Reshoring generic drug manufacturing to the United States sounds like an obvious fix for chronic drug shortages, and yet it rarely happens. We sat down with Lori Hoffman of PharmaBioSource, a veteran of generic and injectable drug manufacturing, to talk through the real barriers to supply chain resilience: the economics, the buyers, the quality tradeoffs, and whether tariffs or nonprofit models can move the needle.
Of everything in the supply chain, why are generics specifically the hardest thing to reshore, harder than branded drugs or even APIs?
In the generic space, it is universal for all dosage forms that margins become razor-thin beyond six months of a generic market forming. Competition is so fierce that although you can have a groundswell of interest in reshoring critical medicines, buyers and reimbursers will ultimately say, “But I can get it for two dollars, not $27 or $52.” Most products, unless the active ingredient is unusually expensive, land somewhere in the $25 to $55 range based mostly on manufacturing time. When the same product comes in from overseas at two dollars or less, the buyer says, “I can look good at the end of the quarter with a cheaper product.” That is the whole problem in one sentence.
Compared with twenty years ago, how did we get to the point where generics are so hard to make in the United States?
A few things converged. FDA inspectors became able and willing to go see facilities internationally, and certain countries had a real entrepreneurial spirit among their scientists and business leaders. Overseas firms that used to serve as the back-end manufacturers, supplying product to a US partner who was the front end, increasingly decided to become the front end themselves and file their own applications.
That is also part of what kicked off the GDUFA fees, the Generic Drug User Fee Amendments. The FDA was receiving so many ANDAs (abbreviated new drug applications) for the same products that it could not keep up. Review times had increased from about 18 months to over 40 months, in some cases. If you track the filings, you see a massive uptick. The agency essentially said it needed substantially more people to manage the volume, so it needed user fees. The hope was that this would also separate the serious players from the rest, because now you have to pay to even get in the door.
Purchasing is concentrated in a few buying groups that compete almost purely on price. How much of the reshoring problem is really the buy side, not the factory?
My gut answer was one hundred percent, but then I settled on maybe 50 to 75 percent. It really is not the factory’s problem. And here is the irony with generics: they are not what drives the high cost of medicines, but they get squeezed anyway, because the buyer says we need this as cheap as possible so patients are not paying a fortune, while an intermediary is adding its own margin on top of everyone involved. So, it is not just the buy side’s demand for the lowest possible price. It is how the whole thing is structured, with pharmacy benefit managers, buyers, GPOs, and hospital groups all in the chain. There is a genuine defect in the demand for generics to be as cheap as humanly possible, and a flaw in a mechanism that takes money out for work that adds little value to the patient.
Trace how those thin margins actually turn into quality failures and shortages. What is the first thing that gets cut inside a plant under that pressure?
A lot depends on who owns it. When private equity comes in, it is often “let’s squeeze down some of these quality people,” or perhaps every department is told to reduce headcount. The one thing pharma takes seriously, at least the companies on the up and up, is that quality has to be there, because people can die. But I have seen enough situations where a buyer comes in and squeezes everything, quality and regulatory included. It is not universal, and it is hard to say who gets cut first; it depends on the parties involved and how the acquired company had been managed. I have seen cases in the last five years where someone came in, quality suffered, and then the buyer turned around and blamed the seller: “Look at this defective organization you sold us,” when really they had just gutted quality themselves.
The combination of bare-bones pricing and quality issues is exactly how shortages start. People spent years trying to analyze shortages or build algorithms to try to predict them, when the reality is simpler: when you grind the price down to nickels, where they used to be dollars, only a couple of players may survive. Then it takes just one quality hiccup to break the whole supply chain.
Say a company wants to build US generic capacity tomorrow. Beyond the sticker price of the plant, what actually stops them?
Deciding which products should be manufactured, and then actually being able to make them. Critical pharmaceuticals are not just slightly different pills. Even with oral solids, you have capsules, tablets, and granules, and each requires specialized equipment. Then you have injectables, liquids, powders, and suspensions, in prefilled syringes, vials, and all sorts of devices. If you want to cover the top twenty-five or, better yet, the top one hundred products, you would need one enormous facility with highly specialized equipment that might be used for only one critical product, which means you are carrying even more overhead, pushing prices even higher. It is nearly impossible.
The Mark Cuban Cost Plus model makes sense for pricing: get access to the products (from current overseas suppliers), add a small fixed markup, and sell direct to hospitals, which takes out the middlemen I mentioned earlier. But when you try to return manufacturing to the US, you are back to the same problem: how much capacity goes to which products, and how do you do it all? There are just too many different technologies, and the footprint required for each is huge.
Even a US finished-dose plant often still imports its active ingredient from China or India. Does reshoring the drug product plant without reshoring the molecule actually reduce our dependency, or just relocate the last step?
It helps, because you may be able to hold some API stock that you believe is enough to support X number of months or years of drug product manufacturing. It’s all about stability and forecasting how much to keep on hand, though. Pharmaceuticals will degrade over time, so even if you stockpile an active ingredient that has only a two-year shelf life with a hard expiry date, you either have to use it or it goes to waste. It really depends on the situation. If you have an API that is rock-solid for five years or more, you have largely solved your problem, because finished-dose manufacturing becomes the real rate-limiter and the piece you most want in-country. But if the API requires a 15-step process and is unstable, the supplier may not even start manufacturing until you place the order. In this case, it would be more critical for supply continuity to have the API produced locally. So it really depends on the product.
GMP validation and qualification take years, not months. How does that timeline break the business case?
It depends on what you are trying to do. If you are trying to seize an unexpected opportunity, “there is a gap in the market, let’s capitalize,” you obviously cannot set up a plant fast enough. You might be able to shift some manufacturing if you already have experience making the drug product and have change parts in hand, but for entirely new opportunities, you will never be fast enough. If, instead, you are developing a product to launch at generic market formation five to ten years out, then yes, you can plan around it. So, a US manufacturing line for generics only makes sense if it is strategic and specialized, tied to first-to-market or patent-challenge opportunities, not “me-too” generics. Otherwise, you are building a facility that sits idle much of the time, incurring overhead expense on capacity you are not using.
The 100% Section 232 tariffs hit branded drugs but exempted generics, and the administration has now floated a phased plan to tariff generics too, 100% starting in 2028 and 200% in 2029. For generics specifically, do tariffs help reshoring, backfire, or miss the point?
I think they miss the point entirely. There are far better ways to incentivize companies to build and manufacture locally than adding a cost that lands on consumers. If you are going to put a percentage tax on anything, you would do it on the highest-dollar products, the branded ones, because the same percentage yields far more dollars. But ultimately most people understand that tariffs are just an added cost to the consumer, and that is the last thing we need for pharmaceuticals in this country.
On what to do instead of tariffs:
For specific drugs we truly cannot afford to lose entirely, such as certain antibiotics and other critical products where going to zero would be catastrophic, we could have a government lab or pilot plant able to make them. That is how the government is supposed to work: a national security and public health approach for the genuinely critical products.
Models like Civica, where buyers commit to volume with a mission-driven maker, get held up as the template. Do they fix the economics, or only work at the margins?
The thing people miss is that a lot of Civica’s early products were not actually made by Civica. They came through private-label distribution deals with the same manufacturers already supplying one or two other companies. So Civica was often selling the exact same product from the exact same plant as its competitors, and the real difference was simply the price it set on the shelf. That tells you the model is, at least at the start, largely a pricing and margin intervention rather than true reshoring of manufacturing.
Of the policy fixes on the table (guaranteed-volume contracts, Medicare add-on payments, mandatory batch testing of imports), which would actually move the needle?
I do not think mandatory batch testing of imports does it. I have seen it in Europe. It slows things down under the guise of providing a safety net, a way of saying “we don’t fully trust this.” It may more closely align costs and quality assurances to current local manufacturers. But I do not see it driving decisions to invest in new local manufacturing. Honestly, I am not sure any single fix does it. It may come down to accepting, as an industry, that a generic injectable drug cannot cost seventy-eight cents a vial.
From where PharmaBioSource sits, is there a real business case for a US generic plant today, even with a subsidy behind it?
That was really the driver for Civica and the others; they likely would not have taken the first step without government money, and if you go far enough back, most pharmaceuticals trace to government-funded research anyway. A US generic plant will only be able to compete on price if it focuses on first-to-market generics or on a specific class of compounds or dosage forms (high barrier to entry). If you try to be everything, you need many different capabilities, and they will not all be used all the time, so you carry a lot of idle overhead.
Modular facilities that let you swap equipment in and out may help, in theory, but I have seen a few and they still are not ideal. It’s not even possible when you are producing antibiotics that are sensitizers or certain oncology products that have to be either fully separated or made on dedicated equipment to prevent cross-contamination. So a generic plant may need to be very specialized, but to make the economics work at all, you have to be among the first to market, while there is still a little margin left in the pricing.
Five years out, do we reshore meaningful generic capacity, or do antibiotics, where the last US penicillin fermentation plant closed in 2004, remain the cautionary tale?
I think the challenges will be difficult to overcome if we intend to rely on the pharma industry alone to fix the problem. The economics just aren’t compelling. For the medicines deemed most critical to public health and national security, it will be necessary to expand the work of the public/private consortia and other programs initiated in 2021-2023 via the Defense Production Act, the Biomanufacturing Initiative, the US Innovation and Competition Act, among others.
The views expressed here are the interviewee’s own and do not necessarily represent those of PharmaBioSource.