August 31, 2018

Audience: Executive Leadership, Shelter/Rescue Staff & Volunteers, Veterinary Team

Organization: Best Friends Animal Society
Investigator(s): Liz Finch, Gary Patronek, Michelle Logan, Stephanie Macgill and Kelly Kramer
Grant Amount: $25,000
Project Type: Phase 1
Project Status: Research Complete

Project Summary

Best Friends Animal Society and Dr. Gary Patronek completed a pilot study at PACC investigating the effect of breed labeling, or lack thereof, on the adoption of shelter dogs. This study also compared breed assignments in the shelter's database with DNA data. However, because DNA analysis results were almost always not back before the dog was adopted, this part of the study involved primarily a comparison of staff guess of breed to providing no breed information. The results of this pilot study revealed that having no breed information on kennel cards contributed to a longer length of stay (LOS). Further investigation is needed before these results should be used to impact shelter protocols.


Primary objectives were to: investigate whether breed labeling, or lack thereof, affects dogs' adoption and length of stay; assess how closely, if at all, staff's breed guess, or opinion of a surrendering owner, matches DNA results; and identify dog characteristics associated with live release and shorter length of stay.


Dogs admitted to Pima Animal Care Center (PACC) shelter were assigned to one of 3 groups for kennel labeling: 1) staff guess or owner-provided information about breed, 2) no breed information, or 3) DNA results. To identify the level of agreement between breed assignment and DNA tests, the results of dogs with DNA tests were statistically compared with either primary or secondary breed entries in PACC's animal management software. Characteristics associated with live release and length of stay were identified by downloading all records from the shelter's database from January 1, 2015 through December 31, 2016, and analysis via statistical software.


  • Absence of breed information on cage cards resulted in a longer length of stay for dogs. The value of this information is limited because this study was done at a single shelter (unknown whether results are generalizable) and because DNA analysis often wasn't back before the dog was adopted.
  • For longer stays, where DNA results were available, 80% of dogs with DNA or breed guess on their kennels were adopted by day 14 or 15, while 80% of dogs without breed on their kennel cards were not adopted until day 20.
  • The same analysis was conducted with dogs whose physical appearances matched stigmatized breeds. The results indicate that for these dogs, as well as non-stigmatized breeds, having no breed information still tends to extend length of stay. However, the effect was more pronounced for dogs who looked like a stigmatized breed.
  • The concordance between breed label by staff guess and DNA test was approximately 63%, with no significant difference for dogs with "pit bull" DNA and dogs without any "pit bull" DNA (p=0.38).


The evidence of poor agreement between staff guess and DNA may encourage more shelters to abandon the idea that visually-based breed labels are accurate or informative. The demonstrated benefits of foster care, particularly for adult dogs, may encourage shelters to expand and more fully utilize foster programs. The two-year data shows that overall, very few dogs were admitted due to severe behavior problems, again contradicting the popular misconception that shelter dogs are somehow "damaged goods." The study showed that dogs returned by adopters did not suffer any adverse effects with respect to live release or length of stay, which may encourage other shelters to be more flexible and open in their adoption/return policies. Lastly, this pilot study found that having no breed information on the kennel card resulted in slightly longer length of stay; however, this result bears further exploration. Although there are numerous reasons why omitting inaccurate information on cage cards may be beneficial to dogs, it is important to determine if this effect is generalizable and if so, identify the reason(s). It's possible that potential adopters prefer more, rather than less information, such as a dog's personality or behavior. Sorting all of these out is important for dogs and shelters, given the trend to remove inaccurate breed information from kennel cards.