ICE’s Own Roster Shows Kentucky at 55 287(g) MOAs Across 53 Agencies — Growth Is Voluntary-Plus-Financial-Incentive, Not the Failed Legislative Mandate

The count, done directly against the primary source

ICE publishes a downloadable roster of every signed 287(g) Memorandum of Agreement (MOA), agency by agency, with a signing date for each. The roster is served from a link on ICE’s participating-agencies page styled as a PDF (.../file-download/download/public/208609) but is actually an Excel workbook. Retrieved and parsed directly on August 30, 2026 (the page’s own as-of timestamp reads “August 30, 2026, 9:39am”):

• National total: 2,411 MOAs across 39 states and 2 U.S. territories — broken down as 1,682 Task Force Model (TFM), 547 Warrant Service Officer (WSO), and 182 Jail Enforcement Model (JEM) agreements. These three sub-totals sum exactly to 2,411, and my independent parse of the roster’s 2,411 non-blank rows reproduces the same three figures, confirming the parse is accurate against the page’s own stated breakdown.
• Kentucky: 55 MOA rows across 53 distinct law-enforcement agencies (two agencies — Grayson County Detention Center and Oldham County Detention Center — each hold two separate model MOAs, which is why rows exceed distinct agencies). By model: 50 Task Force Model, 3 Jail Enforcement Model, 2 Warrant Service Officer. The three JEM signatories are Grayson County Detention Center, Oldham County Detention Center, and Kenton County Sheriff’s Office.

This 287(g) roster is a different dataset from Kentucky’s ICE detention-contract (IGSA) count. The “11 county jails” figure that appears in the task framing and in KY Center for Economic Policy’s reporting refers to jails holding ICE detainees under separate detention/IGSA-type contracts — tracked in this KB’s kentucky-ice-detention-overview-2025-2026 note — not to 287(g) law-enforcement-authority agreements. Only 3 of Kentucky’s 53 287(g) agencies hold the Jail Enforcement Model variant that authorizes jail-based immigration processing; the other 8 of the 11 detaining jails hold ICE detainees under a distinct contractual authority with no corresponding 287(g) MOA in this roster. Headline counts that merge “287(g) participants” and “ICE-contracting jails” into one number are combining two different legal instruments.

The count is a floor, not a ceiling, and grows continuously. Sorting Kentucky’s 55 rows by signing date: the earliest is Grayson County (March 5, 2025); the most recent in this pull is Simpson County Sheriff’s Office (August 17, 2026) — thirteen days before this retrieval. Of the 55 signings, 19 occurred in 2025 and 36 occurred in 2026 (through Aug. 17) — meaning the pace of new signings roughly doubled year over year, and the “36” figure that appears in the KB task framing (drawn from an earlier KY Policy snapshot cited as “36 LE agencies”) is coincidentally close to, but not the same measurement as, “36 agreements signed in 2026” found here. The KY Policy article’s own figure at the time of the original task (an earlier pull) read 24 agencies/30 contracts; a more recent KY Policy page revision, extracted the same day as this roster pull, still shows the 24-agency figure as its most recent stated number — meaning the advocacy org’s public-facing count is now roughly 18 months stale relative to ICE’s own roster. Always count from the primary roster and date the count; a citation to any secondary tally, including this KB’s own prior entries, will already be out of date by the time it is read.

Mechanism: voluntary participation, structured by a strong federal financial incentive — not a mandate

Kentucky’s 287(g) expansion is voluntary at the point of signing: individual sheriffs, police chiefs, and jailers choose whether their agency enters an MOA, and Kentucky’s Oldham County Sheriff’s Office is a documented example of an agency that publicly declined for a period (see kentucky-ice-detention-overview-2025-2026, “Note on Oldham County”) before ultimately signing a Task Force Model MOA on May 5, 2026 per this roster. No Kentucky statute currently requires any agency to sign. That distinguishes Kentucky’s actual, operative mechanism from HB47 and SB86 — the two 2026-session bills that would have made 287(g) participation compulsory (“require all Kentucky State Police posts,” in HB47’s case, and all state and local agencies in SB86’s) — both of which died without floor votes when the Kentucky General Assembly adjourned sine die in mid-April 2026 (see kentucky-hb47-mandatory-287g-legislation, this KB). The word “voluntary” in the task framing is accurate as far as it goes, but incomplete, for the reason the April 28 South Carolina entry (2026-04-28--graham-blackburn-287g-expansion-act-sc-h4764-not-a-companion.md) already established for a different state: a strictly voluntary label can obscure a strong financial lever operating on that voluntary choice.

That lever is documented, second-hand from primary reporting rather than independently re-derived from statutory text here: per KY Center for Economic Policy (citing the “One Big, Beautiful Bill Act,” OBBBA, effective Oct. 1, 2025) and corroborated by NPR/KPBS (May 5, 2026), DHS now fully reimburses participating agencies for trained 287(g) officers’ annual salary and benefits (including overtime up to 25% of salary) and pays quarterly “performance bonuses” — up to $1,000 per eligible task-force officer for agencies that locate 90-100% of ICE-identified undocumented immigrants, scaling down for lower rates — funded through the same package that appropriated the roughly $75 billion in interior immigration-enforcement funding NPR and KY Policy both report. That is a national, not Kentucky-specific, structure — it applies to any agency that signs, in any of the 39 participating states. This is functionally the identical instrument S. 4405 (the federal “287(g) Expansion Act”) proposed to formalize nationally — increased federal funding conditioned on 287(g) participation — except OBBBA’s reimbursement/bonus mechanism is already in effect and already operating on Kentucky’s 2026 signing surge, while S. 4405 (introduced April 28, 2026, still pending in Senate Judiciary as of the SC entry’s last check) would add a further JAG-grant incentive on top of it.

So: voluntary is technically true and substantively incomplete. No Kentucky agency is legally compelled to sign. But every agency that signs after October 1, 2025 does so against a backdrop of full salary reimbursement plus a bounty-structured bonus scheme — a federal incentive strong enough that KY Policy’s analysis explicitly likens it to a bounty. The correct characterization, following the SC entry’s discipline, is voluntary participation under a strong federal financial incentive, distinct from both a true no-strings-attached voluntary regime and a legal mandate. Function, not stated intent: what the arrangement converts is discretion (a sheriff’s choice whether to enter local immigration enforcement) into a revenue-and-staffing decision, by attaching salary coverage and bonus pay to the “yes” answer — regardless of whether any individual sheriff’s stated reason for signing is public safety, cost relief, or ideological alignment with the enforcement push.

Kentucky’s legislative mandate track remains dead, not merely dormant

For completeness against the “bill is not a law” discipline: HB47 (primary sponsor Rep. TJ Roberts, R-Burlington; introduced Jan. 13, 2026; referred to House Judiciary Jan. 21, 2026) and SB86 (10 sponsors) both received no floor vote in either chamber before the Kentucky General Assembly’s 2026 Regular Session (26RS) adjourned sine die in mid-April 2026. No news source located between the existing KB entry’s last check (May 6, 2026) and this pass (Aug. 30, 2026) reports a revival, pre-filing, or committee action for a 2027 session equivalent. Status: introduced, referred to committee, stalled — not enacted, not law. The mandate track and the voluntary-plus-incentive track are, as the SC entry found for the federal/state pair, two different instruments pursuing the same policy goal on parallel tracks — not sequential stages of one laundering pipeline. Kentucky’s 2026 signing surge (36 of 55 total MOAs) happened without the mandate passing, which is itself evidence that the financial incentive alone is sufficient to drive rapid voluntary uptake — the mandate bills would have made mandatory what the incentive structure is already producing voluntarily.

Why this belongs on detention-pipeline as its own entry

The task’s underlying instinct — that “voluntary expansion” is structurally distinct from “legislative mandate” and deserves separate tracking — holds up under primary-source testing, with the correction that “voluntary” needs the financial-incentive qualifier attached every time it’s used, and that the 287(g)-agreement count and the ICE-detention-jail count are two different rosters that should not be added together or substituted for one another in headline figures.

Research Gaps

• KY Policy’s most recent published page (extracted same day as this roster pull) still shows a 24-agency figure; unclear whether that reflects a stale unrevised page or a narrower definitional scope (e.g., counting unique agencies signed before a specific cutoff). Not resolved here — flagged for a future KY Policy methodology check rather than assumed.
• Whether HB47 or SB86 (or successor bills) will be pre-filed for Kentucky’s 2027 legislative session is unknown as of this pass; the 2025-2026 biennium technically remains open per the SC entry’s precedent reasoning about two-year general assembly cycles, though Kentucky’s assembly structure should be confirmed independently rather than assumed identical to South Carolina’s.
• The 8 of 11 ICE-contracting Kentucky jails that hold no corresponding 287(g) JEM MOA in this roster were not individually re-verified against the jail-contract list in this pass; cross-referencing the two Kentucky datasets (287(g) roster vs. IGSA/detention-contract list) by facility name is a natural follow-up task.

Related Entries

• 2026-04-28–graham-blackburn-287g-expansion-act-sc-h4764-not-a-companion (the “voluntary vs. incentivized vs. mandatory” distinction was first established here for South Carolina/federal; this entry applies the same discipline to Kentucky)
• kentucky-hb47-mandatory-287g-legislation (detention-pipeline KB — the dead mandate track)
• kentucky-ice-detention-overview-2025-2026 (detention-pipeline KB — the separate 11-jail ICE detention-contract dataset, and full Kentucky enforcement context)

Conductor QC, 2026-08-31 — the roster was re-downloaded and re-counted independently, and every figure reproduces.

I pulled ICE’s own file from the same endpoint (ice.gov/file-download/download/public/208609, HTTP 200, 273,956 bytes, confirmed Microsoft Excel 2007+ despite the .pdf-styled link), unzipped the XLSX and parsed the sheet XML directly:

The national breakdown matching is the strongest validation available — it means the parse covered the whole file, not a filtered subset, so the Kentucky figures are counted from the same complete roster ICE publishes.

This is what “parse the primary list, do not cite anyone’s total” produces. The entry demonstrates the cost of the alternative in the same breath: KY Center for Economic Policy’s public page still shows 24 agencies — roughly 18 months stale against the primary roster. Citing that secondary tally would have understated participation by more than half.

The dataset conflation in the ticket’s own framing is correctly caught. “36 LE + 11 jails” merges two different ICE datasets — the 287(g) agreement roster and the separate detention/IGSA jail-contract list. Only 3 of Kentucky’s 53 agencies hold a Jail Enforcement Model MOA. Reporting the three support types separately, rather than as one headline number, is what makes this usable: they are materially different authorities.

The “voluntary” test is the finding, and it holds the SC precedent’s triad. Voluntary at signing — no statute compels it, and HB47/SB86 both died without a floor vote. But incomplete as a label: DHS’s OBBBA-funded salary reimbursement plus performance bonuses took effect 2025-10-01, and 36 of 55 Kentucky signings occurred in 2026 alone versus 19 in 2025. The pace roughly doubled after the incentive landed, with no mandate ever passing.

That is a second state confirming the mechanism triad established on South Carolina today — voluntary / incentivized / mandatory — and it points the same way: the incentive track alone is sufficient; the mandate track keeps failing. Two states is a pattern worth naming and not yet a national claim.

Date kinds held: roster retrieved 2026-08-30 (stated, as it must be — this file changes continuously), incentive effective 2025-10-01, signings counted by signing year.