What Killed 27,000 Fewer Americans in 2024 — and Why Washington's Answer Matters Enormously

The Kentucky Harm Reduction Coalition, a small nonprofit in Louisville, learned on a Thursday in late April that the federal government had reconsidered what counts as acceptable public health spending. The organization — which has been distributing fentanyl test strips, clean syringes, and naloxone to people who use drugs in the region — lost its $400,000 SAMHSA grant after the agency issued new guidance on April 24 prohibiting federal funds from purchasing test strips and drug-use supplies. The coalition had roughly one month of operational reserves remaining. The email that delivered the news was six lines long.

The timing is the point. That notification arrived as the Centers for Disease Control and Prevention was releasing data confirming the steepest single-year drop in American overdose deaths since the federal government started counting them. From 2023 to 2024, deaths fell from 105,007 to an estimated 79,384 — a 24 percent decline. The sharpest one-year reduction in recorded history. The administration that cut the test strip funding is operating with an implicit theory about what caused that decline. The stakes of being wrong are exactly the number of people who might not make it through the next year.

The numbers that changed everything — and the fight over why

Overdose deaths in the United States peaked at roughly 110,000 in 2022. They have now fallen to an estimated 70,000 in the 12 months ending December 2025 — a 36 percent drop in three years. In absolute terms, that is tens of thousands of people who did not die from drug overdoses who would have, had the prior trajectory held. The scale of the reversal is difficult to overstate. For two decades, American overdose deaths rose almost without interruption. The line was essentially a wall. Then, sometime in late 2023, it bent.

There are two schools of thought on why, and they are not compatible with each other in their policy implications even if they’re not mutually exclusive in their mechanisms.

The first explanation centers on harm reduction and treatment infrastructure: naloxone distribution, fentanyl test strips, expanded access to buprenorphine and methadone, syringe service programs, low-barrier walk-in services. The argument is that years of cumulative investment in these tools reached a tipping point, and the body count reflected it. The Drug Policy Alliance noted in early 2026 that harm reduction tools “kept tens of thousands of people alive” and pointed to the expanding use of treatments that cut overdose risk roughly in half. More people than ever are in treatment with buprenorphine or methadone. More communities have naloxone in schools, pharmacies, and libraries. The infrastructure argument says this shows up in the data because it had to.

The second explanation looks at what changed in the drug supply itself. A December 2025 preprint by researchers published on medRxiv and covered by STAT News in January 2026 found that the 2024 decline was almost entirely confined to fentanyl-involved deaths — a pattern that is more consistent with a change in fentanyl’s toxicity or availability than with the broad-spectrum effects of expanded treatment and naloxone. “Access to treatment medications did not expand suddenly in either the U.S. or Canada,” the preprint noted. “Naloxone did not see a major jump in use.” Harm reduction services did not markedly increase. Something happened in the illicit fentanyl supply — likely related to international precursor chemical controls — that, for a period, made the drug less deadly. The decline followed.

The administration has read that research and drawn a conclusion.

The supply-side case: when the drug gets less potent, fewer people die

The supply-side argument is not a conservative talking point dressed up as epidemiology. The underlying data is solid. The 2024 decline was concentrated in synthetic opioid deaths; deaths from stimulants, alcohol, benzodiazepines, and other drug classes did not drop proportionally. If harm reduction infrastructure were the primary driver — if naloxone distribution and buprenorphine access had reached a scale that bent the curve — you would expect a broader effect across substance classes. People who overdose on opioids and people who die in alcohol emergencies and people in stimulant psychosis crises share overlapping service populations, overlapping communities. Infrastructure that genuinely penetrates those communities tends to show up across categories. The 2024 data shows a fentanyl-specific decline.

People who overdose on opioids and people who die in alcohol emergencies and people in stimulant psychosis crises share overlapping service populations, overlapping communities.

The implication of the supply-side theory is that the current improvement is conditional. It holds as long as the fentanyl supply remains less potent or less available. When the supply changes — new precursor routes, new analogues, new manufacturing chemistry — the conditions change. The infrastructure argument says: if we had built durable systems, the next supply shift would find us better equipped to handle it. The supply-side argument says those systems didn’t actually bend the curve, so the investment was misallocated.

This is where the policy divergence becomes concrete. If you believe the supply-side explanation, then SAMHSA funding for fentanyl test strips was marginal spending on a marginal intervention in a problem being driven by chemistry. Cutting it is not a tragedy — it’s a reallocation. If you believe the harm reduction explanation, or even a mixed explanation that credits both factors, then cutting that funding is a serious error with predictable human costs.

The harm reduction case: five times more likely — and why the math runs deeper than survival

The pure survival data is only part of the harm reduction argument. The other part is what harm reduction services actually do beyond the immediate overdose reversal — and this is where population-level research from 2026 and earlier work on syringe service programs becomes consequential.

People who engage with syringe service programs are five times more likely to enter treatment for substance use disorder than nonparticipants, according to data cited by the National Association of Counties. Not because syringe programs are treatment programs — they explicitly are not — but because they represent non-judgmental, low-barrier contact with a health system that most people who use drugs have learned, often through repeated bad experience, to distrust.

This is the psychological reality of harm reduction that the supply-side framing misses: for someone in active use, the distance between where they are and a treatment program is not measured in miles. It is measured in shame, in previous encounters with providers who made them feel like a problem to be managed rather than a person to be helped, in uncertainty about insurance, in fear of legal consequences, in the simple cognitive load of crisis. A syringe service program that asks nothing except presence — no insurance, no sobriety requirement, no paperwork — becomes for many people the first point in a chain that eventually leads somewhere different.

When you close that service point, you do not eliminate the need. You eliminate the entry. The Kentucky Harm Reduction Coalition wasn’t just distributing test strips; it was the place where, for some subset of its clients, the conversation that eventually led to treatment began. That conversation doesn’t have a line item in SAMHSA’s budget request. It is harder to count than a box of naloxone. But it is real.

Los Angeles County, which has invested heavily in this kind of infrastructure, saw a 22 percent decline in drug-related overdose deaths in 2024 and a 37 percent drop specifically in fentanyl deaths — the most significant decline in the county’s history. The county’s public health department credited a “multifaceted public health approach.” It did not credit a change in fentanyl supply potency, though supply-side factors were likely present there too.

The American Society of Health-System Pharmacists — a professional organization of pharmacists, not a harm reduction advocacy group — formally opposed the SAMHSA restrictions in April 2026. “The evidence is clear that syringe services programs save lives and do not increase drug use,” ASHP wrote. When pharmacy professionals whose members dispense the medications that treat addiction formally align themselves against a federal drug policy decision, the scientific consensus is not on the side of that decision.

What the April 24 guidance actually does — and what it costs

The SAMHSA guidance issued April 24 is specific: it prohibits federal grant funds from purchasing fentanyl test strips, xylazine test strips, medetomidine test strips, syringes, and smoking kits. The July 2025 guidance had made an exception for fentanyl test strips, framing them as still permissible. The April 2026 letter reversed that exception.

The July 2025 guidance had made an exception for fentanyl test strips, framing them as still permissible.

The reasoning offered by SAMHSA — that federal funding should support “comprehensive treatment rather than medication-only models” — does not survive contact with what test strips actually are. Test strips are not a medication-only model. They are a risk-reduction tool used by people who are going to use drugs regardless of federal policy preferences. A person who uses a fentanyl test strip and discovers their supply is contaminated, and who changes their behavior in response — uses less, uses in the presence of someone who has naloxone, decides not to use at all that day — is not receiving treatment. They are surviving long enough to eventually have the choice.

The organizations that lose grants cannot simply substitute other funding on short notice. Foundation grants have 12-to-18-month application cycles. State dollars vary dramatically by jurisdiction. The $400,000 the Kentucky Harm Reduction Coalition lost wasn’t a rounding error — it was the operational core of a program that served a specific population in a specific geography. When that program closes or contracts, the people it served don’t automatically migrate to another service point. They often disappear from the system entirely.

Arizona is watching — and the calculation is concrete

In Arizona, where the state ranks 49th out of 51 in behavioral health access and where fewer than 1 in 20 Arizonans with opioid use disorder receives buprenorphine or methadone, the federal funding shift lands with particular weight. The state committed $1.215 billion in opioid settlement funds over 18 years to the recovery infrastructure — a sum that represents one of the most significant investments in addiction treatment in state history. AHCCCS spent $582 million on substance use disorder treatment in state fiscal year 2025 alone. The State Opioid Response IV grant brings in $34.8 million per year. That is substantial investment in treatment capacity.

But treatment capacity is not the same as treatment access. The people who would use Arizona’s growing treatment infrastructure first have to find their way to it. Harm reduction programs — the syringe exchanges, the test strip distribution vans, the naloxone points — are, for many people, the path. Cut those programs, and Arizona’s $1.2 billion commitment to recovery becomes a system waiting for people who never arrived.

Five Arizonans die from opioid overdoses every day. That rate has been declining. Whether it continues to decline depends in part on whether the infrastructure that connects the most isolated, most underserved people to the health system remains funded.

The administration has made its bet

The SAMHSA guidance of April 24, 2026, reflects a theory about what caused the overdose decline and what it will take to sustain it. The theory is that supply-side factors drove the improvement, and that harm reduction infrastructure was at best neutral and at worst counterproductive. It is a theory with real evidence behind it. It is not the only theory with real evidence behind it.

What is not in dispute: the organizations that distributed fentanyl test strips were reaching people who use drugs in circumstances where almost nothing else reaches them. The list of 47 names a harm reduction worker kept in her notebook — people who checked a strip, found contamination, and made a different choice — is a real list. It exists in cities across the country, in a hundred different forms, in the records of programs that are now in funding crises.

The supply of fentanyl will change again. It always has. The infrastructure question — who is positioned to meet people where they are when that change arrives — is not an abstraction. It is a bet on whether the 27,000 people who didn’t die in 2024 represent a sustainable shift or a temporary respite. The administration has placed its bet. The people in the service area of the Kentucky Harm Reduction Coalition, and programs like it across the country, are living with the consequences of the wager.

The infrastructure question — who is positioned to meet people where they are when that change arrives — is not an abstraction.

---

The harm-reduction category covers evidence-based overdose prevention tools, syringe service programs, naloxone access, and drug-checking services. For the opioids substance pillar and Arizona coverage, follow those sections for ongoing updates.