HASS RTH Benchmarking Tool

Since a count of RTH outcomes alone makes it difficult to compare performance across shelters of varying sizes, we normally define a shelter’s RTH “rate” as the percentage of stray intakes that have an RTH outcome. This ensures the rate is measuring RTH in the context of lost and found pets, as opposed to cases like custody or seizures, in which the owner is known. However, from our engagement with shelters across the industry, RTH rates are not always measured this way. In building this tool, we wanted to include the largest number of organizations possible in our dataset, for which data from Shelter Animal Counts (SAC) is most helpful. Unfortunately, from data reported to SAC, it is impossible to know which of the reported outcomes were stray intakes. As a proxy, SAC reports the percentage of RTH outcomes out of the total outcomes as an RTH rate. We found that while this rate varies from what we would consider the “true” RTH rate in individual shelters, it is fairly accurate on average. As it enables us to access the breadth of SAC data, this is the formula used to calculate RTH rates in this tool.

We wanted to provide our pilot shelters, and any other interested organizations, with a data-driven method to come up with a goal for RTH rate to guide lost pet reunification efforts and understand where they stand relative to other organizations. We recognized that different kinds of organizations have different capacities for RTH – for example, we assumed that an organization taking in 200 stray animals annually will have an easier time returning 30% of them home than an organization taking in 2000 animals. There are also factors other than intake volume that would presumably affect RTH rates. So the task at hand was to first identify significant factors that influence RTH rates, and then be able to produce estimated RTH rates for organizations based on them. We used simple statistical modeling to achieve just that, and the results of this model are used to select the factors that will be used to generate the estimates, and to identify how each factor should affect the result. The factors we identified appear in the input section or are derived from them (for example, we ask for a US state, but the model assigns this to the region each state is classified into – see further explanation later). An organization can insert these few simple details and receive a “predicted” RTH rate, which can be interpreted as the RTH rate that is expected, on average, for organizations of their kind, given the RTH rates of all organizations in the data.

• Organization type - using SAC’s classification: Government shelter, Nonprofit with contract, Nonprofit without contract. Rescues were not included as they do not typically take in stray/loose pets and partake in significant reunification efforts.
• Region - a standard division based on the chosen US state into Northeast, Midwest (with RTH rates equivalent to each other), West (higher rates), and South (lower rates).
• Annual stray intakes - referring to stray intake only, and grouped into four categories: up to 500, 500-1500, 1500-3000, and 3000+. The higher the group, the lower the expected RTH rate. While the essence of this is that intake volume matters, there is nothing set in stone about these particular thresholds. They arose from the available data as a good fit.
• The percentage of all intakes that are stray - this reflects the centrality of stray intakes and in turn the potential importance of lost/loose pet reunification efforts - higher percentages predict higher RTH rates.
• Social Vulnerability Index (SVI) for the county in which the organization is located - this is a rough-yet-effective way of capturing multiple dimensions of the socioeconomic situation of people served by the shelter. A higher SVI score, which means higher level of vulnerability, means a lower RTH rate expectation. Other, more specific indicators like household income and poverty rates were also attempted and behaved similarly.

• The data used to model RTH rates is the full SAC database de-identified from years 2019-2022 inclusive, provided generously by SAC.
• Only organizations with complete contributions were included on a per-year basis (e.g an organization that had a full year in 2022 but 10 months in 2021 will have the former included in the data, but not the latter). A zero-intake month also counted as a missing month. Help grow our database by submitting your complete data.
• As mentioned above, organizations classified by SAC as rescues were excluded, as were organizations with less than 100 stray intakes, a threshold used to obtain a reliable RTH rate while accommodating for organizations of different sizes.
• At first, separate models were built for cats and dogs, as their RTH rates normally vary significantly. As analysis unfolded, we found that due to the low cat RTH rates in most shelters (~2%), there isn’t a whole lot of effective modeling we can do to provide benchmarks. Therefore, the tool was designed for dog RTH rates only (hence asking for annual and stray dog intake only). Some data about cat RTH data is mentioned below.
• After this preprocessing, data from the remaining 1009 organizations was used to build the model and share its results through this tool. These include 442 Government shelters, 380 nonprofits with contracts and 187 nonprofits without contracts. 468 of the organizations are located in the South, 260 in the West, 213 Midwest, and 68 Northeast.

Cats have significantly lower RTH rates than dogs, so low that it made elaborate benchmarking unhelpful. The average RTH rate for cats across 1,095 organizations with complete cat data (filtered like the dog data above) was 2%. One adjustment we made was including only cats reported in the Adult age category (as opposed to Youth or Age Unknown, the other two SAC categories), as young cats are less likely to be owned. This brings the average to 5% – still pretty low. Nevertheless, there are organizations out there achieving really impressive numbers – 212 organizations recorded over 10% RTH rate for cats, and 18 have exceeded 30%! These higher cat RTH organizations exist across all regions, organization types, and levels of social and economical vulnerability. One organization that achieved a 35% cat RTH rate had over 3,000 stray cat intakes that year, and another returned home 16% of 9000+ stray cats!

To view all HASS resources related to RTH and lost pet reunification programs, see the Lost Pet Reunification pathway section of the HASS Playbook and view supporting resources on our pathway page.

Of course! We would love to hear your suggestions or hear any feedback. Please contact hass-project@americanpetsalive.org