Dropping Numbers: Drugs ODs, Homicides, Motor Vehicle Accident Deaths

Looking at absolute numbers and not rates for now

It is nice to have good mortality news: [I filtered on all opioid OD deaths specifically - mostly fentanyl — this is not all drug OD deaths]

Let me explain this graph, because some may misinterpret: this is not the number of fentanyl OD deaths per month in the U.S., but a rolling-12-month count, ending with the count in August 2025.

So the 21,000 death decrease is comparing the peak of 111K opioid OD deaths over July 2022 - June 2023 to the 69K deaths of September 2024 - August 2025.

The 69K deaths bring us back to a pre-pandemic count, last seen in October 2019. Again, to keep this post simple, I’m only doing death counts and not rates.

Various explanations

8 Jan 2026, Science: Did the illicit fentanyl trade experience a supply shock?

Abstract

Fatal overdoses from synthetic opioids, most notably fentanyl, steadily increased more than 25-fold in the United States over 15 years, peaking at 76,000 in 2023 (1). This trend began to sharply reverse in mid-2023, dropping the annual rate of fentanyl overdose deaths (ODDs) by over a third by the end of 2024 (1). Explaining this unexpected drop is of major scientific and policy interest. Whether a supply shock could account for a substantial part of the decline is challenging to determine because drug trafficking organizations operate in secret. Synthesizing data from the US and Canadian governments and from discussions on the social media platform Reddit, we suggest there was a major disruption in the illicit fentanyl trade, possibly tied to Chinese government actions, that translated into sharp reductions in overdose mortality beginning in mid- or late-2023 and continued into 2024 across both the US and Canada.

The footnote has a stale link to a CDC press release, which I found:

25 Feb 2025, CDC Press Release: CDC Reports Nearly 24% Decline in U.S. Drug Overdose Deaths

“It is unprecedented to see predicted overdose deaths drop by more than 27,000 over a single year,” said Allison Arwady, MD, MPH, Director of CDC’s National Center for Injury Prevention and Control. “That’s more than 70 lives saved every day. CDC’s public health investments, our improved data and laboratory systems for overdose response, and our partnerships with public safety colleagues in every state mean that we are more rapidly identifying emerging drug threats and supporting public health prevention and response activities in communities across America.”

While this national decline is encouraging news, overdose remains the leading cause of death for Americans aged 18-44, highlighting the importance of sustained efforts to ensure this progress continues. President Trump first declared opioid overdose to be a public health emergency in 2017, a designation that remains in place, and the subsequent public health investments to CDC from Congress have transformed the nation’s ability to use data to save lives. In the most recent data, 45 states showed declines in overdose deaths, but five states—Alaska, Montana, Nevada, South Dakota, and Utah—still saw increases in overdose deaths, highlighting the continued need for rapid local data and tailored response. In addition to the large provisional drop in fatal overdoses, we also see smaller decreases in nonfatal overdoses, as measured by emergency department visits for overdose, and welcome continued decreases in self-reported youth substance use.

Multiple factors contribute to the drop in overdose deaths, including widespread, data-driven distribution of naloxone, which is a life-saving medication that can reverse an overdose; better access to evidence-based treatment for substance use disorders; shifts in the illegal drug supply; a resumption of prevention and response after pandemic-related disruptions; and continued investments in prevention and response programs like CDC’s flagship Overdose Data to Action (OD2A) program.

A drop in the supply seems more credible than Narcan availability, given the huge drops in deaths.

Sorry, y’all, but what has often happened was that people were saved from their current OD event by Narcan… but would only have a fatal OD later. The bigger thing that saved them was just not having the fentanyl to kill them.

10 Jan 2026, Forbes, Jesse Pines: Lower U.S. Fentanyl Deaths Linked To China’s Policies, Study Says

What caused this dramatic change? A landmark study published this week in Science offers a compelling explanation. The evidence suggests the decline was not primarily driven by U.S. domestic policy. Instead, a major disruption in the global fentanyl supply chain appears to have played a substantial role, most likely originating with controls on precursor chemicals in China.

Why Treatment and Enforcement Can’t Explain the Drop In Fentanyl Deaths

When overdose deaths fell, some credited domestic policy interventions such as expanded naloxone access to reverse overdoses, medication-assisted treatment to bridge people off the addictive drugs and intensified law enforcement. Yet the timing and scale of the decline do not align with these domestic policy interventions.

The sharp drop beginning in mid-2023 occurred almost simultaneously in both the United States and Canada. This synchrony points to a shared, upstream effect, likely a shock to the drug supply from policies enacted in China. If U.S. border security or domestic enforcement were the primary drivers, Canada’s trend line would likely have remained largely unaffected. It did not.

Contemporaneous reporting from the U.S. Drug Enforcement Administration (DEA) also emphasized difficulties obtaining key precursor chemicals rather than claiming enforcement victories, further supporting a supply-side explanation.

….

When access to fentanyl became constrained, the shrinkage appeared in reduced purity. In fact, the average purity of seized fentanyl powder fell sharply from a peak of roughly 25% in mid-2023 to about 11% by the end of 2024. Fentanyl pills showed a similar, slightly delayed decline in potency.

Shrinkflation is a well-established economic indicator of supply shortages. Lower average purity reduces the chance of overdose risk at the population level, even as individual risk remains high due to unpredictable variability in street drugs.

….

Additional insight comes from an unconventional but increasingly valuable source: social media. Researchers analyzing discussions on drug-related Reddit forums found that mentions of fentanyl “droughts” and shortages spiked in mid-2023 and surged again later that year. Users reported that “fetty” (i.e. fentanyl) was harder to find and weaker than before. While not traditional surveillance data, these unfiltered signals closely tracked laboratory purity data, seizure trends and declining overdose deaths.

….

This timing coincides with stepped-up Chinese regulatory and enforcement actions in late 2023 and throughout 2024, following renewed diplomatic engagement between the U.S. and China. Chinese authorities reported shutting down online chemical marketplaces and removing tens of thousands of advertisements related to precursor sales.

Now, again, while they may credit foreign policy in engaging with China… I have been wondering whether China was having issues internally with fentanyl. They’ve already shown problems in controlling other things inside their own borders that perhaps was intended for foreign consumption only (as it were).

There was something noted as awry in the Science article:

8 Jan 2026, Washington Post, David Ovalle and Cate Cadell: A study offers a surprising reason for plunging U.S. overdose deaths

China agreed to internal restrictions on fentanyl-related substances during the first Trump administration. But that led to Mexican criminal groups synthesizing illicit fentanyl in secret labs in Mexico with precursor chemicals bought from companies in China. Since 2023, the Chinese government has shut down some of those companies as part of a broader crackdown.

….

The Science researchers caution that the precise scope of China’s crackdown is difficult to assess, given the opacity of enforcement in the country. China’s cooperation with U.S. drug authorities on fentanyl has long been fragile, often collapsing when broader tensions flare.

That changed ahead of a November 2023 summit between President Joe Biden and Chinese leader Xi Jinping, when the two governments agreed to launch a multiagency crackdown on Chinese chemical suppliers tied to the fentanyl trade. Chinese authorities subsequently arrested about 300 people and moved to restrict roughly 55 additional synthetic substances — steps Beijing had previously resisted.

The summit, however, happened months after overdose deaths had already begun to fall — a timing mismatch the researchers acknowledge. Humphreys theorizes China may have begun crackdowns months before the agreements were announced.

Again, China may have already started for reasons that have not much to do with the U.S. directly, but the fentanyl production caused trouble for the official Chinese Communist Party internally.

I am not seeing any of the people talking on the record considering this possibility.

“What’s really striking is that parallel across the two countries, even though the two countries [Canada and U.S.] have very different domestic policies,” said Jonathan P. Caulkins, a Carnegie Mellon University professor who researches the criminal drug trade and was a study co-author.

Yeah, it’s not about what Canada or U.S. did… which brings me to the next topic: homicides.

Homicides are down all over the U.S.

I’ve mentioned this a few times over 2025, especially in reference to Chicago:

In that post from September 2025, I pointed out that while yay! Chicago had its homicides way down for 2025… so did:

• Jackson, Mississippi - 50% decrease

• Birmingham, Alabama - 51% decrease

• St. Louis, Missouri - 28% decrease

• Memphis, TN- 25% decrease

• Baltimore, MD - 22% decrease

• Detroit, MI - 19% decrease

• Cleveland, OH - 46% decrease

• Louisville, KY - 30% decrease

• Indianapolis, IN - 30% decrease

• Oakland, CA - 21% decrease

Some people were getting cutesy, trying to say: oh, they’re just not reporting crimes!

Guys, it’s difficult with even the most corrupt police departments to get away with hiding dead bodies… especially so many across the U.S. Or misreporting homicides as different causes of death, given so many homicides are via gun or knife as a weapon.

But even if they were all lying about drops in homicides, there needs to be an explanation as to why they were all lying in concert.

Let’s check in with my favorite Chicago crime site, Hey Jackass!: The 2025 Year in Review

We speculated at the end of 2024 that “maybe the less cops, less tools, more rules has been the winning formula missing for all these years”. Possibly that along with a State’s Attorney that actually enforces the law has pulled off a decade low in hi-speed lead-related incidents despite the Kakistocratic leadership that runs the city and state.

2025: 433 killed, 1,592 wounded
2024: 613 killed, 2,444 wounded
2023: 650 killed, 2,500 wounded
2022: 740 killed, 2,936 wounded
2021: 856 killed, 3,745 wounded
2020: 800 killed, 3,472 wounded
2019: 521 killed, 2,291 wounded
2018: 592 killed, 2,465 wounded
2017: 685 killed, 2,935 wounded
2016: 808 killed, 3,658 wounded
2015: 513 killed, 2,546 wounded
2014: 465 killed, 2,239 wounded
Avg: 658 killed, 2,839 wounded

A breakdown of the year’s non-demonizable, silly decisions because they’ve been starved of opportunities in their own communities where law enforcement is a sickness that needs to be eradicated shows:

…..

We mistakenly believed we had bottomed out in 2024, only to watch the downward trend persist through 2025. In hindsight, we underestimated the significance of replacing a pro-criminal State’s Attorney with one who simply enforces the law as written (despite pro-criminal judges still being a persistent issue).

And here is their graph of homicides by year in Chicago:

If 2014 were included, the count would still be lower than 2014.

I think the owner of the site has the cause-effect wrong, because pretty much everybody does. They think that something that the local politicians and cops did had something to do with the big drop.

If it happened everywhere across the U.S., what do you think?

Unfortunately, I will not have good data for 2025 homicides until about July 2026, because of how external causes of death are released in data from the CDC.

Total Shot trend in Chicago, from Hey Jackass!

But I do have good data through June 2025, and you can see there was already a separation in shootings in Chicago by the halfway mark each year. I may be able to do an analysis to see the large drop year-over-year and see if there were geographical differences… just as there was a large increase in homicides in 2020.

The Geography of Homicide -- States, Base Rates, Increases, and Correlations

Motor Vehicle Accident Deaths

This one is not as large in its movement as the ones above.

December 2025, National Highway Traffic Safety Administration, Crash Stats: Early Estimate of Motor Vehicle Traffic Fatalities for The First 9 Months (January–September) of 2025

It’s only comparing year-over-year, so let me compared the worst year (2021) against 2025: at 14% decrease.

That is not as impressive as the large decreases in drug ODs and homicides — but note, the motor vehicle traffic deaths did not increase as much during the pandemic as homicides and ODs did.

The number of deaths is about what we saw in 2018.

Consider, too, that generally, for motor vehicle accident deaths, we usually measure the rate based on the number of vehicle miles traveled… but that’s for another time.

Here is the change in count by region:

The report is only 6 pages long, and includes a table with a state-by-state breakout, so you may wish to look at it for yourself.

Getting back to normal?

At least with respect to the “nasties” of the external causes of death, it may be that some of problematic causes of death may be coming down to less crazy rates that we saw during the pandemic… for a variety of reasons.

People need to have a little humility, though (especially politicians… not that they’re known for being humble), in claiming credit for any of these improvements.

In some of the cases of all of these causes of death, there was a large jump up in deaths during the pandemic, and in the case of drug overdoses, it was a problem before the pandemic. There is still much to be done.

Claiming victory while the death counts are still fairly high is not helpful. And claiming cause-effect chains that do not hold is also not helpful.