# Burrow Watch — Theory of Change

## Why this exists
The quarterly report should show whether the program is actually moving
toward its intended outcomes, not just list activity. This logic model is
the frame the report generator skill checks findings against.

## Logic model

**Inputs**
- Volunteer rangers (~22 across 5 zones)
- Contract vet (Dr. Suresh Okafor)
- Grant funding (regional biodiversity grant) + public donations
- Landholder access agreements

**Activities**
- Patrols and disease detection
- Mange treatment
- Burrow/habitat condition monitoring
- Landholder liaison and access maintenance

**Outputs** (what the program directly produces — the quant data tracks these)
- Burrows monitored per zone
- Mange cases detected and treated
- Patrol hours logged
- Landholder relationships maintained/expanded

**Outcomes** (what should change as a result — this is what the report
should actually assess progress against)
1. **Reduced mange prevalence** in monitored zones over time
2. **Stable or improving burrow occupancy**, indicating habitat recovery
3. **Sustained volunteer capacity** — enough active rangers to maintain
   patrol coverage without burnout-driven attrition
4. **Sustained landholder support**, enabling continued or expanded access
5. **Funding continuity** sufficient to maintain vet and patrol capacity

**Impact** (long-term, harder to observe directly this cycle)
- Recovery of wombat population and habitat resilience in Wandarra Valley
  following bushfire damage

## Assumptions worth naming
- Early detection meaningfully improves treatment success (supported by
  the vet's own observation in interviews)
- Volunteer capacity is a limiting factor on disease response, not just a
  "nice to have" — if this breaks, outcome 1 and 2 are both at risk even if
  funding and vet capacity hold

## How the report should use this
For each outcome above, the report should state: **on track / at risk /
insufficient data**, with a one-line reason drawing on the quant data and
interview themes. This turns "here's what happened" into "here's whether
it's working" — the actual evaluator question.