Top
Best
New

Posted by marojejian 1/10/2026

Overdose deaths are falling in America because of a 'supply shock': study(www.economist.com)
218 points | 241 commentspage 2
indrora 1/11/2026|
Discussion by an author on the paper: https://unrollnow.com/status/2009340857909170395

Original: https://x.com/KeithNHumphreys/status/2009340857909170395

idoubtit 1/11/2026||
The reporter rightly queried other researchers about this article, and all of them were skeptical that a "supply shock" could be the cause, or even the main cause. My own skepticism is because the death rate went down many months before any sign of shortage appeared.

I haven't read the paywalled Science paper, but The Economist extracted a graph which shows that the purity of Fentanyl pills was stable till the first months of 2024, then dropped sharply. The purity of the powder peaked in 2023, then went down in 2024, back to its older levels. They suppose that it proves the supply was short, but another researcher even states that the supply of Fentanyl precursors didn't change until the end of 2024.

Anyway, the epidemic plateaued by the start of 2022, then went down after August 2023; Source https://www.cdc.gov/nchs/nvss/vsrr/drug-overdose-data.htm

Why did the death rate slow down for one year, then go down many months before any sign of supply changes?

jfengel 1/11/2026|
The article says that deaths peaked in mid 2023. Narcan was approved for over the counter use in March 2023.

That suggests a plausible alternative cause.

SV_BubbleTime 1/11/2026||
What is your supposition here? That addicts are keeping narcan around just in case? That good friends of addicts are standing by with the spray in case it is needed? That your local opium den had staff with it on hand?

Narcan should be available, but short of a few users that know they need to keep it around, I don’t buy that making it available has meant a significant change in total outcomes because of timely deployment.

squigz 1/11/2026|||
Some of those ... absolutely, yes?

You might have got some at a rehab centre, or someone might live with a non-addict friend or partner. Community outreach workers (in cities that have embraced this stuff) might carry some around to administer.

I would be surprised if widespread availability to Narcan didn't decrease ODs.

cindyllm 1/11/2026||
[dead]
swamp_donkey 1/11/2026||||
First responders would carry narcan or equivalent. I am sure it is readily available in areas where people are dying daily from overdoses.
SV_BubbleTime 1/12/2026||
Why would OTC even remotely apply to first responders?

I’m an EMT-B. I’ve had narcan in my personal kits for years.

fc417fc802 1/11/2026||||
Yes to all of the above. I knew of addicts who managed to get their hands on it many years ago when it required a prescription. Most weren't that resourceful though.
patmorgan23 1/11/2026|||
See as you can buy narcan out of a vending machine now, yes it's wider distribution probably has a downward effect on opioid deaths.
phtrivier 1/11/2026||
How good is overdose _deaths_ as an indicator of the epidemic of drug consumption ?

My point being : killing your customer en masse is bad business practice in the long run. (Or even in the medium run.)

So, the drug dealer's best interest is to reduce the potency of the drug, therefore limiting the overdoses but keeping the customers alive, and willing to get the next dose.

If it happens when the prices are high, and you're able to cut your product and see it with a higher margin, it's even more value for the sharehol... Sorry, wrong analogy.

Anyway, is the number of people _using_ fentanyl also going down ? Where are the quarterly sales number published ? What's the trend ? When is the IPO ?

asah 1/11/2026|
Drug trade participants don't consider the long run.
bitmasher9 1/11/2026||
Some drug trade participants do consider the long term. Specifically production and bulk transportation benefits from large long term investments.
AngryData 1/11/2026||
I would postulate that people learning how to more safely handle and dose fentanyl would be the biggest reason for ODs to drop.
defrost 1/11/2026|
How would you explain the increase in hospitalisation, Emergency Services deployments, and fentanyl drug purity in Q1 2025 then?

Sudden unlearning of aquired knowledge seems unlikely.

See: Figure 1 graph set page 4 - https://www.science.org/action/downloadSupplement?doi=10.112...

AngryData 1/11/2026|||
Changes in purity, especially when unknown to users, is going to affect hospitalizations and such no doubt, but the people using it also adapt to the purity of a drug over a longer timespan.

Even if people wanted to its not like they can all just bring a sample of their old heroin and a sample of their stronger high fentanyl laced heroin and test their purity and calculate dosages. Which is part of the problem of the war on drugs, many methods of harm mitigation and recovery are barred from users and 90% of their drug information is based on hearsay or personal experience.

defrost 1/11/2026||
> part of the problem of the war on drugs, many methods of harm mitigation and recovery are barred from users

That is a problem for the US, sure. Australia, where I live, has supervised shooting galleries and more of an addiction as health issue approach.

That said, if you had a chance to look at the US graphs linked above - there was a plateau period of high deaths in the US of some three and half years showing no much evidence of users learning to "safely handle and dose fentanyl" followed by a sharp decrease in deaths that corresponds more with a change in policy than an increase in user knowledge.

I would suggest this may be a somewhat more complex and multivariate issue than your initial upthread postulate acknowledges.

1vuio0pswjnm7 1/12/2026||
Works where archive.is is blocked

   echo " 
   foreground=no
   [ economist ]
   accept=127.0.0.127:80
   client=yes
   options=NO_TICKET
   options=NO_RENEGOTIATION
   renegotiation=no
   sslVersion=TLSv1.3
   sni=www.economist.com
   connect=172.64.145.237:443
   "|stunnel -fd 0 -- 50b76b93

   x=https://www.economist.com/united-states/2026/01/08/why-overdose-deaths-are-falling-in-america
   (
   echo GET /${x#*//*/} HTTP/1.0 @
   echo host: www.economist.com @
   echo user-agent: "Mozilla/5.0 (Linux; Android 14) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/127.0.6533.103 Mobile Safari/537.36 Lamarr" @
   echo @
   )|tr @ '\r'|nc -vvn 127.127 80 > 1.htm
   kill $(busybox pgrep -f 50b76b93)

   firefox ./1.htm
ventegus 1/15/2026|
How's this supposed to work? It doesn’t even make it through the Cloudflare captcha.
topranks 1/11/2026||
I wonder if the eradication of the heroin supply from Afghanistan following the Taliban takeover has had an effect?

Possibly re-directed some of the fentanyl to other markets where addicts could no longer get heroin? Thus reducing supply elsewhere?

ciropantera 1/11/2026|
At least in Europe the heroin gap was mostly filled with crack from Latin America.
mrkramer 1/11/2026||
From what I've seen homeless people overdose the most with fentanyl and homelessness level increased during COVID so they were the ones overdosing. Also everybody knows by now that one pill can kill you so that's good enough deterrent.

The opioid epidemic was caused by COVID pandemic and its devastating economic effects and also by cheapness of the fentanyl pills which were going as low as $1 a pop on the streets.

jimnotgym 1/11/2026||
This suggests to me that the government could reduce this even further by simply outcompeting with illegal sources.

One problem mentioned was that other drugs were being laced with fentanyl. Simply supply a licensed, guaranteed clean version through a legal source at a lower price?

Then people who want actual fentanyl, supply that in the same way too.

whimsicalism 1/11/2026|
And how do you know this wouldn’t result in significantly expanding the number/frequency of drug consumers?
jimnotgym 1/11/2026||
Coffee is legal in my country, but I don't drink it. Alcohol is legal, but I drink it infrequently and in moderation. I can get codeine over the counter, but I don't take it every day.
whimsicalism 1/11/2026||
i don’t find this a compelling argument and i think it undermines the case for the policy you are proposing.
znnajdla 1/11/2026||
That example of "shrinkflation" sounds like plain old fraud to me. Having a dimple at the bottom of a peanut butter container so it looks like it has more than it does should be illegal fraud, plain and simple.
whimsicalism 1/11/2026|
the containers say how much volume they have and usually are sold with a unit price as well. seems like a much simpler and general solution than defining some legal shape of jars
yatopifo 1/11/2026|||
The real issue is that those dimples prevent you from getting the last bit. I also find it very annoying that in Canada quantity is often reported in ounces. Aside from the troy ounce, i have absolutely no idea how much an ounce is and whether it measures volume or mass. The only reason we still have ounces is because of trade with the US. Since no Canadian should be buying US made stuff, we should just ban most non metric units at this point.
wasabi991011 1/12/2026|||
I don't remember the last time I've seen an item in a Canadian grocery store that doesn't also include a metric amount (possibly in parenthesis) on the label itself. Not to mention the shelf price has per unit, almost always per metric unit (except rarely meat being per lb).

Are you sure about what you are seeing, is it possible this is just for a few US imports and maybe you aren't looking at the shelf sticker? Or maybe it's a province-specific thing?

Edit: Found the regulation. In general,

> On consumer prepackaged foods, the net quantity must be declared on the principal display panel in metric units [221, 232, SFCR]. However, consumer prepackaged foods that are packaged from bulk at retail, other than individually measured foods, can declare the net quantity on the principal display panel in Canadian units [241.4(2)(b), SFCR].

https://inspection.canada.ca/en/food-labels/labelling/indust...

aitchnyu 1/12/2026||||
I saw a reddit post about somebody cutting up an empty (cant extract any more) squeeze tube of some beauty product and around half was stuck on the walls. I gained a new respect for those who silently chose transparent containers and dispensers.
whimsicalism 1/11/2026|||
moral Canadians should probably also refrain from participating in any American-run website like HN
znnajdla 1/12/2026||||
We don't need to define a legal shape of a container. We can just make deceptive shapes illegal and let a judge/jury define what deceptive means on a case-by-case basis. In fact I don't think any new laws are needed for this, there must be some existing legislation for which a case can made for fraud here.
nathan_compton 1/11/2026||||
I don't know - its very easy to buy what looks like the same jar and find it has less. The consumer should not be expected to be some fucking food detective, constantly working to make sure that they aren't being ripped off. Packaging should be simple enough that the volume presented to the purchaser is the actual volume.
whimsicalism 1/11/2026||
unit costs should be larger on the label. as large as the standard price
nathan_compton 1/12/2026||
Literally no reason to allow companies to use misleading package sizes or to expect customers to check the unit price every time they walk into the grocery store. Its great the unit prices are there, but misleading packaging still sucks.
aitchnyu 1/12/2026|||
Customers should also be informed of quality/expectations differences between versions. For example, there was a car which got a facelift, but they cheaped out on everything inside so it can compete with cheaper cars but unsuspecting buyers had the wrong reputation in mind.
silexia 1/12/2026|
https://www.cbp.gov/border-security/frontline-against-fentan...
More comments...