A label's a label, no matter the dog: Evaluating the generalizability of the removal of breed labels from adoption cards.

A common barrier to entry for New York City (NYC) dog adopters trying to rent apartments is the breed label the animal shelter assigned to their dog, despite the fact the labelling is primarily based on intuition and appearance. Bideawee, a limited admission shelter with three locations in the greater New York area, including one in NYC, phased out breed labels from their adoption cards in December 2017. In this study, we evaluated the generalizability of previous findings, specifically, that the removal of breed labels from adoption cards affected length of stay. Moreover, due to Bideawee's multi-location structure, this study provided a unique opportunity to compare variables across different shelter sites while having shelter administration practices held constant. Data from 16-month time periods before and after breed labels were removed was compared. The median length of stay of a dog at Bideawee decreased by 11.3 days (-37.3%) once breed labels were removed (Mdn = 19.0) compared to when breed labels were in place (Mdn = 30.3). A Mann Whitney test indicated that this difference was statistically significant (U(Nno breed labels = 1259, Nbreed labels = 987) = 386309.5, z = -15.41, p < .001). Dogs with a "green" behavior assessments (on a scale of green, blue, yellow, red) were almost four and a half times more likely to be adopted faster than "red" dogs (HR: 4.495, 95% CI 2.755-7.335, p < .001) before breed labels were removed, but only two times as likely to be adopted faster afterwards (HR: 2.220, 95% CI 1.514-3.254, p < .001). The return rate stayed constant across the two time periods at 6%. These findings provide new insights on dog adoptions in the NYC area and suggest that the removal of breed labels will help all dogs get adopted from animal shelters.