datamatters24's picture
Add full CFPB summary DB + dataset card
880dcbf verified
|
Raw
History Blame Contribute Delete
2.63 kB
---
license: cc0-1.0
task_categories:
- tabular-regression
- time-series-forecasting
tags:
- cfpb
- consumer-finance
- credit-card
- complaints
- finance
pretty_name: Consumer Harm CFPB Summary
size_categories:
- 1K<n<10K
---
# Consumer Harm — CFPB Summary
Aggregated [CFPB Consumer Complaint Database](https://www.consumerfinance.gov/data-research/consumer-complaints/) summaries for the **Consumer Harm** story dashboard (credit cards: relief vs. the void).
## Study windows (fixed calendar)
| Window | Range |
|--------|--------|
| Study | 2011-01-01 → 2024-12-31 |
| Early comparison | 2011–2017 |
| Recent comparison | 2018–2024 |
About **7.2 million** complaints fall in the study window (built from the full CFPB bulk CSV).
## Contents
| File | Description |
|------|-------------|
| `cfpb_summary.db` | Complete SQLite aggregate DB (~4 MB) |
| `SOURCE.txt` | Download provenance notes (when present) |
### Tables in `cfpb_summary.db`
| Table | Rows |
|-------|------|
| `channel_by_product` | 112 |
| `credit_card_companies` | 1,107 |
| `credit_card_company_monthly` | 13,456 |
| `credit_card_issue_monthly` | 3,778 |
| `credit_card_issue_response` | 299 |
| `credit_card_issues` | 58 |
| `credit_card_monthly` | 157 |
| `credit_card_states` | 64 |
| `credit_card_subissue_monthly` | 7,745 |
| `credit_card_subissues` | 124 |
| `credit_card_subproducts` | 8 |
| `credit_card_theme_monthly` | 878 |
| `meta` | 11 |
| `product_monthly` | 1,458 |
| `product_stats` | 21 |
| `response_by_product` | 108 |
### Build meta
```
{'study_start': '2011-01-01', 'study_end': '2024-12-31', 'early_end': '2017-12-31', 'recent_start': '2018-01-01', 'study_label': '2011–2024', 'early_label': '2011–2017', 'recent_label': '2018–2024', 'period_mode': 'fixed_calendar', 'built_at': '2026-07-25T17:32:35.644409', 'total_complaints': '7231945', 'source_csv': '/home/grimdolf/opportunity_harm/cfpb/extracted/complaints.csv'}
```
## What is not included
The raw ~8 GB `complaints.csv` is omitted — download from CFPB. This dataset is the analysis-ready summary used in production.
## Live demo
- Landing: https://consumer-harm.vercel.app
- Dashboard: https://consumer-harm-production.up.railway.app
- Code: https://github.com/tedrubin80/consumer-harm
## Load example
```python
import sqlite3
import pandas as pd
conn = sqlite3.connect("cfpb_summary.db")
products = pd.read_sql("SELECT * FROM product_stats", conn)
cc = pd.read_sql("SELECT * FROM credit_card_companies ORDER BY complaints DESC LIMIT 20", conn)
```
## License
Aggregates packaged as CC0. Underlying CFPB data remains subject to CFPB terms of use.