How counting foster time changes shelter length of stay
Many shelter systems record "foster" as an outcome and a return from foster as a new intake. That stops the length-of-stay clock while the animal is still in the shelter's care, so stays look shorter than they are.
What we did
Three of the eight public shelters in our data code foster this way: Dallas, TX; Louisville, KY; and Norfolk, VA. We linked each foster exit to the same animal's return-from-foster intake and counted the stay through to the final outcome. That covers 258,463 completed stays.
What changed
- Average stay: 8.2 days as recorded, 17.8 days with foster time linked.
- Median stay: 3 days → 5 days.
- Stays over 30 days: 5.7% → 15.5% of animals.
- Share of all days in care used by stays over 30 days: 40.3% → 68.8%.
- Counting only days in the shelter itself (excluding time in foster homes), the average is 9.8 days.
| Shelter | Median days | Average days | Over 30 days |
|---|---|---|---|
| Dallas, TX | 3 → 5 | 6.6 → 16.0 | 4.1% → 12.6% |
| Louisville, KY | 4 → 8 | 10.2 → 20.7 | 7.6% → 21.1% |
| Norfolk, VA | 4 → 9 | 12.9 → 23.2 | 10.4% → 23.0% |
Why it matters
If foster time is left out, long-stay animals look like short stays, and the animals that most need a plan disappear from the numbers. Benchmarks that mix shelters which code foster with those that don't also compare unlike things. Our benchmarks and home page figures now count foster time wherever a shelter records it.
For your own data
If your software records foster placements, our member upload tools link foster stints to the same animal automatically. If it doesn't, the results note that foster time can't be counted. See methodology and foster program impact.
Source: Shelter Insights analysis of city open-data outcome records for Dallas, Louisville and Norfolk, processed Oct 22, 2026.