Methodology
v1.0 · Last updated April 2026This site visualizes signal convergence — the overlap of multiple independent indicators that a county is being targeted for ICE detention facility expansion. A single signal means little; multiple signal types in the same county indicate a pattern.
How the score works
Each county receives a heat score based on three components:
- Weighted signals: Each entry is scored by its signal type (see weights below). Higher weights mean more predictive signals.
- Per-type caps: Each signal type is capped so that, e.g., 20 287(g) entries don't outscore 1 IGSA + 1 commission vote. The cap limits how many entries of one type contribute.
- Convergence bonus: +10 points per signal type beyond 2, and +15 more if 5 or more types are present. This is the key insight: three different signal types in one county is far more predictive than ten entries of the same type.
Worked example: Broward County, FL (Score: 129)
Note: Actual score may vary slightly as the heat score script counts all entries with matching FIPS, including entries that may have been added since this example was written. View Broward County →
Signal types and weights
| Signal | Weight | Cap | Description |
|---|---|---|---|
| IGSA Facility | 10 | 5 | Existing Intergovernmental Service Agreement for ICE detention |
| ANC Contract | 8 | 3 | Alaska Native Corporation federal contract in detention/security |
| 287(g) Agreement | 7 | 3 | Local law enforcement deputized for immigration enforcement |
| Commission Activity | 7 | 5 | County commission agenda item, vote, or hearing on detention |
| Job Posting | 7 | 3 | Detention consultant or contractor recruiting for this area |
| Sheriff Network | 6 | 3 | Sheriff conference attendance, pitch to commissioners |
| Comms Discipline | 6 | 3 | Polished messaging, NDA citations, opposition framing |
| Budget Distress | 5 | 2 | County budget shortfall making it vulnerable to the pitch |
| Real Estate Trace | 2 | 2 | Warehouse or county building that could be converted |
| Legislative Trace | 1 | 2 | State legislation to block or enable detention agreements |
Coverage depth
Each county page shows a coverage badge indicating how thoroughly it has been investigated:
You can help move counties from "automated only" to "well researched" by investigating locally and submitting what you find. The FOIA request generator produces ready-to-send public records requests for any county.
The thesis
Detention consultants target counties with budget distress, cooperative sheriffs, and existing infrastructure. The pitch follows a playbook: sheriff conference recruitment, closed-session presentations, NDA-protected negotiations, then a commission vote framed as fiscal salvation.
This site surfaces the signals early — when multiple independent indicators converge on the same county, that county is likely being pitched. Early detection enables democratic response. See the county fights section for documented cases.
Data sources
| Source | Data | Accessed | Count |
|---|---|---|---|
| Vera Institute of Justice | IGSA facilities, FIPS-coded | Mar 2026 | 459 |
| Prison Policy Initiative | 287(g) agreements, compiled from ICE data | Feb 17, 2026 | 1,311 |
| USAspending.gov | ANC subsidiary federal contracts (API) | Apr 2026 | 244 |
| Legistar API | County commission agendas and minutes | Ongoing | 3 |
| USDA ERS | County typology and economic distress indicators | Mar 2026 | 34 |
| Public career pages, LinkedIn | Detention consultant job postings | Ongoing | 6 |
| Local news, FOIA, court records | Commission activity, sheriff networks, comms patterns | Ongoing | 7 |
| The RAMM | Investigative reporting on detention pipeline — contractor profiles, revolving door, warehouse purchases | Ongoing | — |
| U.S. Census Bureau | County FIPS codes (2020) | 2020 | 3,235 |
Disclosure
This site is published by The RAMM. Some entries cite The RAMM's own investigative reporting as a source — these citations are marked "(publisher)" in the source list. All investigative findings cited are published and publicly accessible. Leaked documents referenced (e.g., the Sabot Consulting briefing package) were obtained and published by independent journalists; the KB entries cite the published articles, not the raw documents.
Limitations
This is an early-warning system, not a confirmed list. A high score means multiple independent signals converge — it does not mean a facility deal is confirmed or inevitable. False positives are expected. Some counties with active pipeline work may score low if their signals haven't been captured yet.
287(g) FIPS resolution covers ~79% of agencies. Municipal police departments (as opposed to county sheriffs) often can't be reliably mapped to counties without geocoding. ANC contract data depends on USAspending.gov's completeness. Commission activity and other human-researched signals are limited by contributor coverage — most counties have only automated signals.