# Ethiopia Financial Inclusion Forecast
Forecasting system for Ethiopia's digital financial transformation, built for a consortium
of development finance institutions, mobile money operators, and the National Bank of
Ethiopia. The project tracks and forecasts two Global Findex pillars:
- **Access** — Account Ownership Rate (`ACC_OWNERSHIP`)
- **Usage** — Digital Payment Adoption (proxied via `USG_P2P_COUNT`, `ACC_MM_ACCOUNT`,
`USG_TELEBIRR_USERS`, and related transaction indicators)
## Project Status
| Task | Description | Status |
|------|-------------|--------|
| 1 | Data Exploration & Enrichment | ✅ Complete (merged from `task-1`) |
| 2 | Exploratory Data Analysis | ✅ Complete (merged from `task-2`) |
| 3 | Event Impact Modeling | ✅ Complete |
| 4 | Forecasting Access & Usage | ✅ Complete |
| 5 | Dashboard Development | ✅ Complete |
## Data & Schema
This project uses a **unified schema**: every record (`observation`, `event`, `impact_link`,
`target`) lives in the same table and shares the same columns. The `record_type` column
determines how each row should be interpreted.
**Key design principle: events are not pre-assigned to pillars.** A launch like Telebirr
affects both Access and Usage; forcing it into one pillar would bias the analysis. Instead:
| Record Type | `category` | `pillar` |
|---|---|---|
| `observation` | empty | ✅ measured dimension |
| `target` | empty | ✅ goal dimension |
| `event` | ✅ event type (policy, product_launch, ...) | empty — no pre-assignment |
| `impact_link` | empty | ✅ pillar of the *affected* indicator |
`impact_link` records connect an `event` to an indicator via `parent_id` (the event's
`record_id`), capturing the estimated direction, magnitude, lag, and evidence basis of the
effect. See `data/raw/README.md` for the full schema documentation
## Repository Structure
```
ethiopia-fi-forecast/
├── data/
│ ├── raw/ # Starter dataset + reference codes (as provided)
│ └── processed/ # Enric …