Hacker News Reader

Comment by opportune | Hacker News Reader

If you think about it, the entire US pharma industry paradigm doesn’t really make sense. A lot of exploratory research into drugs is publicly funded, but for some reason we then rely on private companies to pass FDA trials and bring the drug to market, at the cost of profits/subsidies/advertising/executive compensation/rent seeking IP.

Why not publicly fund R&D all the way up to when a drug passes FDA trials, then allow any private company to produce a generic version of the drug? That way you have competition on the price of manufacturing without all the unnecessary costs involved with private IP

View on HN

This is a good question that is difficult to answer.

I think the answer is that drug development is the kind of process which public institutions are particularly ill suited to manage. That isn’t to say it’s impossible, but the chance of messing it up is very high. A public institution simply cannot fund a research process purely based on facts. There are always other political motivations involved, about where the money goes, who gets the money, who is getting too much money etc. A drug company is free to pursue and fund development of drugs based on merit alone.

> I think the answer is that drug development is the kind of process which public institutions are particularly ill suited to manage.

> A drug company is free to pursue and fund development of drugs based on merit alone.

What?

The argument that only they can provide what the public needs is just an excuse to avoid competition, as it is everytime you hear it being made in any industry: Municipal broadband is not able to fulfill the requirements of the fast-pacing telecommunications market so we should outlaw it and subsidize the private sector even more so you can finally have that 1MB dsl line you were due in 1999.

The Pharma industry has been "caught" many times preferring and in some cases preventing development of "cures" to instead sell long lasting treatments. Like most businesses, the only merit they care about is what makes them the most money, and there is nothing wrong with that, but let's not make them the gallant saviors of the masses when their evident greed becomes apparent every quarterly report and lobbying all over the world to avoid any form of competition is just the latest example of it.

>The Pharma industry has been "caught" many times preferring and in some cases preventing development of "cures" to instead sell long lasting treatments.

https://www.cnbc.com/2018/04/11/goldman-asks-is-curing-patie...

I don’t see any argument refuting my point that governments could not replace big pharma. I agree with you that big pharma are like any other large corporation at the end of the day, it’s all about the bottom line.

It is however an indisputable fact that big pharma have delivered novel drugs which do make a large difference to many patients’ lives. The scale of their research operations dwarfs all other biomedical research by orders of magnitude. It is highly speculative that a public institution could do research at this scale. If you have ever worked in an academic department this would be immediately obvious.

> I don’t see any argument refuting my point that governments could not replace big pharma.

I'm not trying to. I'm refuting the argument that they "develop drugs based on merit alone" and thus somehow because of that don't need competition.

Should be if a company gets a public grant for research their research must become public domain. Wishful thinking though, that'd never happen.

Is that the basis on which drug companies currently pursue and fund development of drugs?

To the extent that merit is equal to profit anyway.

Yes, in my experience.

Because the US govt can't file for most patents. The constitution says any IP developed by the US is in the public domain (there are secret patents but they may not survive a skeptical supreme court). The US govt wants American companies to own patents so the profits come to the US.

US developed products and research used to be free to everyone (the US govt funded BSD). The US govt stopped that when Japanese companies started bring new inventions to market that American companies didn't want to because it threatened it's current product line.

With tax shifting strategies and countries negotiating deep discounts while Americans pay much higher prices; it is difficult to justify the current practice now.

But I am suggesting that these developed drugs do become part of the public domain! Let the government develop it and let anyone who meets manufacturing standards manufacture it. It sucks when Japan does that to cars, but it would suck a lot less if Japan did that for lifesaving pharmaceuticals. In that case it’s not an investment in technology that could be used to strengthen the economy directly, but instead an investment in public health (which will of course benefit the economy indirectly)

I love the spirit, but "let the government do it" usually means contacting.

Edit: which is considerably more expensive than it should be.

The gov. could "contract" the way they do with research grants, and could also expand the FDA/NIH to do more in house. Traditional government contracting sucks, I fully agree

Makes me wonder. Can the usps file for patents? Can they have trade secrets? What about trade marks?

Not too sure about an independent agency or a GSE, but the courts have allowed third parties to trademark certain designs and transfer them to the fed govt.

They have definitely enforced trade mark. I remember hearing that they had sued the band "The Postal Service".

"ZIP Code" was originally trademarked, but I believe it has expired now.

The State of New Jersey actually sued a Florida pizza place for trademark violations: https://www.nbcnewyork.com/news/local/New-Jersey-Loses-Trade... they lost, however, because nobody ever confused a Florida pizza place with a New Jersey highway.

That’s bizarre. Lots of countries have postal services. Does USPS think they have a monopoly on the concept of delivering mail?

The settlement was The Postal Service playing a small concert for USPS employees, so the end result was kind of cute :)

Trademarks are inherently geographically limited. They apply in regions where the trademark holder does business under that name, not over the whole world.

The bizarre part is trying to apply their trademark to an area of business that is not related to what USPS does. USPS does not play music gigs. There would be no danger of someone confusing USPS with a band. Trademarks are also limited to the line of business that the trademark holder operates in. This is why Apple Computers and Apple Records (the record label for the Beatles) were able to operate without trademark conflicts for decades until Apple Computers started getting involved with music and then needed to come to an arrangement with Apple Records.

Apple computer and Apple records early on had a trademark spat, and subsequent settlement. The reason why the Beatles took so long to show up on iTunes was related to the fallout of breaking that agreement.

https://en.wikipedia.org/wiki/Apple_Corps_v_Apple_Computer

> Does USPS think they have a monopoly on the concept of delivering mail?

Under the name "The Postal Service" and within US borders? Yes, they do think that, and their trademark would agree with them.

Yes they legally have a monopoly mail service. You cannot compete against them. Package delivery services like ups or FedEx are a different type of service and distinctly defined. But mail carrier is exclusive to USPS.

Why are people down voting my above comment? It had up votes until recently and people are not posting any counters.

It’s factually incorrect. The US government can hold patents. https://www.law.cornell.edu/cfr/text/34/6.3

More widely, if someone argues that changing the law to do X is a good idea. You can’t argue that the law prevents Y because changing the law is the idea.

> It’s factually incorrect. The US government can hold patents. https://www.law.cornell.edu/cfr/text/34/6.3

Fine, that small part about filing is incorrect. But most of my argument is correct.

The US govt allows private companies to monopolize public funded research to keep jobs and tax revenue that research would make in the US, because the US govt can't legally restrict them by filling patents of it's own (except for certain secret patents I read in Wired magazine).

> More widely, if someone argues that changing the law to do X is a good idea. You can’t argue that the law prevents Y because changing the law is the idea.

I was answering OP's question. The US govt wanted a return for it's funding of ALL RESEARCH and this was a way to deal with it ~40 years ago. This is now having issues with drug patents today.

If you actually read my argument more carefully you would see I agree with OP at least with drug patents.

> With tax shifting strategies and countries negotiating deep discounts while Americans pay much higher prices; it is difficult to justify the current practice now.

Please actually comment instead of just downvoting; this isn't reddit.

The amount of innovation and capital at risk in private biotech companies is similar to that for the tech industry.

If we got rid of the tech industry and had the govt do all of it, do you believe we’d be where we are today?

I’d argue we wouldn’t be.

One of the big reasons why small biotech’s outperform big pharmaceutical companies is the lack of bureaucracy and high level of risk taking. Two things the govt is bad at.

As well, one reason why we have so many drugs is there are thousands of companies placing different bets on an idea. If that was all pulled under a centralized committee that made decisions on what to pursue, we’d be missing out on a ton of long-shots that every “smart” person said would never work.

Thank you for your opinion! :)

You make it sound like there is a hard barrier between these two options but like most things in reality it is a continuum. You can have mainstream R&D be publicly funded all the way and not necessarily restrict private companies from competing.

I personally believe that there is a good chance that a centralized publicly funded and nonprofit pharmaceutical organization could outperform most for profits on various important metrics. At least something that should be investigated and not dismissed outright.

Regarding long-shots, I think you fall victim to the survivorship bias here. Even if some good things came out of it, the question is at what cost? If many tries fail you really need some big wins to cover those costs. Thus, it’s really something that you should look into on an empirical level before making strong judgements.

Out of curiosity, do you have any specific proposals thoughts about what this centrally funded / non profit approach would look like, and why it would work?

My contention is that drugs are expensive because developing drugs is expensive. Developing drugs is expensive because biology is hard. Public vs private funding would not change that

Hey, sorry for getting back to you so late but I had a full weekend :)

I am not an expert in the pharmaceutical space so take everything I say with a big grain of salt. I would think it could be possible to have some agency like NSF or FDA be tasked with financing studies to develop medicines and do medical trials, etc. I don’t think it would need to look much different what there is now, maybe higher and more long-term grants for applied R&D. Manufacturing could then be outsourced, Apple style. I guess the main challenge would be to have reporting and grants be distributed effectively and fairly. There should be some level of friendly competition/ cooperation to ensure that progress is being made. But I think, in principle this already works pretty well through peer review and on some level metric tracking for individual researchers (although I would argue we still often track the wrong metrics). So it’s all about designing the right incentives work work to get done, profit is but one of the possible metrics to achieve that.

This of course does not change that the work itself is incredibly hard. The benefit of doing it in a more „structured“ and nonprofit way is that it is possible to create a more specific/ tailored/ guided set of outcomes given the ability to align the incentives with societal goals. Everything you see going on in blockchain is work in that direction. How can we design economic systems where incentives are aligned to create an overall desireable outcome for individuals, the community and society at large?

The notion ghat “the government” is ‘bad’ at spurring innovation in high-risk scenarios is both common and flawed. It was NASA that sent men to the moon and, more pertinently, it was ARPA that shouldered the risks upon which the predecessor of the Internet (and packet-routing networks generally) was built, as well as (to a lesser degree) the microprocessor. I’m not quite sure why this is such a persistent impression, parroted by folks who, if they thought it through properly themselves, would reach the opposite (correct) conclusion.

I think the concern is not that the government can't spur r&d, but publicly funded r&d would be less efficient than private capital

Back in the 1960s no other organization but the government had enough money to fund something as ambitious as the space program. And the government spends a ton of defense research and always has. Sheer volume of capital can get things done, but that doesn't mean it is efficient

The belief is that private markets incentives more efficient research than public funding. So if there is enough private capital to support some r&d goal, it is preferred to fund it privately vs publicly

I'm sure there's research on this but as someone who has worked on r&d at small companies, large companies and with government funded grants, there's little doubt in my mind what groups are more efficient. If you look at r&d productivity at big pharma companies vs small, small companies are almost an order of magnitude better.

And out of 47 new drugs approved last year, 40% were developed by small companies, 30-35% by big companies, and 1 from a public private partnership (the rest were mid sized companies). That's despite small companies getting an order of magnitude less funding than big ones, and half as much funding as NIH

Does that 40% figure refer to taking drugs all the way through clinical trials and FDA approval, or does it also include e.g. drugs that small companies brought through earlier stages of the pipeline before well-capitalized BigCorp acquired/partnered to finish approval? My impression is that drug development costs are heavily weighted toward the later phase clinical trials, and that it’s rare for a startup to complete them without help/acquisition from one of the big incumbents. I could be wrong about this. My lay perspective is that of someone who used to be a (non-pharma) chemist and who enjoys reading Derek Lowe.

But if I’m right, and small companies usually can’t finish fully independently, that suggests a role for government drug development: funding the clinical trials and completing approval. Whether a government agency or GSK provides that funding doesn’t seem to much affect innovation of small companies in earlier stages.

Here is Derek’s blog post and reference to the original paper. They were quite explicit on what counts as “discovering” a new drug. It’s whoever first created the drug and identified it as having potential to treat a disease.[1]

[1]http://blogs.sciencemag.org/pipeline/archives/2010/11/04/whe...

Thanks, I actually hadn't seen that. The data I was referring to is from my own analysis of 2017 new drug approvals on FDA's website

That paper covers 1998 to 2010, roughly coinciding with the end of the booming era of cardiovascular and psychiatric small molecules (prozac, lipitor etc) and ending right before the patent cliff in 2012, which saw the pharma industry's revenue shrink for the first time ever

Since then, R&D has shifted strongly to smaller companies from big ones, as the data reflects.

I'll have to check my data. You are right re late stage studies being most expensive component of drug dev cost in terms of direct costs, but when you account for cost of failure, early stage dev is actually more expensive. Phase 2 failure (drugs that work in the lab and animals not working in people) is the single biggest cost driver. That cost is linked highly to the difficulty of biology and biochemistry (the stuff Derek Lowe often talks about :) and it isn't clear to me how more public funding would address this

Your proposal of having govt fund later stage r&d makes more sense to me that replacing IP altogether but i still think the current situation would be better at getting new drugs approved more cheaply

Reason being is that early stage innovation is tightly linked to what assets pharma companies want. Allocation of venture capital by therapeutic area is very close to allocation of approved drugs by therapeutic area and to big pharma acquisitions of small biotechs by therapeutic area

If government chose which drugs to fund at later stages, that would probably resemble the current NIH grant approval procedures which are fraught with politics and bureaucracy. There are some NIH / govt initiatives now to fund clinical studies (companies I've worked with have applied to two, CIRM and an NIH funded diabetes research group) and the application processes were brutal and took years. And I don't think those initiatives have yielded much in the way of approved drugs

I don’t think that government taking drugs through clinical trials would significantly reduce development costs. I think it could reduce all-in costs for drug buyers if some new drugs were “born generic” in terms of IP restrictions. Everybody’s revenue is somebody else’s cost. The idea is to shrink the cost/revenue for new drugs that government takes over the finish line, whereas private investment toward the same goal needs to recover R&D costs plus make a profit.

I don’t think that lower drug prices is a panacea for health care costs in the US! Drug spending is a small slice of an oversized pie. But that pie is so big that even small improvements at the margin could save patients billions - mostly unnoticeable at the individual level, but worthwhile considered at the macro scale.

I think that having the government take drugs through clinical studies would increase development costs, and the increase in development costs (funded by taxpayer dollars) plus the fact that you have to replace the entire commercial infrastructure (since no private company will take part in this) would be larger than any savings you'd get from eliminating IP. People would end up paying way way more in new taxes than they save in insurance premiums and copays (people spend roughly $30 a month on rx drugs today, govt run drug dev and distribution would probably cost $60-hundreds+ extra a month, unless you're ok getting 10% of the life saving new drugs our society gets today)

The profits in pharma accrue to the top 15 or so companies by revenue, because they have invested in this commercial infrastructure over decades. You could argue that having large scale clin reg and commercial infrastructure is actually more valuable than IP. Especially in the new era of large molecules and cell / gene therapy, where IP doesn't matter nearly as much as in small molecules. Regualtory and manufacturing issues are a bigger barrier to competition in those areas than IP is. Look at biosimilar development as an example -- patents gone, no new competing drugs in most areas. An exec at a top antibody developer told me they know competitors can work around patents, they just need a 6 month head start and then their best in industry antibody dev capabilities will get them the edge they need. Avexis, Kite and Juno were acquired for ~$10B each not because of their products, but because they had the largest scale manufacturing and clinical logistics set up in gene and cell therapy

I differ from most people on this forum as I fundamentally attribute the high price of drugs to the high cost of developing drugs, not to IP granting monopolies. IP does not exist to grant rent-seeking powers, it exists to incentive publicly useful research that otherwise would not get done. I think that if you look critically at the cost of drug development and the economics of the industry, that's the only conclusion you reach

There's a lot of emotional and biased information about drug pricing out there. It is one of the top political issues of the day, one of the few with bipartisan supprt, it has become an emotional punching bag for a lot of people who have been really poorly treated by an opaque and impersonal healthcare system, and i can understand the frustration people feel. But if you look at the facts and take an objective standpoint, and invest the time into understanding the industry, the public dialogue becomes quite concerning

edit: apologies if this comes off as aggressive, i just feel strongly about this :)

I fundamentally attribute the high price of drugs to the high cost of developing drugs, not to IP granting monopolies.

I agree with you more on this than I disagree. Biology is hard. It is getting harder to find new pharmaceutical interventions that treat currently untreatable conditions, or that are enough better than old ones to repay developing them.

I don't think that government involvement can change the shape of the cost curve for new drug development. I think that government involvement may be able to reduce costs of completing drug development, essentially because it doesn't need profit on top of R&D costs. (Not reducing the big-O cost of drug development, but improving constant factors, if you'll permit the mixed metaphor.)

You seem to think that the costs of reproducing private development infrastructure would be prohibitive, and that no private company will let the government pay for use of its infrastructure toward the development of drugs that are born patent-free. You may be right about the costs of infrastructure. As for the latter, a government agency could offer quite a bit of money to become that partner with institutional knowledge and existing infrastructure. (If you think that private drug companies are in a prisoner's dilemma regarding cooperation with a government looking to undermine their long term profitability, you might substitute "defector" for "partner.")

This has to be my last post of the day but I've enjoyed your responses and perspective.

9 of the 21 drugs I referred to as developed by small companies were owned by big companies at time of approval. Some of the drugs were acquired earlier in clinical dev, some were acquired after NDA was already filed

Thanks. I would credit "acquired after NDA was already filed" new drugs to small companies even if they were owned by big companies at time of approval. Either way, if 12 out of 21 small-company drugs were developed all the way to the way to approval without bigger companies, my earlier guess that the majority required big company help to finish was incorrect.

Agreed.

Don’t get me wrong, I think the gov’t should still play a role in funding “big ideas” that the market won’t. Antibiotic development is a good one.

The other thing to keep in mind are all the failures. By pushing R&D to private companies, the tens of billions spent on Alzheimer’s drugs, of which every single one has failed, didn’t come from the pockets of tax payers, rather they came from private investors.

>If we got rid of the tech industry and had the govt do all of it, do you believe we’d be where we are today?

Aren't a lot of core inventions/breakthroughs in most of our current technology (at least in USA) made by Military?

https://www.businessinsider.com/the-us-military-is-responsib...

The centralized model you refer to in the last paragraph is akin to how a lot of big pharma r&d groups work. And they get about 10x less innovation per dollar spent than small companies (innovation measured as drugs approved)

Imagine how much worse that would be if it was govt rather than a big company

which drugs do you fund? I doubt that public funding of drug R&D will be as efficient as a privately funded effort. If it were all publicly funded, the incentives for selecting drug candidates throughout all stages of the pipeline are shifted.

The government already does decide which drugs to fund, just at a different stage of the pipeline.

Also, it honestly probably will be about as efficient, since we subsidize a lot of medicine’s private R&D already (eg orphan drugs), not to mention eliminating the rentiership of private IP and (some of) the waste of pharmaceutical marketing

Public health outcomes. The model would be something like the UK's NICE. It has many fewer perverse incentives than the current model (which focusses on diseases of wealth).

It's not at all clear that it wouldn't be markedly more efficient in terms of public health outcomes.

How do you apply NICE’s framework to in early R&D when you have no idea what the clinical benefit might be?

As I'm sure you could guess, the reality is much much more complicated. Drug dev is really hard, and really expensive. Private money funds an order of magnitude more R&D than the government. That private money takes a huge amount of risk, and most investors lose money (even in the last few years when VCs and public equity investors have made money, those profits will probably be offset by poor returns to pharma companies for their expensive acquisitions of risky assets). The actual rent seekers in the US healthcare system are large hospital systems, although that's not something that is discussed much in the public debate (hospital spend accounts for 30% of US healthcare spend compared to 10% for prescription drugs, and the strategy of hospitals the last decade or so has been to consolidate into local monopolies so they can capture market share and drive higher rates with payers -- pretty much the definition of rent seeking)

Drug company R&D productivity has been declining for years, and probably will be zero or negative in a decade barring breakthroughs. From my limited experience working with NIH and FDA, the government would be way, way worse -- there is very little accountability so simple things are really hard and slow.

NIH spends ~$34B on R&D. Private industry spends over $200B (mostly US pharma companies, with a good number of EU and some Japanese as well). Top 15 biopharma companies spend like $70B per year on R&D in aggregate. Global venture funding is on pace to be $15-20B this year for biopharma

Despite all this money, we got 47 new drugs in 2017 in the US. Most for cancer and rare disease, because those are the only remaining therapeutic areas where you have enough of a shot of making money to attract private capital. If you look at the top 10 causes of death in the US, only one -- cancer -- receives much private funding and new drugs -- very few new drugs for diseases like heart disease, diabetes, mental illness etc. That doesn't even account for the world's most devastating diseases, many of which receive very little funding and innovation, because the economics dont work

To replace for-profit R&D with government R&D in the US, we'd need to probably quadruple or quintuple the NIH budget -- assuming NIH is just as efficient as private industry. If we wanted to allocate spend to the biggest unmet medical needs rather than those that can make a profit, we'd probably need to spend 10x the ~$300B+ the world spends now on drug R&D (that's basically a guess but I don't think it is crazy directionally, considering how little investment goes into cardiovascular disease, psychiatric disease, infectious disease, stroke, pretty much anything other than cancer)

Apologies for the rant and i know this is an unpopular opinion, but I think the public debate on drug pricing and healthcare costs in general is way off and is a huge obstacle to actually making healthcare better and developing new medicines

Regarding risk and most investors losing money, that makes government funding actually more attractive because it pools risk in basically the largest pool possible.

Additionally regarding the private industry paying $X while the NIH pays $Y, private industry would not have $X for funding (minus the money spent on new drug companies) without at least the prospect of consumers paying for those drugs - that is, the funding is still coming from taxpayers' insurance / Medicare / Medicaid / pocketbooks. If we ignore insurance (because that may introduce a lot more factors), this configuration can only be desirable if we determine that, dollar for dollar, government R&D is less effective than private industry's R&D + all other costs regarding marketing/executive comp/profit that wouldn't already be incurred from their drugs' manufacturing.

I'm in favor of letting industry compete with itself in a market as a way to keep things efficient - except this only works if things really resemble a market. The use of exclusive drug IP to extract money resembles a monopoly even more than regional hospital oligopolies, as need/choice of drug is not typically up to the user. If there were no private IP, nobody would develop new drugs, so why not just eliminate IP and pay for development with the money that taxpayers were going to be spending on R&D anyway?

My premise is that drugs are expensive because it costs a lot of money to get a drug developed, not because IP enables companies to collect value they didn't create. IP is a mechanism that incentivizes investment into a field that is publicly valuable. And it's not that great of an incentive because even with IP most drug r&d is not a great investment

But most importantly, nationalizing drug research would probably make healthcare more expensive, not less. It's a red herring in the healthcare cost debate. Honestly my tinfoil hat theory is that hospitals, which have a ton of political clout (AHA lobby spends about as much as Phrma lobby, but hospitals are top employers in most areas so have a ton of soft political power), drive the drug pricing narrative to distract from the bad stuff they get away with. Tinfoil theory but I don't think it's too crazy

That aside, US spends like $320B on rx drugs. Let's say that goes to zero with your plan. Then we increase NIH budget by $150-200B -- assuming government is just as efficient as private industry. That just covers R&D. You still need infrastructure covering prescription processing, drug manufacturing, distribution, quality control, etc. you'd need to set all that up from scratch, because you wouldn't have a private infrastructure. Could that be $100B? Maybe. So you're breaking even

And if government is half as efficient as private industry? If it's an order of magnitude less efficient? Ouch

My reasoning may be off bc it's Friday night but I think it's directionally correct.

But more importantly: why are we focusing on this in the first place? Do we actually think we can pass legislation to quintuple the NIH budget even if we should (And I don't think we should)

Wouldn't our efforts be better spent addressing the largest source of healthcare spend and biggest source of rent seeking behavior?

Your theory seems good at first blush. However, we all know the US is the only major industrialized country without socialized medicine, correct? So doesn’t that mean that other countries should have succeeded in developing new medicines according to your theory? It appears the vast majority of effective drugs are coming out of the USA. This is a difficult problem.

The argument would be much easier to make if the US had single payer healthcare as well. "We the people" already have to pay, it's just that the costs are hidden from us so the majority of the population gets to ignore that they're really paying even though it might not look like it.

we're primarily paying for early stage basic science, which is the desired purpose of public funding. The huge investments in biotech and pharma (>$50M to bring a drug through clinical trials) are privately funded

I've wondered the same thing.

I suspect it was argued that people betting their own money will be better at choosing winners from the pool of results coming out of publicly funded labs. I doubt that this effect is large compared to the rent-seeking it enables, though.

Good point. Could the fda just run the trials? That would be interesting.

Reply on news.ycombinator.com