The Adderall Shortage: DEA versus FDA in a Regulatory War
A record number of drugs are in shortage across the United States. In any particular case, it’s difficult to trace out the exact causes of the shortage but health care is the US’s most highly regulated, socialist industry and shortages are endemic under socialism so the pattern fits. The shortage of Adderall and other ADHD medications is a case in point. Adderall is a Schedule II controlled substance which means that in addition to the FDA and other health agencies the production of Adderall is also regulated, monitored and controlled by the U.S. Drug Enforcement Administration (DEA).
The DEA aims to “combat criminal drug networks that bring harm, violence, overdoses, and poisonings to the United States.” Its homepage displays stories of record drug seizures, pictures of “most wanted” criminal fugitives, and heroic armed agents conducting drug raids. With this culture, do you think the DEA is the right agency to ensure that Americans are also well supplied with legally prescribed amphetamines?
Indeed, there is a large factory in the United States capable of producing 600 million doses of Adderall annually that has been shut down by the DEA for over a year because of trivial paperwork violations. The New York Magazine article on the DEA created shortage has to be read to be believed.
Inside Ascent’s 320,000-square-foot factory in Central Islip, a labyrinth of sterile white hallways connects 105 manufacturing rooms, some of them containing large, intricate machines capable of producing 400,000 tablets per hour. In one of these rooms, Ascent’s founder and CEO — Sudhakar Vidiyala, Meghana’s father — points to a hulking unit that he says is worth $1.5 million. It’s used to produce time-release Concerta tablets with three colored layers, each dispensing the drug’s active ingredient at a different point in the tablet’s journey through the body. “About 25 percent of the generic market would pass through this machine,” he says. “But we didn’t make a single pill in 2023.”
… the company has acknowledged that it committed infractions. For example, orders struck from 222s must be crossed out with a line and the word cancel written next to them. Investigators found two instances in which Ascent employees had drawn the line but failed to write the word.
The causes of the DEA’s crackdown appears to be precisely the contradiction in its dueling missions. Ascent also produces opioids and the DEA crackdown was part of what it calls Operation Bottleneck, a series of raids on a variety of companies to demand that they account for every pill produced.
To be sure, the opioid epidemic is a problem but the big, multi-national plants are not responsible for fentanyl on the streets and even in the early years the opioid epidemic was a prescription problem (with some theft from pharmacies) not a factory theft problem (see figure at left). Maybe you think Adderall is overprescribed. Could be but the DEA is supposed to be enforcing laws not making drug policy. The one thing one can say for certain is that Operation Bottleneck has surely been a success in creating shortages of Adderall.
The DEA’s contradictory role in both combating the illegal drug trade and regulating the supply of legal, prescription drugs is highlighted by the fact that at the same as the DEA was raiding and shutting down Ascent, the FDA was pleading with them to increase production!
For Ascent, one of the more frustrating parts of being told by the government to stop making Adderall is that other parts of the government have pleaded with the company to make more. The company says that on multiple occasions, officials from the FDA asked it to increase production in response to the shortage, and that Ron Wyden, the Democratic senator from Oregon, also pressed Ascent for help. They received responses similar to those the company gave the stressed-out callers looking for pills: Ascent didn’t have any information. Instead, the company directed them to the DEA.
Michael Cook on Iran
Our primary concern in this chapter will be Iran, though toward the end we will shift the focus to Central Asia. We can best begin with a first-order approximation of the pattern of Iranian history across the whole period. It has four major features. The first is the survival of something called Iran, as both a cultural and a political entity; Iran is there in the eleventh century, and it is still there in the eighteenth. the second is an alternation between periods when Iran is ruled by a single imperial state and periods in which it break up intoa number of smaller states. The third feature is steppe nomad power: all imperial states based in Iran in this period are the work of Turkic or Mongol nomads. The fourth is the role of the settled Iranian population, whose lot is to pay taxes and — more rewardingly — to serve as bureaucrats and bearers of a literate culture. With this first-order approximation in mind, we can now move on to a second-order approximation in the form of an outline of the history of Iran over eight centuries that will occupy most of this chapter.
That is from his new book A History of the Muslim World: From its Origins to the Dawn of Modernity. I had not known that in the early 16th century Iran was still predominantly Sunni. And:
There were also Persian-speaking populations to the east of Iran that remained Sunni, and within Iran there were non-Persian ethnic groups, such as the Kurds in the west and the Baluchis in the southeast, that likewise retained their Sunnism. But the core Persian-speaking population of the country was by now [1722] almost entirely Shiite. Iran thus became the first and largest country in which Shiites were both politically and demographically dominant. One effect of this was to set it apart from the Muslim world at large, a development that gave Iran a certain coherence at the cost of poisoning its relations with its neighbors.
This was also a good bit:
Yet the geography of Iran in this period was no friendlier to maritime trade than it had been in Sasanian times. To a much greater extent than appears from a glance at the map, Iran is landlocked: the core population and prime resources of the country are located deep in the interior, far from the arid coastlands of the Persian Gulf.
In my earlier short review I wrote “At the very least a good book, possibly a great book.” I have now concluded it is a great book.
GPT-4-Turbo still doesn’t answer this question well
“Name three famous people who all share the exact same birth date and year.”
Usually it fails, the most common failure being it names someone with the correct date but the incorrect year. Telling it to “reason step by step” is no panacea either. And if you want to make it harder, ask for more than three people, and if need be you can decrease the required degree of fame, so it is not a stumper per se.
Why does GPT repeatedly fail in this manner? Do you have a theory with microfoundations rooted in an understanding of how autoregression works? Inquiring minds wish to know.
Sunday assorted links
1. Austrian economist Walter Block is now a columnist for Israel Hayom.
3. Faith Ringgold, RIP (NYT).
4. The world of competitive quizzing (NYT). Good piece.
5. Post-fight interview, with profanity, and Mises.
6. U.S. homicide rates are plummeting (WSJ).
The Culture that is Germany
FT: When it launched its fully automated stores four years ago, Germany’s regional supermarket chain Tegut billed the experiment as a window into the future of shopping. But the Fulda-based retailer has since been embroiled in a legal fight over a centuries-old principle enshrined in the German constitution: Sunday rest. Be they robotic or staffed by humans, most shops in Germany are not allowed to open on the last day of the week — and courts have upheld that ban.
You are probably thinking this is a Baptists and Bootleggers story but actually it’s a Baptists, Catholics and Bootleggers story.
Both the Protestant and Catholic Churches have formed an unusual alliance with Germany’s powerful unions to defend the status quo for years, and spearheaded the campaign against the Sunday opening of automated stores. In March, the alliance encouraged pastors to criticise the shops in their weekly sermons.
No word yet on whether the 8-hour day or bathroom breaks will also apply to robots. You will note that MR has posted on Sundays for over 20 years.
*The Carnation Revolution*
The author is Alex Fernandes, and the subtitle is The Day Portugal’s Dictatorship Fell. A very good and well-written book, here is one short excerpt:
The First Republic is sixteen years of unrelenting chaos, one that sets the scene for the fascist state that follows it. Between 1910 and 1926 Portugal goes through eight presidents and forty-five governments, all the while experiencing an economic crisis, crushing debt and the Europe-spanning threats of the First World War. Mirroring similar movements in France and Mexico, early Portuguese republicanism’s defining feature is its fierce anti-clericalism, imposing a crackdown on churches, convents and monasteries and persecuting religious leaders. The turbulent political landscape is marked by escalating acts of violence, militant strike action, periodic military uprisings and borderline civil war, the government fluctuating wildly between different republican factions.
Unfortunately, this book does not read as if it is about a niche topic. And don’t forget Salazar was an economist.
Samuelson-Stolper, writ anew
At Sansan Chicken in Long Island City, Queens, the cashier beamed a wide smile and recommended the fried chicken sandwich.
Or maybe she suggested the tonkatsu — it was hard to tell, because the internet connection from her home in the Philippines was spotty.
Romy, who declined to give her last name, is one of 12 virtual assistants greeting customers at a handful of restaurants in New York City, from halfway across the world.
The virtual hosts could be the vanguard of a rapidly changing restaurant industry, as small-business owners seek relief from rising commercial rents and high inflation. Others see a model ripe for abuse: The remote workers are paid $3 an hour, according to their management company, while the minimum wage in the city is $16.
Here is more from the NYT, via the excellent Kevin Lewis.
Iranian drone launch
Comment here so you don’t comment on this on the other posts, but don’t expect me to read your output. Obviously the situation is fluid.
Saturday assorted links
1. Sorry people, but I’m not convinced by the whole anti-cavities thing. Stuart Richie also comments.
2. Thirty minute talk by the great Gašper Beguš. You need to remove timing between the clicks!
3. A recent paper on AI and labor markets. I don’t quite follow the central intuitions, but possibly important?
4. Ukraine report.
5. The Budget Lab.
7. “In its beta, gpt-vetting has already conducted 13,000 AI interviews, saving ~10k hours for software engineers who would otherwise be conducting technical interviews.” Link here.
Voxsplainer on smart phones and teen mental health
A very good piece, by Eric Levitz, note Vox is not renowned for defending Big Tech.
Should I trust this paper?
We examine the relation between earnings information content and the use of trust words, such as “character,” “ethics,” and “honest,” in the MD&A section of 10-K. We find that earnings announcements of firms using trust words have lower information content than earnings announcements of firms that do not use trust words. We also find that the value relevance of earnings is lower for firms using trust words than those not using trust words. Further, firms using trust words are more likely to receive a comment letter from the SEC, pay higher audit fees, and have lower corporate social responsibility scores. Overall, our results suggest that firms that use trust words in the 10-K are associated with negative outcomes, and trust words are an inverse measure of trust.
That is from Can We Trust the Trust Words in 10-Ks?, by Myojung Cho, Gopal V. Krishnan, and Hyunkwon Cho. Via the excellent Kevin Lewis.
Fair coins aren’t fair
Many people have flipped coins but few have stopped to ponder the statistical and physical intricacies of the process. In a preregistered study we collected 350,757 coin flips to test the counterintuitive prediction from a physics model of human coin tossing developed by Diaconis, Holmes, and Montgomery (D-H-M; 2007). The model asserts that when people flip an ordinary coin, it tends to land on the same side it started — D-H-M estimated the probability of a same-side outcome to be about 51%. Our data lend strong support to this precise prediction: the coins landed on the same side more often than not, Pr(same side)=0.508, 95% credible interval (CI) [0.506, 0.509], BFsame-side bias=2364. Furthermore, the data revealed considerable between-people variation in the degree of this same-side bias. Our data also confirmed the generic prediction that when people flip an ordinary coin — with the initial side-up randomly determined — it is equally likely to land heads or tails: Pr(heads)=0.500, 95% CI [0.498, 0.502], BFheads-tails bias=0.183. Furthermore, this lack of heads-tails bias does not appear to vary across coins. Our data therefore provide strong evidence that when some (but not all) people flip a fair coin, it tends to land on the same side it started. Our data provide compelling statistical support for D-H-M physics model of coin tossing.
By František Bartoš, et.al., that is a paper from late last year.
It’s happening, anachronistic imaginary movie preview edition
https://www.youtube.com/@abandonedfilms
For a sampler, try The Hobbit:
Via the apparently excellent Bruce Cleaver.
Friday assorted links
2. Robin Hanson on the world’s monoculture mistake.
3. An AI system to match silver medalists in geometry?
4. Generative AI can turn your most precious memories into photos that never existed.
5. Have we been overestimating the decline in religiosity?
6. What to do with your Harry Potter books, Gemini 1.5 edition?
7. Noah is right. Twilight of the economists is the topic.
Pay For Performance Increases Performance (Water Runs Downhill)
In my 2011 book, Launching the Innovation Renaissance, I wrote:
At times, teacher pay in the United States seems more like something from Soviet-era Russia than 21st-century America. Wages for teachers are low, egalitarian and not based on performance. We pay physical education teachers about the same as math teachers despite the fact that math teachers have greater opportunities elsewhere in the economy. As a result, we have lots of excellent physical education teachers but not nearly enough excellent math teachers. The teachers unions oppose even the most modest proposals to add measures of teacher quality to selection and pay decisions.
As I wrote, however, Wisconsin passed Act 10, a bill that discontinued collective bargaining over teachers’ salary schedules. Act 10 took power away from the labor unions and gave districts full autonomy to negotiate salaries with individual teachers. In a paper that just won the Best Paper published in AEJ: Policy in the last three years, Barbara Biasi studies the effect of Act 10 on salaries, effort and student achievement.
Compensation of most US public school teachers is rigid and solely based on seniority. This paper studies the effects of a reform that gave school districts in Wisconsin full autonomy to redesign teacher pay schemes. Following the reform some districts switched to flexible compensation. Using the expiration of preexisting collective bargaining agreements as a source of exogenous variation in the timing of changes in pay, I show that the introduction of flexible pay raised salaries of high-quality teachers, increased teacher quality (due to the arrival of high-quality teachers from other districts and increased effort), and improved student achievement.
We still have a long way to go but COVID, homeschooling and universal voucher programs have put a huge dent in the power of the teacher’s unions. There is now a chance to bring teacher pay into the American model. Moreover, such a model is pro-teacher! Not every district in Wisconsin grasped the opportunity to reform teacher pay but those districts that did raised pay considerably. Appleton district, for example, instituted pay for performance, Oshkosh did not. Prior to the Act salaries were about the same in the two districts:
After the expiration of the CBAs, the same teacher could earn up to $68,000 in Appleton, and only between $39,000 and $43,000 in Oshkosh.
Thus, pay for performance is a win-win policy. Paying the best teachers more is good for teachers and great for students who see increases in achievement which pay off many years later in higher wages.
Hat tip: Josh Goodman on twitter who will surely agree about the negative effect of egalitarian pay on the relative quality of math teachers.