting, where members of the Parish Sheriff’s Office were to speak.  Albright brought one of his service dogs, Zoey.  The manager of the café, Teresa Cutts, blocked Albright from entering the café with his dog.  Albright produced documentation that Zoey was a service dog, but Cutts continued to bar his entrance. 

The Rapides Parish Sheriff’s Office responded to the dispute.  Deputy Gunter spoke separately to Albright and Cutts.  A second deputy, who was not identified, arrived and told Albright he had to leave.  Deputies Davis and Walters, who had come not because of the incident but to speak to the Neighborhood Watch group were nearby.  Albright, a former police officer who had known Davis and Walters professionally, found them and asked them to intervene.  They declined, saying it was not their job to enforce federal law. 

This was apparently not the first time that Albright had dealt with the Rapides Parish Sheriff’s Department on a service dog matter.  His pleadings indicate that he had been blocked from entering the Rapides Parish Courthouse by a deputy several years before, though in that case a senior deputy intervened and permitted Albright to bring the dog inside.  Albright had discussed the courthouse incident with Deputy Davis, who Albright says agreed with him that more training on the Americans with Disabilities Act and service dogs was needed for law enforcement personnel in the Parish.  No courses or training had been initiated by the time of the restaurant incident. 

The district court determined that the Rapides Parish Sheriff’s Office “is a public entity within the meaning of Title II of the ADA.”  To establish discrimination under the ADA, the court said that Albright had to establish that (1) he is qualified for protection under the statute, (2) he was “excluded from participation in, or being denied benefits of, services, programs, or activities for which the public entity is responsible, or is otherwise being discriminated against by the public entity,” and (3) the exclusion, denial of benefits, or discrimination was by reason of the disability. 

To receive compensatory damages for ADA violations, a plaintiff must show intentional discrimination.  Employers are vicariously liable for the discriminatory acts of their employees.  Title II, however, does not allow actions against a person in his individual capacity, only in his official capacity.  Therefore, insofar as Albright sued Deputies Walters and Davis in their individual capacities, their motion for summary judgment was granted by the district court. 

The public entity involved was the Rapides Parish Sheriff’s Office, and any official capacity suit could only be brought against Sheriff Hilton, the current sheriff of the Parish, not against former Sheriff Wagner.  Nor could an official capacity claim be filed against deputies Walters and Davis, and therefore claims against Walters, Davis, and Wagner were dismissed. 

As to the three ADA claim requirements enumerated above, Albright filed medical records on his diagnosis and treatment.  For purposes of the motion for summary judgment, this information was adequate for the court to accept that Albright was a qualified individual under the ADA.  Also for purposes of the motion, “the presence and participation of Deputies Walters and Davis on behalf of the sheriff’s office is a service, program or activity performed by a public entity, in this case, the Rapides Parish Sheriff’s Office.”

The most difficult of the three issues that had to be resolved to assign liability under the ADA was the third, whether the exclusion was due to a plaintiff’s disability.  Louisiana district courts are within the Fifth Circuit Court of Appeals.  The Eighth Circuit has determined that to place liability on a public entity, that entity must have had “a policy, practice, or custom in place discriminating against disabled individuals with service animals.”  (Paraphrasing Pona, discussed above.) 

The Louisiana federal district court distinguished the situation before it from the one in Gipson v. Popeye’s:

“We agree that whether a service dog is permitted in a restaurant is a straight-forward question of law, but disagree that an officer would be liable if he did not convince the restaurant manager to permit plaintiff to enter with his service animal. The instant case is not one in which a deputy valiantly defended the civil rights of the plaintiff only to be refused by the restaurant manager.  Rather, the officer did not educate the restaurant manager about the ADA nor did he inform Mr. Albright of his civil rights against the restaurant. We are troubled that Defendants may have intentionally discriminated against Mr. Albright and been deliberately indifferent to his civil rights. This case is also distinguishable [from prior cases] because Mr. Albright had a previous experience in which he was refused entry into a courthouse with his service animal and personally notified Deputy Davis, a Defendant in this matter. This is very different from Sears in which an individual with a service animal was delayed entry into a courthouse and the county responded by holding a training session to address the needs of disabled individuals with service animals and display the county's new service animal policy in public buildings…. Here on the other hand, the parties agree the officers receive no training about the ADA, and Defendants seems to have no intention of training their officers about the ADA.”

The court also emphasized that Sears was “factually distinguishable from the case at bar as it involved a delay as opposed to completely barring entry and the Court found the police officer did not intentionally discriminate against the disabled party.” (emphasis added)

Therefore, the case can now proceed on Albright’s ADA claim against the sheriff’s office.  Also, the court allows for the possibility of compensatory damages, saying that although intent is necessary for such damages, the refusal to take any action upon Albright’s approaching the officers may indicate intent. 
Motions for summary judgment on the other federal and state law claims against the officers were granted, including claims for deprivation of various constitutional rights under 42 U.S.C. 1983 (as duplicative of the ADA claim, tantamount to allowing Albright “two bites at the same apple”).  Motions for summary judgment under Louisiana state law claims were also granted the officers, including a Louisiana White Cane Law claim and a claim under Louisiana human rights statutes (because the restaurant manager, not the officers, had denied Albright’s entry into the restaurant with the dog).  Nevertheless, a significant part of the case withstood the defense efforts to stop the litigation, and it is to be hoped that a lesson will be taught and learned. 

Proliferation of Service Animal Types and Bogus Claims Make Enforcement Difficult

It must be acknowledged that some of the service dogs described here were accompanying individuals whose disabilities were likely not evident to the businesses and police involved.  This does not mean the dogs were not legitimate service animals under the ADA, but it does mean that some level of inquiry was appropriate.  Users of service dogs should be familiar with where the boundaries lie, as I have detailed in Service and Therapy Dogs in American Society and in prior blogs here.  The rights of those with disabilities as to specially trained dogs do not extend to dogs that do not meet regulatory standards provided by the Department of Justice or by case law in jurisdictions where courts have analyzed the ADA and other relevant statutes. 

The ADA does not provide the same protections to pets, companion animals, and emotional support animals (as opposed to psychiatric service dogs), though emotional support animals receive similar protection to service animals in certain transportation and housing contexts.  What cannot be tolerated, any more than blind prejudice against users of qualified service animals, is the increasing alacrity with which some people claim that their pets are service animals merely because they do not want the inconvenience of leaving the dog at home. 

A recent case demonstrates that users of bogus service animals can be a problem for law enforcement officers.  In Lerma v. California Exposition and State Fair Police, 2014 WL 28810 (ED Cal., 2014), Regina Lerma attempted “to bring a pet Cocker Spaniel puppy into an amusement park and pass it off as a trained service animal under the ADA.”  When California Exposition and State Fair Police Officer Siegrest “asked plaintiff what task the dog had been trained to perform, plaintiff responded by stating ‘all I have to tell you is it’s a service dog and I’m going to sue you.’  … When asked how she would handle the dog’s need to relieve itself or whether it was housebroken, she responded again that she was going to sue the officer.”  At a deposition in the case Lerma filed, she “admitted that her dog was not individually trained to perform any task for her…. [P]laintiff conceded at her deposition that she took the dog to the Park because she ‘needed the dog to be able to get through the day,’ to help her feel better, and because the children wanted to bring it there.”  The defendants’ motion for summary judgment was granted.

There may be cases where police do not take the action they should because they doubt a claimed service animal is legitimate. The more people make such inappropriate claims for their pets, the more difficulties they create for those whose dogs are necessary for their disabilities. The ADA is not a magic spell that anyone can cast to ward off inconvenient questions about a dog that does not qualify as a service animal.  
Is There Too Much of a Shield for Police Involvement in Service Animal Discrimination?

What is the current state of the law with regard to the potential liability of the police in responding to a complaint by someone with a service animal?  It appears that if the police engage in some dialogue with the parties, even without making any attempt to explain to the business that its refusal may be a violation of an individual’s civil rights, which could result in significant liability, there will very likely be no liability on the part of the police.  Nor need the police be in any hurry to respond so long as they eventually do so, though at some point delay will be tantamount to denial and it must be hoped that courts would draw a line. A failure on the part of a police authority to educate staff as to rights of individuals with service animals will add considerable force to plaintiff’s arguments.       

I began writing this blog after the Popeye’s case was issued several years ago, but wanted to find a way to argue that there could be recovery against police officers abetting denials of service to individuals with service animals.  The best I could find at the time was to take some of the language from the non-prevailing views of the three-judge panel in Pona, the 1998 case, and rather remote extrapolations from non-service dog disability cases.  My arguments at the time, however, struck several of the people I asked to comment and frankly even me as too close to one of those rants in a law review where a professor who has never practiced takes an impossible moral high ground that would never be accepted by a judge in a trial and might not even be seriously argued by a disability rights lawyer in an actual case.  Now Albright gives me hope that things may be changing and I have finally completed this blog.


It is too late in the history of American law on service animals for such things to be happening.  Everyone would be shocked if a restaurant refused to seat a customer based on race, and even more shocked should a police officer help the restaurant throw the customer out.  Failure on the part of police to understand the rights of individuals with service animals, and a lack of any educational effort at the departmental level, should deserve a severe reprimand and in some cases should result in liability. An exclusion of someone with a legitimate service animal from a restaurant is as much an exclusion on account of a disability as an exclusion of someone with a guide dog, a walker, or a wheelchair.

By allowing the police to participate in an exclusion from a public accommodation, courts permit the police to become agents of private enterprises that are denying rights guaranteed under the Americans with Disabilities Act.  This must stop. 


The above discussion takes a particular trajectory through four cases and does not discuss every issue by which plaintiffs have attempted to impose liability on officers and their supervising agencies. The following table fills in some of the gaps for those wishing this level of detail. Since there were three separate opinions in Pona v. Cecil Whittaker’s, the judges in that case are listed individually. 

Title II
42 U.S.C. 1983
Qualified Immunity
State Law Claims
Pona v. Cecil Whittaker’s
No recovery
No recovery
Not discussed
Dismissal of trial court affirmed
MS Arnold
No recovery (no policy)
Because ADA claim fails, 1983 claim based on it must fail; also ADA violation is not actionable under 1983 because it must be presumed that enforcement mechanism of ADA must be exclusive
Doubts facts would satisfy state law grounds, dismissal appropriate
No recovery (policy not targeted at persons with disabilities)
Agree with MS Arnold as ADA Title II and III claims are not cognizable as 1983 claims
Denial of services on account of service dog was denial of services on account of disability, but agrees that state law claim should be dismissed (without further explanation of latter point)
RS Arnold
Recovery possible (policy effectively withheld services from disabled citizens)
Not discussed
Not discussed
Gipson v. Popeye’s
No recovery (Title II services provided by response and attempt to resolve conflict)
Not in issue  
Not in issue
Not in issue
Sears v. Bradley County (defendant Sergeant Brown)
No recovery in individual capacity but could be heard as to official capacity, but intent could not be shown
Could have been liable had Sears shown she suffered violation of clearly established constitutional right of which a reasonable person would have known, but burden not met
Sergeant Brown entitled to qualified immunity on 1983 claim
State law claims dismissed without prejudice
Id. (defendant Bradley County)
Failure to supervise not valid theory under ADA
No recovery (no unconstitutional policy or custom, nor actionable failure to train or supervise)
Albright v. Rapides Parish
Recovery possible as to Parish (see text discussion for distinction between liability of officers and sheriff’s department)
Duplicative of ADA
Not discussed
Summary judgment granted officers

Thanks to Leigh Anne Novak for suggestions.

Wednesday, October 8, 2014

Does Therapy Work Stress Dogs? Should Therapy Dogs Be Allowed to Work Off Leash?

The benefits of therapy dogs for various types of populations, including hospital patients, students, children learning to read, nursing home residents, and others, have been well documented, but how this work affects the dogs themselves has been less studied.  Having made therapy dog visits with Chloe for over six years, I have known a number of teams that have stopped because the handler concluded a therapy dog was burning out.  One handler told me that she took her boxer out of service because the dog had an expression that she had not seen before when a child picked up a fold from the dog’s back.  “He was not happy.  I could not risk that it might go further.” 

Concern for stress in therapy dogs was expressed by James Serpell, Raymond Coppinger, and Aubrey Fine in 1999, and various research teams have been looking to identify and measure stress in dogs doing therapy work, with some studies being largely anecdotal, some behavioral, and some looking for chemical markers that might indicate whether therapy dog work is stressful, and when it is too stressful for the dogs.  Although levels of oxytocin and insulin have been studied in the search for chemical profiles of stress, the studies discussed here were particularly focused on measuring levels of cortisol.  These studies often included measurement of specific behaviors and generally—though not consistently—found that therapy dogs did not have higher cortisol levels while they were doing therapy dog work or had recently completed therapy assignments. 

Cortisol is a steroid hormone secreted naturally in response to numerous mental and physical stimuli, some but not all of which are negative and stressful.  Cortisol rises in response to sexual stimuli and activities such as hunting and guarding.  Also, since dogs have long been valued for waking a village at night at the approach of intruders, it is possible that being approached and petted by strangers, as happens in therapy work, may produce some discomfort and raise cortisol.  Short-term effects of elevated cortisol are useful in allowing an animal to regulate bodily responses to situations requiring energy and focus, but prolonged high levels can lead to stress-related diseases and have negative effects on an animal’s health.  (Glenk et al., 2014)  Therefore, research has begun to focus on a combination of behavioral indications and high cortisol levels to establish that a dog is under stress. 

Research Issues

A number of questions have been posed by the research teams involved in this work, which are listed here with pithy summaries of the answers reported so far.  After this summary of issues, I will attempt some additional elaboration on some important studies.
  1. Are cortisol levels higher while therapy dogs are working than when they are at home or in other environments?  Some studies say yes, but an increasing number have found no significant difference during or after therapy work.  It is not clear that a reconciliation of inconsistent results is yet possible or wise, but I will hazard some thoughts on why the disparate results may have been reached. 
  2. Do the length and frequency of therapy dog work sessions affect cortisol levels?  Most studies looked at dogs doing nearly identical regimens of therapy work, though there may be indications that particularly long and frequent assignments lead to elevated cortisol levels.
  3. Does providing breaks to therapy dogs during work sessions avoid cortisol levels elevating during work days?  A study that was designed to answer this specific question found that the breaks made no significant difference in cortisol levels between a group of dogs that got breaks and a control group that did not.  
  4. Does working on leash increase cortisol levels for therapy dogs as opposed to working off leash? A recent study found that working on leash resulted in higher cortisol levels. I will discuss policy implications. 
  5. Are dogs stressed when working with specific target populations?  There is some evidence that working with children under age 12 is stressful for dogs, but other groups, such as nursing home residents, Alzheimer’s patients, and drug addicts do not seem to increase canine cortisol levels.
  6. Does positive reinforcement as a training approach reduce stress in subsequent therapy work by dogs?  A number of studies emphasize the use of positive reinforcement in training dogs for therapy work, but this has not been isolated as a factor in identifying which dogs will be more stressed by the work.  Because of ethical considerations, I doubt that it will be. 
  7. Are particular canine behavioral patterns correlated with higher cortisol levels?  Since some studies did not find higher cortisol levels from therapy work by dogs, correlations could not be provided between behavioral patterns and higher cortisol levels.  Nevertheless, a number of research teams have proposed lists of behaviors that should be recorded when seeking to verify that higher cortisol levels indicate stress rather than arousal from some other stimulus.  A standard list of such behaviors should be developed so that studies from various research groups can be correlated.
Before looking at individual studies, it is appropriate to provide a summary of their approaches and most important results, which is attempted in the following table.  The lead authors and dates of publication are listed in the first row, followed by summary characteristics of the teams, their training, the facilities where they work and what therapeutic functions they have.  Because one paper in particular raises the question of whether use of a leash may be a significant stress factor, a row is provided for what information is available from each study in this respect.  Finally, a summary of the behavioral and cortisol results is provided. 

Figure 1. Characteristics of Dogs, Training, Facilities, Activities, and Results of Cortisol Studies

Haubenhofer & Kirchengast, 2006
Piva et al., 2008
King et al., 2011
Glenk et al., 2013
Ng et al, 2014
Glenk et al., 2014
Number of dogs/teams
27 selected (data from 21)
15, but one team eliminated because of atypical results
5 dogs 3-10 years old
15 (4 spayed)/3
(1 neutered)
1 female intact (sometimes in heat)
7/14 (all spayed)
8/6, all fixed or spayed
3 (2 fixed)/2 (1 spayed)
Teams members of Tiere als Therapie
Basic obedience commands
Certified by Therapy Dogs, Inc. (Cheyenne)
Austrian AAI organization, certified dogs had ≥ 1 year experience in mental healthcare facilities
Vet Pets, Therapy Dogs International, AAA sessions, averaging visits from an 1 to 10/ month
Certified in Italy for AAI work, all with ≥ 2 years working experience; all trained with positive reinforce-ment and averaging ≥ 1 visit/ month
Facility/ population
Various facilities (hospitals, rehabilitation centers, retirement homes, schools)
Alzheimer’s patients
Edward Hospital, Naperville, Ill., most patients 41-60 years old, but some children
8-10 patients
Students in common area of U. Penn. Dormitory, 30 to 56 individuals in same room during study breaks
Drug-addicted inpatients
Dogs participate in “all types of work in AAA and AAT,” ranging from 1 to 8 hours (!)
20-minute AAA sessions
2-hour shifts, involving petting, hugging, watching the dog (AAT)
50-60 minute AAI sessions (~AAT)
Highly structured petting, scratching interactions in 60 minute sessions
Multi-professional animal assisted intervention (similar to AAT); sessions 55-60 minutes
Variable number of sessions, sampling over 3 months
3-4 times/week
2 sessions/ week
1 time weekly for 8 weeks
Two 2-week periods during March & April 2012
1/week with 8-10 participants (and 1 dog and 2 therapists)
Use of leash
Not described
Allowed to roam free
Certifying organization requires use of leash, so presumably applied here
For experiments, 7 certified dogs were on leash and 7 were off; dogs in training were on leash
Off leash at home, on leash in novel and AAA environments, but could be attached to collar, harness, or head collar
Off leash
Not described
Praise, allowed to be free in garden and chase lizards
2-minute quiet-play time-out session (found not to significantly alter cortisol levels)
Praise, food treats
Treats prohibited during AAA sessions
Food treats (cheese) were sometimes given, but sometimes only allowed to be smelled to stimulate salivation
Behavioral stress indications
Not described, though authors recommend several days of rest after each session
Frequency of emotional and behavioral responses and calming signals and body language noted by staff
Stress behaviors such as air licking, yawning, panting, pupillary dilation, and whining were observed more frequently in dogs identified as having higher cortisol levels
No behavioral observations in published article
Ethogram (see next table); only behavioral difference noted was dogs standing more in AAA setting than novel or home; but more in novel than home
No differences in activities on the ethogram were found during working sessions; lip licking and other behaviors also did not indicate stress
Cortisol results
Salivary cortisol levels significantly higher on days dogs worked; higher in short sessions; generally higher at end of session; higher in morning than afternoon; cortisol levels increased with number of sessions/week
Fecal and hair cortisol measured supported “positive adaptation to the new environment and role of the dog”
10 dogs showed significant signs of stress, but 11 were either neutral or negative to stress levels of cortisol; experienced dogs were less likely to show stress levels of cortisol
No significant increases in salivary cortisol levels in any of the groups compared to baseline and home levels; positive interactions, quiet play, and affiliate behaviors were linked to reduced cortisol levels; dogs on lead showed high cortisol levels than dogs off lead
Cortisol measure in home, administrative office (novel), and dorm (AAA) environments; no significant difference between home and AAA environments
Salivary cortisol levels indicate dogs were not acutely stressed, but significant decreases were only found in experienced therapy dogs working off leash.

Undoubtedly more studies will be released in the coming months and years, but the following will provide a picture of some of the more dramatic results reported so far.  

2006 Study Finding Elevated Cortisol Levels in Dogs after Therapy Work

Two researchers from the Institute of Anthropology at the University of Vienna, Dorit Karla Haubenhofer and Sylvia Kirchengast (2006), studied salivary cortisol levels in 18 therapy dogs, 15 females (4 spayed) and 3 males (1 neutered).  All dogs were members of an Austrian therapy animal organization, Tiere als Therapie.  The working methods of the dogs were apparently variable as a “therapeutic session was defined as one visit to a facility and included all types of work in AAA and AAT.”  Sessions ranged from one to eight hours, though longer sessions apparently involved many breaks.

The number of sessions in which teams participated in the three-month period during which measurements were taken ranged from 9 to 50, so a team that had 50 sessions could have been averaging visits almost every other day.  The authors found that cortisol concentrations were significantly higher on days that the dogs did therapy work than on other days.  Cortisol levels after sessions were generally significantly higher than those taken before sessions began.  Also, this was more pronounced in the morning than the afternoon, and was higher in short sessions than in long ones.  The number of sessions per week was an important factor:

“Another variable related to physiological arousal was the number of therapeutic sessions done each week, with cortisol concentrations increasing significantly with the number of sessions. This suggests that several days of rest after each therapeutic session might prevent extreme arousal, which could lead to signs of chronic stress.”

Although cortisol levels were higher in short sessions than long sessions, it seems that the intensity of the work for some of the participants was very high, and eight hours in a therapeutic environment is not something I have personally encountered at any institution or facility.  The general nature of the description of the sessions, “all types of work in AAA and AAT,” makes speculation as to how these results may be correlated with others difficult.

The Time-Out Study

Two researchers from the Biology Department of Northeastern Illinois University and one from the Chicago Zoological Society, King et al. (2011), looked at both behavioral factors and cortisol levels in dogs working in animal-assisted therapies (AATs).  A specific purpose of the study was to determine whether therapy dogs allowed to have a quiet-play, time-out session during a working shift would “show fewer signs of stress both physiologically and behaviorally.”  A quiet-play, time-out session was described as consisting “of 2 minutes of visiting with the handler, allowing the dog to chew on a toy, petting and talking to the dog, or providing mental stimulation in the form of obedience commands.”

A total of 27 therapy dog teams were recruited, with 14 female and 13 male dogs, but 6 dogs were dropped from statistical sampling for various reasons.  Handlers had scheduled two-hour shifts at the Edward Hospital in Naperville, Illinois.  They note that this environment often involved “gathered crowds of people, or alarms on hospital equipment.”  Dogs were certified through TDI, which apparently here means Therapy Dogs, Inc. of Cheyenne, Wyoming (i.e., not Therapy Dogs International). 

Participating handlers filled out a survey providing information concerning the dog’s behavior during sessions and “the age of the patient the therapy dog visited, and the type of interaction between the patient and the AAT dog.”  Most patients visited by the dogs were between 41 and 60 years old.  Most engaged the dogs by petting, but some by hugging or just watching the dog.  A researcher also observed the dog’s behavior at the AAT office in the hospital after each work shift, recording, in particular, “panting, air-licking, tremor, pupillary dilation, or yawning 3 times.”

As to the question of how a quiet-time session would affect cortisol levels, the researchers found no significant difference between dogs that had such sessions and those that did not.  Some handlers did report “an initial hesitation in the dog when they returned to work after the time-out session.”  They also found that “AAT-hander survey reports were consistent with physiological indicators of stress,” though “cortisol levels decreased in several dogs at the end of their work shift, whether they had the ‘quiet-play’ time-out session or not.” 

This and the previous study are frequently cited as examples of results indicating that therapy work can be stressful in dogs.  Although I have encountered handlers who take their dogs for shifts as long as two hours, I have always restricted visits with Chloe to about one hour.  This was not initially my own limitation but because I sensed that Chloe’s interest in visiting more people in hospitals and nursing homes declined after about an hour.  Again, I wonder if the length of the sessions was a factor in producing stress. 

Shelter Dog in Alzheimer’s Facility

An early study that did not find unusual cortisol levels in a dog at an Alzheimer’s facility was somewhat different from other studies discussed here because the dog was not brought to the facility by a handler, but was “rehomed” to live in the facility after being at a shelter. (There has been some legal recognition of such resident companion animals in the United States.  See Service and Therapy Dogs in American Society, at 104.)  The research team, composed of veterinarians from the University of Bologna, Piva et al. (2008), chose Daisy, a six-year-old intact female English setter, from a group of eight candidates for the project.  Daisy was chosen because she was kind and cooperative when handled by strangers, interested in people, lacked aggression, was not excitable, had a stable and tolerant temperament, and acquired and carried out simple obedience commands easily.  The commands included Sit, Come, Wait, Sit on the Chair, and Down. 

Daisy (courtesy Dr. Elisabetta Piva)
Daisy was gradually introduced to the facility, initially by weekly visits, first with employees, then with patients. Meetings were held in a large room where she had the opportunity to escape.  Finally, she was allowed to explore the entire facility and randomly meet patients and employees.  Once in residence, Daily began to have 20-minute AAA sessions three to four times a week with four Alzheimer’s patients chosen by the staff psychologist. 

“These sessions include activities such as calling the dog from varied distances with vocal commands or gestures, brushing, playing, walking, feeding, and using hand signals to practice obedience exercises, depending on the patient involved.”  After sessions, Daisy was allowed to unwind and praised for relaxing.  She was also offered the chance to be free in the garden where she could hunt lizards.

Cortisol levels were measured:

“Fecal cortisol levels were evaluated in samples collected from every spontaneous defecation during the gradual introduction phase starting 2 days before to 2 days after each visit, and 2 days before to 7 days after adoption. An increase in fecal cortisol concentration indicates an activation of the adrenocortex in response to stressful conditions….”

The researches argued that their “findings suggest that despite small changes in the environment (more human patients in the facility) and in daily routines (fewer walks outside during AAA because of the hot and sometimes stormy summer weather that was considered dangerous for elderly people), the dog was healthy and her level of social interaction, exploration, and playfulness increased during the course of the AAA program. The dog displayed no aggressive or sexual behavior, even when in heat.  Autogrooming detected previously decreased in frequency and an acral lick granuloma that she had had gradually disappeared. Over-grooming, shown frequently by Daisy during the first examination in the shelter, was displayed less frequently as time went by in her new social and physical environment.”

The authors concluded:

“Daisy was able to cope with the new environment. The presence of enriched social contacts and the opportunity of increased physical activity positively affected her well-being…. The hormonal trend, especially with hair cortisol, seems to be correlated with the clinical and behavioral findings, all supporting a positive adaptation to the new environment and role of the dog…. Although this report refers to a single subject, Daisy, the results of our work represent an encouraging basis for further studies on a wider scale. Our dog, besides her ability to adequately carry out AAA sessions, showed progressive benefits (re-homing, increased activity and social interaction, decreased signals of stress), as definitively assessed by statistical analysis.” 

Although not highlighted in the results, Daisy was frequently off leash, which may be significant in light of a study to be discussed below.  The results also suggest that Alzheimer’s patients are not inherently stressful subjects for a therapy dog to work with.  

Leash Study

Six researchers from a number of institutions in Austria, Glenk et al. (2013), sought to measure changes in salivary cortisol levels of dogs participating in animal-assisted interventions (AAIs, essentially AAT sessions), collecting samples on working and non-working days of 21 dogs.  The researchers concluded that both therapy dogs and therapy dogs in training were not stressed by their participation in the activities. They did detect, however, that cortisol levels were significantly different depending on whether the dogs were working on- or off-lead in the sessions.  The paper argues that the effect of a leash on cortisol levels of therapy dogs may have been underestimated.  Beerda et al. (1998) had previously demonstrated that pulling dogs on a lead cased increases in cortisol when the dogs were confronted with sudden noises or electroshocks. 

Teams were recruited from an Austrian AAI organization whose members regularly work in dog-assisted group therapy.  All handlers were women who also took part in the therapy sessions.  An experimenter also attended the sessions for sampling procedures, with the approval of staff members of the facility where the sessions occurred.  Seven dogs were males and 14 were females, ranging from 1.5 to 14 years. Nine of the females were spayed, but the other females were not in estrus or pregnant during the experiments. 

To become a certified therapy dog team, “dogs and owners undergo special training, during which animal handlers can decide whether they want to work with their dog on- or off-lead.”  This is not typical in the United States. The rules of the organization to which Chloe and I belong provide: “Dogs must be kept on leash at all times when visiting, except when warranted (during a demonstration).”  There was another significant difference from my experience.  Among interaction behaviors recorded during the study were food treats being given to the dogs.  My organization's rules state that the “use of food or treats is prohibited while visiting (exception—during a demonstration, the handler can treat the dog).”

There were three groups used for the study, one group of dogs trained to work on lead (CTD-ON, in the acronym used in the paper), and another group off lead (CTD-OFF).  Dogs used for the study “had a minimum of one year working experience in mental healthcare facilities.”  The third group consisted of dogs that were in training (TDT-ON), whose owners had not yet decided whether to work on- or off-lead, but which were on lead for the experiments.  There were seven dogs in each group. The existence of such a training stage indicates a higher level of training than is often found for therapy-dog qualification in many countries. 

Sessions lasted 50 to 60 minutes and were described as follows:

“All therapy sessions consisted of theory parts, interpersonal communication and interaction parts with the therapy dog. Each therapy programme started with a group of ten adult patients and was run weekly for eight weeks with the same individuals (animal handler, therapy dog and patients) present. There was no patient turnover during these eight weeks. AAI programmes were supervised by each institution’s staff members and participation depended on the respective physical or psychological condition of the patients. In the experimental sessions, 8–10 patients were present. The patients were informed previously how to interact in an appropriate way with the therapy dog before the dog was first introduced to the group.  During therapy, the patients were seated in chairs and instructed by the animal handlers when and how to interact with the dog (ie stand up, call, touch, grab or pull the dog’s lead).”

Certified dogs participated in the sessions while dogs in training watched and did not interact with the working dog.  Dogs were trained and handled by “positive reinforcement and gentle handling.”  Salivary cortisol was measured at home, before and after AAI sessions, and on non-working days.  Saliva samples were also collected during some therapy group sessions. 

“[O]ur study results reveal that performance in group-AAIs in adult mental healthcare did not stimulate significant increases in salivary cortisol stress responses in CTD-ON, CTD-OFF or TDT-ON when working cortisol levels were compared to baseline levels and home levels. These are important findings, considering that in dogs, an elevation in cortisol has been associated with stressful conditions resulting from fear….  On the other hand, positive interaction with humans, quiet play and affiliate behaviours were linked to reduced cortisol levels in dogs….”

Cortisol Levels Before and During Sessions (courtesy Dr. Lisa Glenk)
The graphic from the paper shows that before therapy (Time 1a) dogs in the three categories, two with leashes (but one in training) had fairly similar cortisol levels, but during therapy (Time 1b), they separated substantially, with dogs off leash having lower cortisol levels then dogs on leash. Dogs in training, with leashes, had the highest levels. As noted above, Beerda et al. (1998) had determined that being pulled on a lead caused stress in dogs and increased cortisol levels, so these results were not surprising. 

The longer sessions lasted, the more cortisol levels may have declined in CTD-ON dogs, but more research will be needed to verify this.  The team notes that there may be some differences between handlers that prefer to work with dogs on lead and those that do not, meaning perhaps that more controlling handlers might add some stress to the therapy situation.  On the other hand, “animal handlers with therapy dogs on the lead should be aware of subtle signals of discomfort in their dogs when they interact with patients and react accordingly.”  My own experience over six years of therapy dog work confirms this.  I have no choice but to work on leash, but more than a few times I have felt that a patient or other individual in a therapy dog visit was discomforting Chloe in some way, perhaps slapping rather than petting, or smothering her, and I have had to extract her from the setting as tactfully as possible. 

As to why Hubenhofer and Kirchengast (2006) and King et al. (2011), both discussed above, reached different conclusions, this paper speculates:

“[I]t is likely that the AAIs investigated by the different authors are not directly comparable because of their different conceptual context (eg therapy content, single patient versus group interventions, familiar versus unfamiliar patients), environment (eg therapy facility such as hospital, prison, geriatrics) and arrangement (eg frequency, intensity and duration of human-animal contact, dog on/off the lead, refuge for the dog).”

This is an important observation.  With the limited number of cortisol and cortisol/behavior studies of therapy dogs conducted so far, more patterns regarding the effects of different types of institutions and different types of programs may come to light. 

Substance Abuse Facility

The same Austrian team published another paper in 2014, looking at behavior and cortisol levels in therapy dogs that participated in sessions at an in-patient substance abuse treatment facility, measuring cortisol levels before and after sessions.  Levels of the chemical decreased in all sessions.  “There was no difference between salivary cortisol levels sampled on a nonworking day at home and work-related levels sampled at the therapy site. None of the behavioral parameters varied significantly over the course of the 5 MTI [multiprofessional animal-assisted intervention] sessions.”  MTI, according to the authors, “meets all require criteria to be considered AAT … and is carried out by 2 human experts with a back ground in psychology, pedagogy, life science, and/or social science….” The paper elaborates:

“Mediated by the MTI professionals, patients used signals to communicate with the dog. These signals were verbal or nonverbal cues including different hand signs, eye contact, mimics, words, and tone of voice. The main goal of MTI is to enhance patients’ emotional and social competence through implicit learning during interaction with a therapy dog…. Accordingly, the participants had been instructed how to interact in an appropriate way with the therapy dog before the dog was introduced to the group. Human-animal contact was initiated by the freely moving dog, which was kept off lead. Human-animal interaction behaviors moreover included verbal contact, where people talked to the dog or spoke in a high-pitched/fluctuating voice to praise the dog. Tactile contact included softly touching, stroking, and/or grooming the dog. To play with the dog, people used dog toys and/or gently gestured with hand, arms, and fingers. For ethical reasons, dogs were never forced into positions and were able to lie down, drink water, or leave the therapy room at any time.”

The researchers concluded that therapy dogs “are not being stressed by repeated participation in in-patient substance abuse therapy sessions.”  They argue that therapy dog handlers need more guidance:

“The development of a practitioner’s guide on dog welfare for AAI professionals and, maybe even more importantly, AAI volunteers shall be a forthcoming endeavor. AAI volunteers are often very dedicated to their work, but they also need to be well aware of subtle signs of discomfort in their dogs. Future research still needs to identify the populations or situations where contact with therapy animals may be potentially problematic or inappropriate for either the animals or the people involved….”

This also is an important observation.  As additional research establishes what sorts of conditions might produce stress in dogs, handlers should receive additional instruction in how to recognize stress.  At the moment, most qualification examinations in the U.S. that I have encountered through my own experience or that of others primarily involves testing the dog’s ability to follow the handler’s commands, not whether the handler can read signals given by the dog. 

Ethogram Study

Four researchers at the Virginia-Maryland Regional College of Veterinary Medicine and two from the vet school at the University of Pennsylvania, Ng et al. (2014), sought to quantify the stress effects on therapy dogs from engaging in animal-assisted activities (AAAs). This team sought to determine how stress levels would differ for therapy dogs in three environments: their homes, a novel setting, which was an administrative area in Penn’s vet school, and AAA sessions with college students in the communal area of a residence hall. 

Dogs were either part of the University of Pennsylvania’s therapy dog program (Vet Pets) or dogs registered with Therapy Dogs International.  All participating dogs were fixed or spayed, seven females and eight males.  Five had been certified for only one year, but one had been certified for six years and one for eight. Some averaged only one therapy dog visit per month, but one averaged between seven and ten days per month.  Session lengths from the experience of the handlers could be as low as half an hour and as high as two hours.  Of 15 dogs, only two had male handlers. 

Saliva was collected after the handler and dog had been in each of the three settings for an hour.  The researchers state that the “goal of the novel setting was to allow the dog to act as it typically would with its owner when waiting in an unfamiliar environment for 60 min without stranger interaction.” Sometimes two of the teams were in the administrative location simultaneously, though dogs were limited to interacting only with their own handlers in the location. 

The AAA sessions were with undergraduates in a communal space of Rodin College House, a Penn dorm.  The room was often used by students for study breaks and there were between 30 and 56 individuals present in the room during the sessions.  The researchers elaborate:

“The room was 9.1 m × 9.1 m, with tables and chairs arranged to evenly spread and demarcate eight, 2.1 m × 2.1 m spaces for each dog-handler team to remain during the session. The dog remained in close proximity to the handler on a1.83-m leash attached to a collar, harness, or head collar that the dog was accustomed to wearing at AAA visits. A maximum of eight dogs used separate spaces of the room simultaneously, but dogs were restricted from interacting with one another.”

The AAA sessions were highly structured with interactions and saliva collection kept in specific sequences:

“Participants were instructed to approach the dog from the side, to extend a hand to allow the dog to sniff, and to pet gently. Participants were instructed to avoid the following: aggressive gestures, making loud noises, leaning over the dog, giving treats, and crowding around the dog. Assigned “petters” were assigned to pet specific dogs during the 5-min petting time … but participants were permitted to interact with any of the dogs outside of that time….

“Each petter was instructed to sit on the floor to the side of the dog opposite the handler, where the petter did not obstruct the video camera view of the dog. In all settings, each petter was instructed to sit next to, rather than facing, the dog and gently stroke, pat, massage, and/or scratch the dog anywhere on the body with at least one hand remaining on the dog at all times. The dog was to be allowed to position itself and behave as it wanted during the 5-min petting procedure, as long as it remained within the assigned space within view of the video camera (accomplished via leash).”

Saliva was collected every half hour as prior research by Vincent and Mitchell (1992), and Handlin et al. (2011) had found significant changes in cortisol levels 15 to 30 minutes after stress events. 

The study also involved looking at behavioral changes in the three distinct environments.  Sessions were videotaped and behavioral patterns of the dogs, particularly while being petted, were analyzed.  An ethogram, a list of behaviors divided into postural states, spontaneous events, oral behavior, and alertness was completed by analysts of the tapes. (See Figure 2, below.)  Postural states included sitting, standing, recumbent, ambulating, exploring, and crouching.  Exploring is defined as “moving slowly, sniffing, investigating the environment.”  Crouching is defined as “rapid, pronounced lowering of posture, sometimes in combination with movements that enlarge distance to eliciting stimulus; posture shows lowered position of tail, backward positioning of ears, legs bent.” 

Spontaneous events were subdivided into eight categories: paw lifting, vocalizing (any form), scratching, body shaking, trembling, jumping (“springing into air, either to make contact with an object or person or for no apparent reason”), repetitively moving head (continuously for more than three seconds), and stretching.  Oral behavior could be panting, neutral (mouth closed), mouth opening (“opening, closing mouth with rapid movements without extending tongue; possibly yawning”), lip licking (“includes tongue out: tip of tongue briefly extended; snout licking: part of tongue shown, moved along upper lip; swallowing; smacking”), licking person, licking object, and self-grooming (“oral behaviors directed toward dog’s own body (licking, chewing skin and coat)”).  Alertness was divide into two categories: alert (“eyes kept open”), and rest/sleep (“eyes closed, dog inactive > 10s”).  

The reason for looking at both cortisol levels and behavior is explained as follows in the paper:

“It is important to consider that the same behavior can correspond to different emotional states of the dog. For example, ambulating may be strictly a motor behavior, but it can also be an indicator of restlessness or anxiety, depending on how the behavior is performed and the type of concomitant behaviors present at the time of ambulating. Therefore, it is necessary to assess behavior in conjunction with a physiologic parameter such as cortisol.”

The only behavioral differences noted had to do with postural states.  As to behavior, dogs stood significantly more of the time in the residence hall than the other settings, but more in the administrative area than at home.  They also spent more time ambulating in the residence hall, but very little in the administrative area and almost none at home.  Recumbent time was highest at home, which is no surprise, but at 60 minutes was higher in the residence hall than in the administrative offices, while at 90 minutes was higher in the administrative offices than in the residence hall.  The researchers speculated that in the administrative offices, the environment was unfamiliar and the dogs “were hyper vigilant to disturbances outside the room.” 

The researchers conclude:

“The 60-min AAA for college students in a dormitory setting did not appear to cause significant HPA activation or increases in stress-associated behaviors in registered AAA dogs….  No physiologic or behavioral indicators of stress, fatigue, or exhaustion were present during the AAA, suggesting that this particular AAA with college students did not negatively impact the welfare of these dogs. Furthermore, salivary cortisol was higher in the novel setting, which may be explained by the unpredictable nature of the setting.”

They speculate that “dogs were likely not stressed during the AAA because it was a familiar and predictable situation.”  In the novel setting, the administrative offices, the dogs may have been slightly stressed by an inability to predict what would happen next.  Although the dogs were not in direct contact with the veterinary hospital, the researchers note that they “could have detected subtle cues of a veterinary hospital environment.  Therefore, the physical environment alone, especially a veterinary hospital setting, may be physiologically stimulating, irrespective of the activity performed.” 

As discussed, this study did not look at dogs in AAT settings, just AAA settings.  The researchers note that “there is no single validated model to test the effect of AAA or AAT because they vary greatly in intensity of interaction, duration, objectives, and demographics of recipients.”  Thus, an AAT session, which “typically involves a continuous interaction with a single or small group of individuals,” might be more stressful for dogs. 

The researchers argue that their study describes “a rigorous method of assessing welfare in AAA dogs that can be applied in larger populations of working dogs.”  This proposition should be considered by other research teams entering this area.  Consistency of research designs will help determine whether and when dogs are being stressed by therapy work. 


Variations in the studies as to the numbers of teams, the training and certification approaches, the types of facilities visited, what activities were conducted inside those facilities, and the length and frequency of visits and sessions, means that comparing studies so far published cannot assure solid conclusions.  American and European studies differ in that the former generally involved dogs working on leashes, without food treats, whereas European studies sometimes involved dogs moving freely in therapeutic settings and receiving food treats from handlers and sometimes others.  Attempts to specify behavioral indicators of stress have had limited success, and correlations with cortisol levels have proved elusive. 

Arguments, such as that made by Ng et al., that a uniform grading system for behavioral measurements would be desirable, should be adopted, perhaps after some major conference on stress and trained dogs.  The behavioral measures suggested by four of the major studies discussed here are listed in Figure 2. 

Figure 2. Behavioral Measures in Cortisol Studies

Piva et al. (2008)
King et al. (2011)
Ng et al. (2014)
Glenk et al. (2014)
Nose and lip licking
Sit or curl up
Lie down, also with head
Closing eyes
Rapid eye movement
Walking, pacing
Low posture
Displacement activity
Redirect activity
Pupillary dilation
Air licking
Postural state:

Spontaneous event:
Paw lifting
Body shaking
Repetitively moving head

Oral behavior:
Neutral (mouth closed)
Mouth opening
Lip licking
Licking person
Licking object


Lip licking
Paw lifting
Body shake
Tail wagging

Response to human action:
Takes a treat
Obeys command

Arguably, a set of behaviors and a recording system should  be used for stress studies of dogs doing not only therapy work but also service, police, rescue work, etc. 


The standard U.S. requirement that dogs be on lead during therapy visits does not come solely from certifying organizations or the mandates of their insurance providers.  Recently, a hospital I have visited with Chloe went through a national accreditation process where one dog, which was regularly turned loose in the psychiatric ward, was banned from the hospital because the accrediting organization stated that therapy dogs visiting the hospital had to be on a leash at all times.  This is also required by some insurance policies for hospitals.  Nevertheless, therapy dog organizations should give further consideration to whether working with therapy dogs off leash might be appropriate in certain confined spaces and structured activities, and if necessary, engage other stakeholders in therapy dog work in discussions on this issue.  It might be appropriate for certifying organizations to consider advanced certifications for dogs that are capable of working off leash under voice command of the handler. Progress in this area might require the major therapy dog organizations to cooperate with each other. 

As discussed, most target populations caused and most research designs found no evident stress in dogs from therapy work.  Nevertheless, Marinelli et al. (2009b) found that the age of individuals involved in animal-related activities and therapies “influenced the expression of stress-related behaviors” in dogs in that more such behaviors were evident when dogs were with children under 12 years of age.  This has been my experience as well.  When Chloe turned six I stopped going to a facility for developmentally disabled children because one teacher who had often accompanied me on visits to the classrooms left and too many other teachers regarded my entry into the room as time to attend to other matters.  The children often overwhelmed Chloe.  As she has gotten older, I have become more concerned with experiences that I sense, without any chemical or scientific behavioral analysis, may be stressful for her. 

These studies do not negate the importance of the handler knowing his or her dog and sensing when it is time to go home or to stop working.  Even the most sophisticated behavioral analysis, with video cameras and behavioral patterns categorized down to seconds if not microseconds, cannot replace the instincts of a good handler.  Nor can handlers be automatons created complete and unchanging after passing a certifying examination.  My work with Chloe has evolved over many years, and I have grown to trust my sense of how she can be most effective, and what will be most helpful to a patient, resident, student, or other person we are visiting when she is working.  She looks at me when she wants me to realize that it is time to move on, or that the patient has had enough.  She has also grown into the work.  Any long-term handler with an experienced therapy dog will say much the same. 

Stress research is a promising area and all those involved or interested in therapy dog work, whether as handlers, trainers, certifying organizations, recipient institutions, or policy makers, should be paying attention as more results come in. 


Beerda, B., Schilder, M.B.H., van Hooff, J.A.R.A.M., de Vries, H.W., and Mol, J.A. (1998). Behavioural, Saliva Cortisol and Heart Rate Responses to Different Types of Stimuli in Dogs.  Applied Animal Behaviour Science, 58, 365-381.

Ferrara, M., Natoli, E., and Fantini, C. (2004). Dog Welfare During Animal Assisted Activities and Animal Assisted Therapy.  Tenth International Conference of the IAHAIO. Glasgow, Scotland (finding dogs did not show stressed behavior from AAT and AAA sessions).

Glenk, L.M., Kothgassner, O.D., Stetina, B.U., Palme, R., Kepplinger, B., and Baran, H. (2013). Therapy Dogs’ Salivary Cortisol Levels May Vary During Animal-Assisted Interventions. Animal Welfare, 22, 369–378.

Glenk, L.M., Kothgassner, O.D., Stetina, B.U., Palme, R., Kepplinger, B., and Baran, H. (2014). Salivary Cortisol and Behavior in Therapy Dogs During Animal-Assisted Interventions: A Pilot Study.  Journal of Veterinary Behavior: Clinical Applications and Research, 9(3), 98-106

Handlin, L., Hydbring-Sandberg, E., Nilsson, A., Ejdeb, ck.M., Jansson, A.U., and s-Moberg, K. (2011). Short-Term Interaction between Dogs and their Owners: Effects on Oxytocin, Cortisol, Insulin and Heart Rate: An Exploratory Study.  Anthrozoos Multidisciplinary Journal. People and Animals, 24, 301–315.

Haubenhofer, D.K., and Kirchengast, S. (2006). Physiological  Arousal for Companion Dogs Working with Their Owners in Animal-Assisted Activities and Animal-Assisted Therapy.  Journal of Applied Animal Welfare Science, 9, 165-172.

King, C., Watters, J., and Mungre, S. (2011). Effect of a Time-out Session with Working Animal-Assisted Therapy Dogs.  Journal of Veterinary Behavior: Clinical Applications and Research, 6, 232–238.

Marinelli, L., Slvadoretti, M., Normando, S., Moro, L., Ravarotto, L., and Bono, G. (2008).  Cane, Monitoraggio de Benessere [Monitoring Dog Welfare]. Obiettivi & DocumentiVeterinari, 6, 36-45 (finding none of the monitored parameters indicated that AAA sessions impaired the welfare of the dogs, but recommending exercise and interaction with conspecifics after AAA sessions).

Marinelli, L., Mongillo, P. Salvadoretti, M., Normando S., and Bono, G. (2009a). Welfare Assessment of Dogs Involved in Animal Assisted Activities. Journal of Veterinary Behavior: Clinical Applications and Research, 4, 84-85. (Salivary cortisol levels and behavioral indications of acute stress were not elevated in therapy dogs visiting a retirement home over a period of seven weeks.)

Marinelli, L., Normando, S., Siliprandi, C., Salvadoretti, M., and Mongillo, P. (2009b). Dog Assisted Interventions in a Specialized Centre and Potential Concerns for Animal Welfare.  Veterinary Research Communications, 33 (Supplement 1), S93-S95.

Ng., Z.Y., Pierce, B.J., Otto, C.M, Buechner-Maxwell, V.A., Siracusa, C., and Were, S.R. (2014). The Effect of Dog-Human Interaction on Cortisol and Behavior in Registered Animal-Assisted Activity Dogs.  Applied Animal Behaviour Science, 159, 69-81

Odendaal, J., and Meintjes, R.A. (2003). Neurophysiological Correlates of Affiliative Behavior Between Humans and Dogs.  The Veterinary Journal, 165, 296-301.

Piva, E., Liverani, V., Accorsi, P.A., Sarli, G., and Gandini, G. (2008).  Welfare in a Shelter Dog Rehomed with Alzheimer Patients.  Journal of Veterinary Behavior, 3, 87-94

Serpell, J.A., Coppinter, R., Fine, A.H., and Peralta, J.M. (1999).  Welfare Considerations in Therapy and Assistance Animals: An Ethical Comment.  Chapter 18 in Handbook on Animal-Assisted Therapy” Theoretical Foundations and Guidelines for Practice. New York: Academic Press, 415-431. (The chapter has appeared in two subsequent editions of this book.)  

Vincent, I.C., and Michell, A.R. (1992). Comparison of  Cortisol Concentrations in Saliva and Plasma of Dogs. Research in Veterinary Science, 53, 342–345.