Science1 distinct publisher3 min readPublished
A five-month records hunt across all 64 Louisiana sheriffs produced the first detailed accounting of the carve-out, and the two tests applied to that spending both came from outside the state.
The Scientist · Science desk

Compiled by The ScientistSomething wrong?How this is made
A Louisiana sheriff runs his own budget and does not serve at the pleasure of any other local official, and no rule requires him to tell the public how he spends opioid settlement cash [2]. The state's five-member Opioid Abatement Task Force advises sheriffs and parishes but controls none of the money, and Louisiana wrote no list of its own naming purchases the funds may not cover, though the task force says it has passed along other states' guidance [9].
That is why both failing grades in this accounting were imported. One came from a panel of three the newsrooms assembled: a Louisiana man in recovery whose son died of a fentanyl overdose, an addiction medicine doctor, and a public health policy analyst who has tracked settlement spending since 2022 [7]. They judged $5.4 million of the $8.1 million reported to be inappropriate, roughly 66%, including salaries and overtime for homicide detectives and for officers running jail shakedowns for contraband [6][11]. The second test was also external: $4.7 million went to items that appear on the unallowable lists of six other states, among them cellphone data extraction technology and automated external defibrillators, which are rarely needed in an opioid overdose [8].
Those are two different tests, and the reporting does not say one sits inside the other. But both apply to the same $8.1 million, so at least $2 million of the disclosed spending failed both: money three reviewers called wrong and that other state governments have already written down as off limits [1]. That is the figure carrying the least discretion, and it is a floor, not an estimate.
The tempting comparison is $8.1 million disclosed against the $10.7 million that twenty non-responding sheriffs were allotted [5][10], but the two numbers measure different things. One is money out the door, the other is money assigned, and a sheriff may be sitting on much of his share. Jefferson Parish's allotment alone, nearly $4 million through 2025, is about 37% of the total kept from view [c10b][2].
Nine sheriffs came through both tests intact, with close to $1.8 million spent on treatment inside jails, on training officers to respond to overdoses, and on telling residents where help exists [12]. That is about 22% of reported spending [3], which is the practical answer to whether handing settlement money to elected law enforcement produces abatement: sometimes, at the discretion of the individual officeholder, with no mechanism to make the good cases the standard. With Purdue Pharma's $7.4 billion bankruptcy finalized this year and settlements paying out for another decade and more [13][14], the sheriffs' share of what has not yet arrived travels the same route as the share nobody could audit.
Ranked by verification strength, evidence, and original report placement.
The review panel included a Louisiana resident in recovery who lost his son to a fentanyl overdose, an addiction medicine doctor, and a public health policy analyst who has tracked settlement dollars since 2022; they judged expenditures using their professional and personal expertise.
Louisiana is the only state directing 20% of its opioid settlement payouts to sheriffs, the largest carve-out for law enforcement nationwide.
Louisiana sheriffs are elected, do not serve at the pleasure of another local official, have independent budgets, and are not required to proactively report to the public or another authority how they spend opioid settlement cash.
Short of public records requests or waiting for official audits, it is difficult for the average person to track the Louisiana sheriffs' settlement money.
KFF Health News worked with The Current, Gulf States Newsroom, and Verite News, contacting all 64 Louisiana sheriffs over five months and often filing public records requests, to produce the first detailed accounting of the spending.
Thirty-eight sheriffs reported spending more than $8.1 million in total.
Follow any of these and your For You feed starts watching them — no settings page required.
Evidence-backed comparisons of source perspectives and observed adoption signals. Read the methodology
Which Builder, Operator, and Investor concerns the observed source mix emphasized—not a truth score.
Evidence, demonstrated adoption, hype gap, incentives, and confidence are assessed independently, each on its own current evidence. How these are measured.
Strong primary records reporting, single publisher
The findings rest on original documentary reporting: all 64 sheriffs contacted over five months, frequent public records requests, and itemized dollar totals that reconcile internally ($8.1 million reported, $5.4 million panel-disapproved, $4.7 million unallowable elsewhere, $1.8 million fully approved, $10.7 million undisclosed). Methodology and panel composition are disclosed. Evidence is capped below the top band because the cluster contains a single publisher observation with no independent replication, the evaluative tests are newsroom-constructed rather than audit findings, and the supplied body is truncated before the full response from sheriffs.
Money demonstrably flowing, disclosure only partial
Real-world use is documented rather than announced: more than $8.1 million already spent by 38 offices, allotments running through 2025, and nearly $1.8 million spent in ways that pass both tests. Adoption is held mid-range because roughly $10.7 million across 20 offices is undisclosed, so the observed usage covers only part of the carve-out and no statewide carve-out total is given for denominator context.
Slightly overstated by framing, numbers hold
The dollar findings are documented and conservatively sourced, but the headline percentages (66% and nearly 60%) are shares of disclosed spending only, while roughly $10.7 million remains unseen, and the 'inappropriate' and 'unallowable' labels come from a newsroom-assembled panel and other states' lists rather than from Louisiana law or an audit - the same story notes the state agreement expressly allows law enforcement expenditures relating to the epidemic. That gap between normative test and legal standard tilts the framing modestly beyond what the evidence strictly establishes; the underlying figures are not overstated.
Strong competing stakes, largely disclosed
The source makes the incentive structure explicit on multiple sides: sheriffs hold independent budgets strained by the crisis and argue seizures and arrests save lives; recovery advocates oppose any flow to what they call a failed drug war; the settlement pool is described as a pot of gold many are vying for; the advisory task force has influence but no control; and the reviewing newsroom both assembled the evaluative panel and stands to gain from an accountability finding, with panelists' personal and professional stakes named. Score reflects high incentive intensity coupled with good disclosure rather than hidden conflicts.
Well-documented findings, one publisher
Confidence is above the midpoint because the numeric claims are internally consistent, method and evaluators are disclosed, and non-response is reported rather than glossed. It is held back by single-publisher sourcing with no independent confirmation, a truncated source body, the absence of a statewide carve-out denominator, and reliance on normative tests imported from outside Louisiana.
science
HIPAA Covers Less Than You Think, And "Anonymized" Is Not A Legal Shield1 distinct publisher
science
US life expectancy hit 79 in 2024, and the cushion over 2019 is about 73 days1 distinct publisher
science
Hospitals want the $1 trillion Medicaid cut back. Finance chiefs must price both answers.1 distinct publisher
science
Coronary calcium earns one narrow job: breaking statin ties in the 3-to-10 percent band1 distinct publisher
Distinct publishers with included, body-backed reporting in this cluster.
1 article · August 27, 2026