I got a preview of Drug Wars by
Robin Feldman and Evan Frondorf. It’s not about the War on Drugs, but about the devices used by Big Pharma to maintain the profits they earn from their intellectual property (ownership of drug patents, brand names and so on) and to stave off competition from generics. Feldman and Frondorf propose a number of reforms to the operation of the patenting system to enhance the role of generics. I’m more interested in a fundamental shift away from using intellectual property (patents and brand names) to finance pharmaceutical research.
Drug Wars covers devices like product hopping and evergreening. The idea is to make small changes to existing drugs, so that generics aren’t exact substitutes. Ideally, this enables a whole new patent (evergreening). Even if not, the change means that generic copies of the old version can’t be substituted automatically when prescriptions are filled. All of this is backed up by massive advertising directed both at consumers (this is only legal in the US and NZ) and physicians.
The cost is mainly, though not exclusively, borne by the US government and US consumers. Other developed countries like Australia bargain with the drug companies to place their product on a list that can be purchased at low cost. If the price demanded by the company is too high, there’s no subsidy and sales are usually very limited. One of the central US aims in both the US Australia Free Trade Agreement and the Trans-Pacific Partnership (killed off by Trump) was to cripple the Australian Pharmaceutical Benefits Scheme and similar measures in other signatory countries. Less developed countries have also resisted the demands of Big Pharma to charge whatever the market will bear, with a fair degree of success.
Could this be done better. The ideal alternative would be an extension of public funding to cover drug development as well as fundamental research (which currently acts as an effective subsidy to Big Pharma). Some of this could be done on the current model of grants for research, but there could also be “prizes”, that is, payments for the successful development of drugs assessed as beneficial. Research results would be shared publicly, and products would not be patented (or, perhaps, patents would be held by the government that funded the work)
Again, ideally, all the developed countries would contribute, and share the results. But the US is large enough (around half of global sales of pharmaceuticals and half of expenditure by pharmaceutical companies) and does so badly under the existing setup, that it could probably act unilaterally.
Here are some relevant numbers, I’ve found around the place. All provisional, corrections welcomed
* Total research expenditure by pharmaceutical companies $76 billion. US share, about $40 billion
* Marketing expenditure by pharmaceutical companies more than $100 billion.
* US Spending on prescription drugs $425 billion
* National Institutes of Health total funding $33 billion
Looking at these numbers, I’d say that if the US government doubled the NIH budget and allocated all the money to drug development it would generate more new and useful drugs than are currently produced by Big Pharma in the US, given that a fair chunk of the $40 billion being spent is devoted either to competitive duplication in the race for new products or to non-productive activities like evergreening. Given that much of the cost of prescription drugs is ultimately borne by the government, even a 20 per cent reduction in total costs would be enough to offset the cost the government, with a big benefit left over for consumers.
{ 59 comments }
Val 05.25.17 at 5:03 am
Sounds good to me.
It would also get rid of the problems of research subsidised by drug companies. You are probably familiar with the work of Lisa Bero. I heard her give a presentation in 2015 that suggested to me there is no way of dealing with these problems.
People used to think ‘declarations of interest’ would resolve then, but I heard a paper from Bero at a conference that suggested the opposite (as I recall, I haven’t been able to find a copy) – ie articles where researchers declared their industry affiliations were no less biased (they may have been more so) than those where they didn’t.
I guess the possible reason is that by declaring their interest (or intending to declare it) they feel they absolve themselves of possible bias.
Anyway it suggests it’s an incurable problem and the only way to get decent research is not to have any drug company funding. I don’t think Bero is actually saying that,. but that’s what it seems to me.
Btw as I may have mentioned before, people like Rob Moodie are also saying there should be no involvement of industries (such as food or alcohol) in policy development, which is a similar issue (and should also apply to research). I actually caused a bit of a stir once by tweeting about the fact that Mars (yep Mars!) was supporting some research done by Baker-IDI heart and diabetes research centre. I also got an interesting response recently when I questioned the involvement of pharma in funding some research on women’s health.
People say eg they are aware of the possible problem of bias but have measures to address it/don’t receive direct funding (ie the funding from companies is allocated at arms’ length)/etc but I don’t think that works.
heckblazer 05.25.17 at 5:20 am
I would note that president Trump’s proposed budget would cut NIH funding to $26 billion. Congress gets the final say and will likely ignore that, but large increase isn’t currently in the air.
Pavel A 05.25.17 at 5:58 am
Not to mention the incredibly low quality of research used by most pharmaceuticals to try to convince people that their products are actually significantly better than generics/the competition/the previous iteration of the same product. See John Ioannidis discussing why most clinical research is basically co-opted corporate trash: https://www.youtube.com/watch?v=N63skNtYaJw
Gareth Wilson 05.25.17 at 6:17 am
You’re essentially proposing to merge all drug companies into one. Have you researched the impact of actual drug company mergers on drug development?
John Quiggin 05.25.17 at 7:30 am
@4 “Essentially” seems to be doing an awful lot of work here. Would you care to spell it out?
reason 05.25.17 at 7:32 am
I have floated an idea of how this could work privately (I don’t like the prize idea – it seems to have a winner-take-all problem to me – science ultimately is like navigating a maze, lots of wrong turns finally finding the way out – why should just the guy making the last turn get the whole prize). The way that this could work is that patents are not allowed to be held by manufacturers and there must be several independent manufacturers. The patents would be held by research institutes and they would have use open bidding processes to issue several manufacturing licences. Now the crunch – the currency used in the bidding is donations to research institutes (who would be non-profit and independent).
Gareth Wilson 05.25.17 at 7:42 am
Drug development is a completely different process from academic research. You can certainly have the government pay for it all, but it’ll be the same drug company employees working in the same labs. The only change is that it’ll be a single $66 billion dollar government department, trying to develop useful new consumer products. Do you have any examples of similar organisations? How did their products turn out?
Tim Worstall 05.25.17 at 9:31 am
“Looking at these numbers, I’d say that if the US government doubled the NIH budget and allocated all the money to drug development it would generate more new and useful drugs than are currently produced by Big Pharma in the US, given that a fair chunk of the $40 billion being spent is devoted either to competitive duplication in the race for new products”
You’re going to reduce prices by reducing competition?
Interesting.
If we really want to reduce drug prices why not try reducing the costs of gaining drug approval? There has long been the suggestion that if the FDA only regulated safety rather than efficacy then costs would come down dramatically.
John Quiggin 05.25.17 at 9:41 am
“Drug development is a completely different process from academic research.”
Can you spell this out? Obviously it’s different now, given the way it’s organized, but you seem to be claiming that it’s inherently unamenable to the methods of academic research.
“You’re going to reduce prices by reducing competition? ”
This kind of appeal to the magic of the market doesn’t work any more, Tim. If you think competition between drug companies is pushing prices down, spell out how it does so.
MrArt 05.25.17 at 10:31 am
Rather than nationalising the drug industry, why not just fix these two problems? Allow older generics to be substituted. Ban advertising – clearly that’s possible, as it’s already banned in most Western countries.
MFB 05.25.17 at 11:45 am
It seems that in the West, and with the collusion of Western governments via TRIPS, drug companies are an oligopoly; they routinely collude against the “consumer”, and of course the “consumer” doesn’t have the kind of choice one would have with cheeseburgers, because unfortunately you can die from the diseases that the drugs are supposed to cure or palliate.
Which makes people more desperate, and makes them willing to pay more for drugs, and willing to look the other way when Western governments collude with the process.
My wife is epileptic, and has to take an enormous variety of drugs, and every time I read the tiny print of all the side-effects of the drugs (many of which, as with ARVs, are prescribed to counteract the side-effects of the other drugs) I shiver all over. But what’s to do? Enjoy a grand mal every fortnight?
MFB 05.25.17 at 11:48 am
Oh, and by the way, apropos Worstall’s neoliberal approach: “if the government only regulated safety rather than efficacy”. Doesn’t that open the royal road to manufacturing snake-oil which doesn’t actually do anything, and then using your marketing budget which is much bigger than your research budget to sell it to gullible consumers via lazy or bribed doctors? Isn’t that rather like saying that car manufacturers should be allowed to lie about their emissions or their kilometre-per-litre performance, so long as their wheels don’t actually come off when you go round corners?
MFB 05.25.17 at 11:53 am
And, since I’m here and regarding Gareth Wilson’s point about how the gummint should keep its hands off our drug research — examples of the government doing research which leads to consumer products include the entire electronics industry and the entire aerospace industry. And, as far as I know, a great deal of the drug industry. (Mould-based antibiotics are a government invention, as far as I can recall; didn’t Fleming work for the British government?)
So your point is rather moot, and as Quiggin’s original post points out, the system isn’t working now.
anymouse 05.25.17 at 12:39 pm
Only 425 billion? Wow! The benefits most utterly dwarf that cost. 40 billion in R&D on 425 billion in sales. Wow! At the margins I have nothing against more public spending on pharma R&D. But the current system is a stunning success. Again only 425 billion in cost and 40 billion in R&D. Unbelievable!
John Garrett 05.25.17 at 1:23 pm
Because of the extended patent exclusivity, there is no direct competition among pharmas for prescription drugs. They do of course compete with each other and generics via payoffs for docs and drug advertising, and we all see the results. Big pharma does not produce new drugs, they buy them from startups or license them from universities. The regulatory process is not the problem: there is no relation between price and cost. Pharmas charge what insurance companies will accept. There is no perfect fix, but the first fix is single payer with an obligation to negotiate price and/or buy overseas.
Pavel A 05.25.17 at 1:50 pm
@8:
“If we really want to reduce drug prices why not try reducing the costs of gaining drug approval? There has long been the suggestion that if the FDA only regulated safety rather than efficacy then costs would come down dramatically.”
So, the logical implication of this statement is that drug corporations would be free to sell you snake oil at a reduced cost! Dr. Ioannidis has already basically shown that under the current co-opted research regime, it’s almost impossible to know whether something actually has the effect that drug corporations claim that it has (hint your GP doesn’t really have the time or expertise to confirm or deny their claims, that’s why drug corps sell directly to GPs using pretty standard sales tactics).
Even from a free market perspective, the safety level as well as the rate of success are both crucial ways to evaluate why you as a consumer should be using one drug over another. The placebo effect might indicate that it’s helping you, but you’ll never know for sure without actual impartial studies and there is a serious opportunity cost to using a drug that doesn’t work over one that does. The idea that we should loosen the process further to gain access to even cheaper placebos is laughable.
“You’re going to reduce prices by reducing competition?”
Oh man, the price of the Epi pen really dropped due to all that competition. I guess you’ve never heard of tacit collusion and inelastic demand.
For further consideration, I would direct you to Dr. Ben Goldacre’s talk about why drug companies are super opaque about research and drug trials (hint: it’s because their results are often no better than placebo, sometimes worse than placebo): https://www.youtube.com/watch?v=XqJ7QMTSKME
steven t johnson 05.25.17 at 2:25 pm
As to the merits of the proposed reform, it is indeed likely that putting the government into drug development would be economically beneficial… if you think the purpose of economic arrangements is to satisfy human needs and wants, without distinction of persons. It is not at all clear than any trained economist believes anything of the sort. The general view of the economy is that it is the activity of a civil society which is prior to and superior to the state. This civil society is properly constituted so that individuals may exercise their rights. Those with property will produce for profit. The proposal to set up the state in competition with those who produce drugs for profit is in principle immoral. Whether it meets some utilitarian standard of efficiency is irrelevant.
The only way to justify such an unfair assault by the state is the need to correct for market failure. Unfortunately, there is no principled way to identify “market failure.” It is in many respects a counterfactual. It is a generalized case of opportunity cost. Market failure is a type of opportunity cost. But unless you can estimate the profits lost to which property owners, the idea doesn’t have any real content.
Most of all, in a pragmatic sense, the idea that any political party will forego campaign contributions from pharmaceutical companies in exchange for a benefit to the polity at large rather than themselves seems to misconceive the role of political parties, at least in the US system. Harry Truman, Dwight Eisenhower and aides like Joseph McCarthy killed off any legitimacy to a state pharmaceutical company that brutalizes free enterprise. And nobody since has contested the post-New Deal orthodoxy. The rejection of something worse than the New Deal has been a fundamental principle accepted by all serious people ever since, even (or especially?) by the relative handful who want to keep the New Deal.
A government operated drug testing program could be turned into a government subsidy of pharmaceutical companies, which would be politically attractive. Such a program might still offer enough transparency and objectivity in testing to be a serious improvement.
If one did accept that property rights of individuals (or their incorporated associations) does not mean that monopolies should not be regulated by the state, then patent laws could be amended to require that the patent holders be compensated at limited rates by whoever uses the patents, rather than giving them the right to charge whatever the market will bear, upon pain of withholding the use of the patent. This is also unlikely to be acceptable to economists, but the population at large might not see it as promoting competition in the free market, rather than the unfair competition of the state.
Jerry Vinokurov 05.25.17 at 2:40 pm
No it’s not. The line between “development” and “research” is totally arbitrary; if it happens in a corporate lab we call it the former and when it happens in an academic lab we call it the latter, but they’re not fundamentally different in any way.
In general, patent protections are far too generous and flexible in the US. There should be no possibility for any manufacturer to keep a generic off the market by slightly modifying an existing drug; once something loses patent protections, the manufacturer should have to turn over the entire process. And to prevent situations in which a drug does lose patent protection but isn’t manufactured as a generic because the manufacturer decides to stop doing so, the government should open its own pharmaceutical factories where generics are manufactured and sold at cost. For all the competition fetishists, this ought to give the industry a swift kick in the ass.
Rob Chametzky 05.25.17 at 2:45 pm
Dean Baker of Center for Economic and Policy Research (CEPR) has been writing about this for a while now. Here’s a link to working paper from last August
http://cepr.net/images/stories/reports/rents-inefficiency-patents-2016-08.pdf?v=2
And his even more recent “Rigged” (e)book goes over this too (the ebook is free)
https://www.deanbaker.net/books/rigged.htm
–RC
Sebastian H 05.25.17 at 4:14 pm
Ugh. The numbers you are citing aren’t even close because their methodology is terrible. It would be a major undertaking to get closer numbers but I can at least outline the problem.
They appear to be taking the top pharmaceutical companies opening up their 10-k filing and looking at the top line R&D numbers as if that was a good proxy for industry cost for the drugs that company sells. They are a little bit of a pain to open, I wouldn’t suggest following this link unless you have a huge amount of memory available but for illustration they appear to be using things like page 39 of Amgen’s 10-k.
That kind of look might have been sufficient in looking at a 1980s model of a drug company, where everything is done in house and therefore captured by the operational R&D line. Like it or not (and we can certainly criticize it in all sorts of ways), the way it actually works for much of pharma R&D is that hundreds of tiny companies work on individual drugs in the hope of developing enough promise to get bought out by the big companies. A huge part of the buyout price is effectively the “cost” of R&D that you are trying to capture in your view (i.e. the cost of looking at semi-promising projects and having a huge number of them fail). I don’t know exactly what percentage you should attribute to it (I’d wildly guess something like 60-70%) but your cited methodology doesn’t notice it at all and therefore attributes 0% which is clearly wrong. You’d have to comb through the 10-ks and figure out which mergers, buyouts, and acquisitions are R&D buyouts. In the 10-K I linked you would at the very least need to include what Amgen calls “business combinations” (page F-12). There is a major acquisition of Dezima listed there. That represented a total of about $300 million in cash PLUS $1.25 billion in payments if the drug actually pans out.
I assume that is one of their largest acquisitions for that year, but that one deal alone represents an additional 1/12 of their operational R&D budget for that year that is totally ignored by the blog you cite. (And that is assuming that the drug DOES NOT succeed). [As far as I can tell, post merger the salaries get counted in R&D but the payments still won’t.] There are certainly other acquisitions, and even more for a more acquisition oriented pharma company than Amgen. Similarly patent acquisitions don’t always (and maybe never) get counted as R&D though some largish percentage of that cost goes to paying off R&D costs.
There method of comparing marketing costs is similarly flawed. You can’t use the “selling, general, and administrative” line of a 10-K and assume that is all marketing. That includes the salaries and operating expenses around everyone who isn’t directly in manufacturing or R&D. So you are counting all the accountants, lawyers, secretaries, janitors and regulatory compliance people as if they were “marketing”. Again, I am fully willing to believe that pharma might overspend in all those areas, but you need them to run any big company–even those that never do research at all.
I’m not sure I want to get into their US “premium” numbers. That’s obviously a hard area and I don’t know where to start with it. But we should notice that it pretty much is the upper bound on the premium, not actually a good estimate of what the premium is.
Cranky Observer 05.25.17 at 4:50 pm
Note that many things one might not normally consider marketing are included in a pharma Marketing budget. For example packaging (including the tissue-thin information insert you promptly throw away) is closely scrutinized by the FDA and requires large amounts of time and money to create, maintain, and change; those costs are typically charged to the marketing dept.
Jerry Vinokurov 05.25.17 at 5:07 pm
The fact that you have to do forensic reconstruction to find these numbers is actually part of the problem. Just like the rest of the medical system where there is zero price transparency and no one knows how much anything costs, here you have a maze that it takes teams of accountants to navigate, so that no one can reliably pin down how much a pharmaceutical company spends on anything.
Here’s a bold idea: require that pharma provide this information up front. If you want to claim that your R&D costs are higher because you are acquiring small companies who do the R&D, you should be required to break that out and explain it line by line. Tell us how much of the administrative cost is janitors and how much is salespeople (and how much is just straight-up bribes to doctors to prescribe your medication). If you don’t want to provide that information then don’t complain when people take the top-line items and draw conclusions from that.
Also, the model of having many small companies that all do different kinds of research is actually horribly inefficient. This isn’t an internet startup where you can get going with a few programmers and commodity hardware, it’s a research project that requires doing things like buying ingredients in bulk. All of that stuff scales more efficiently when centralized rather than having small fragmented teams, each replicating the basic infrastructure of a research lab so they can do their work.
Brian Hanley 05.25.17 at 5:19 pm
Since I have a pharma startup, I see the workings of big and small pharma pretty well.
Pharma does very little basic research. Calling what pharma does (including what I do) research is pretty much a misnomer. Pharma does translational development. When pharma does something, it generally knows there is a something out there, or it picks up something out of academia. There is a fair amount of high throughput screening done by pharma, but this is mostly to find alternatives, similars, to current drugs. But even that HTS stuff is done in the context of knowledge of pathways, etc.. Pharma sees a new pathway that is identified and targets things based on that.
Academia does the basic research. That is not to say that what pharma does is like rolling out of bed in the morning. An analogy I’ve used (that has issues, but the all do) is to liken pharma to what SpaceX is doing. Sure, we know we can get to Mars, but it takes some work to do it.
The big problem I see in pharma is the business model straitjacket. I got my education in this from two people. One was a kindly venture capital investment officer who told me, “I like this, I like you. I am sure you can do it. I think it will work. But I can’t fund you because I know I can’t sell your company in Phase III clinical trials to pharma.” He explained that the pharma industry wants one thing – drugs that get repeat business. And they want drugs that repeat daily. Pharma hates the “surgical model” which fixes a problem and makes it go away, or treats it so the patient doesn’t need to come back for a long time.
The other was a professor at UC Davis who is probably the top anti-inflammatory scientist in the world. I was chatting with him about a new paper that found the strongest, longest-acting anti-inflammatory drug ever identified. The professor told me that he had lots of those. Drugs that last 36 hours, 48 hours, etc. Those are no problem. But, he screens them out. Instead, he chooses drugs that last 4 to 8 hours because he knows that anything else is a total waste because no pharma company will pick it up.
This is also a problem in the venture capital system. VC’s want to plan to turn their investment in 5 years. Kauffman Fund has made it clear that this idea of VC firms is malarkey. They don’t do that. But, they all still make their plans that way. This is the best paper on venture capital I know of.
http://www.kauffman.org/~/media/kauffman_org/research%20reports%20and%20covers/2012/05/we_have_met_the_enemy_and_he_is_us.pdf
Because of this accepted standard operating procedure, which is robotic cookie-cutter malarkey that passes for expertise, VCs can’t fund companies that would change the way health care works.
There is the apparent counter-example of Amgen which came out with Enbrel, a biologic for rheumatoid arthritis that lasts for 3 days. But when they were in phase III, nobody would pick it up. That forced the VCs to pony up around $40 million more to build out the channel to get it to doctors. Only when they had done that did Pfizer pick it up. (I have a cousin who has made his career selling Enbrel, first at Amgen, then he went with it to Pfizer, and now he’s back at Amgen. As a result of this problem, Amgen has made more money, a lot more. Also as a result of this, pharma has come up with competitors like Humira so each of the big companies can have their own version. They don’t like to cede the field to each other. But that’s only a me-too competition, which tells you more about the psychology of the rather dull people who run big pharma than it does about what help the world.
One would think that this would make VCs sit back and say, “Hey. Maybe we should change the way we operate, because if things are good in phase III clinical trials, it’s pretty much golden at that point. Risk is low.” But they don’t. As I started to allude to above, and the Kauffman paper pretty much says, VCs are mostly not terribly smart. The people I have interacted with pretty much either apply a cookie-cutter and have connections, or they take a random walk in a target-rich environment and keep operating a confused enough ship that they keep their walk random.
Because of what I have discussed, pharma does not pick up and commercialize a lot of what comes out of NIH funded research. Pharma doesn’t pick it up and VC’s don’t. This leaves a great deal, and I think, some of the most important, best work languishing, untouched.
MPAVictoria 05.25.17 at 6:17 pm
Sebastian H money spent on buying up other companies isn’t “R&D” spending; it is speculation.
Sebastian H 05.25.17 at 6:20 pm
To be clear, I don’t think all mergers count in what you are trying to capture.
What we are trying to trace back is how drug money payments at the end trace back to R&D along the way. A huge proportion of the drug research companies never get payments directly from the drugs they are researching. That is true for two basic reasons: the drug washes out somewhere along the way so nobody ever buys it, or because the initial research company doesn’t have the expertise to make it through all the FDA hurdles. The drug companies your article talks about are all the end-level drug companies. They are essentially the only ones who get payments from medical consumers (either personal level or insurance or government payments). They do a fair amount of research themselves, but a very large percentage gets done further down the chain. Under the current system, if any of those people are to get paid, they get paid by being bought out by one of the top level companies (or get funded by investors hoping to get paid by being bought out). The analysis you link to treats that whole process like it isn’t happening at all.
Maybe it is horribly inefficient and we should do something about it. Fine. But it represents real costs that are actually being spent now, that don’t make it into your numbers.
But I’m skeptical that the government can do much better on the huge scale we are talking about. If it were so easy, why doesn’t one of the non-US governments just do it? Why doesn’t Australia do it? Or the UK government? Or Germany? They could get the US patents and make all of the money that the pharma companies make. Why don’t they just do it?
Matt 05.25.17 at 6:22 pm
In the late 1990s the pharmaceutical industry had great profit margins. It has since undergone massive consolidation and shed tens of thousands of high wage jobs, but its profit margins are no better (perhaps worse?) now than 20 years ago. Most of the low hanging fruit seems to be plucked. “Biologics” (genetically engineered proteins produced in living cells, as opposed to traditional synthesized small molecule drugs like lidocaine, atorvastatin, abacavir…) still seem to be making strong progress but they have high R&D and very high production costs compared to small molecule drugs.
I think we’re looking at an industry that’s compensating for weakness in its old age with increased marketing and mergers. The consolidations, layoffs, and stock buybacks to boost price are a symptom of the underlying slowdown in technical progress; and in particular, technical progress that can be monetized. Maybe there is still some really low hanging fruit to be plucked in terms of diseases that affect mostly tropical regions with dirty water. We won’t find out from the for-profit pharmaceutical industry because there’s no money in treating the afflictions of those “consumers.” Even when a company comes up with e.g. a better antiretroviral drug that Americans will use it’s not going to be as lucrative as the first effective antiretrovirals.
An industry that is suffering from slowing technical progress and increased gimmickry to keep its profits and share prices up seems like a great candidate for increasingly direct government provision. I don’t want the government trying to out-apple Apple in designing portable electronics; that’s an inessential industry that’s still changing at high speed. But I’m perfectly happy living in a public utility district, getting my potable water, sewer, and electrical service from the government. These services are important, slowly changing, and I just want a reliable supply at a price that just covers long run costs. I don’t feel like I’m missing out on any great dynamism by getting sewer services from my city instead of competing private businesses.
I don’t think we would miss out either if a Bonneville Pharmaceutical Administration started (at least) manufacturing off-patent drugs and supplying them near cost, much how the actual Bonneville Power Administration provides low price electricity to my region. Public utility districts demonstrate that government services actually can match private businesses on service and beat them on costs. I see no fundamental reason that a government manufacturer of generic drugs couldn’t perform similarly well. It might not lower costs much in the average case but it would be good insurance against price shocks engineered by the Martin Shkrelis of the world. (As enticement for the right wing punishment brigade, I note that a government drug manufacturer could easily make the thiopental and pentobarbital that private companies will no longer sell to executioners.)
Trader Joe 05.25.17 at 6:28 pm
Agree 100% with what Sebastain says @20. The “right” R&D number is far higher than $76B…there’s bound to be a better figure. A portion of these costs is also capitalized every year as well, so you have to look to cash costs. Because of taxation and the high margins, companies have every reason to undercount the amount they call R&D (some of which goes on the balance sheet and is amortized) and plenty of reasons to expense and not so label.
Two other notes: 1) Remember that whatever the R&D budget is, that pays for all of the failures as well as the things that are successful. A government program designed around explicitly paying for failures would be a very capricious thing to fund – no doubt the many academics here appreciate how difficult mainting grants is – imagine applying for drug research funding for a treatment that is struggling to get traction but might work over time.
2) A lot of the R&D spend relates to processes to synthesize different compounds, non-active ingredients that aid in dose absorbtion, delivery (i.e. capsule, spray, liquid) and side effect mitigation….much of this R&D is general insmuch as it can be applied to multiple final products, but doesn’t necessarily relate to any in specific. This also would need funded and is a cost that is recaptured across successful products.
I’m fully in agreement that the current formulary is broken, I’m a little skeptical as to where the right entry point is for government involvement – I rather like capital markets funding all the failures and all the ancillaries (effectively the fixed costs) maybe regulating the returns like a utility is better balance of risk, reward and opportunity.
John Garrett 05.25.17 at 6:43 pm
Further to #20 – for both tax and marketing reasons it is in pharma’s interest to maximize their R&D expenditures on their 10-K and publicity, by including lots of administration and management (the layers in a pharma makes the government seem horizontal). As Sebastian points out, it’s almost impossible to find their real R&D costs in their reports, since they don’t want the truth known (a little like figuring out how your hospital bill got to 15K in three days).
John Quiggin 05.25.17 at 7:31 pm
@20 No need to be rude. I said in the OP “All provisional. Corrections welcome”.
That said, you’re complicating things needlessly with the discussion of acquisitions. The correct version of your claim is “a large share of R&D is undertaken by small companies which have been omitted”. Once they are included, acquisitions represent double counting. I’ll look into this.
The marketing number isn’t really important for the argument. The only thing that matters is that it’s big.
MPAVictoria 05.25.17 at 8:00 pm
“Because of taxation and the high margins, companies have every reason to undercount the amount they call R&D”
Actually the opposite is true.
SamChevre 05.25.17 at 8:12 pm
The most important thing in drug pricing in the US for the average consumer (not payer), in the US, was Wal-mart’s $4.00 generic program. It covers a lot of the common generics, and is massively cheaper than drugstore pricing. (I know at one time a maintenance SSRI I took was $28 at a drugstore and $4 at Wal-mart.)
I have no idea what the best way to fund the development of the next Sovaldi, but I think it’s important to note that the very high prices are overwhelmingly an issue with new drugs. A better system for generics could help, but only a little.
Biotech Guy 05.25.17 at 8:27 pm
Like some others here, I have worked in pharma, biotech, and life science VC.
Some thoughts:
1. Generic entry is one of the great points of inefficiency and opportunity for windfall profits in pharma, and perhaps on earth. If you can extend the patent life of a $2B/year molecule by 6 months, that buys a lot of lawyering — hell, 6 *days* buys a lot of lawyering. Almost without exception the expiry of every dug brings an orgy of zero-sum activity, and there are great opportunities for reforms. I look forward to reading the book, and I hope they deal with the case of biosimilars (generics for large molecule drugs) which have been a fraught area from the standpoint of regulation and science, and where higher regulatory barriers and consequently lower price erosion post biosimilar entry can really warp investment decisions.
2. However, moving from inefficiency in the generic substitution process to “let’s have the government run all pharmaceutical R&D” is a giant non-sequitur. It’s of course possible that one could get efficiency by moving to a single government run pharma company (this would, as others have pointed out, look *nothing like* the average academic lab funded by NIH). It’s also possible we should close down Apple and restore consumer surplus by having iPhones made by a government smart phone company (I’m not entirely joking). A person’s view on the likelihood of this working out well probably tracks with overall ideological commitments/views about economic systems.
3. But we could give it a try! My modest proposal would be, start the government pharma co, give them a $1B/year ($10B?) budget for 10 years, and see what they come up with. Better yet, let’s have someone other than the US do it, because the ROI is likely in my view to be abysmal. I do think small ‘c’ conservatism suggests that until we have an example of a government owned pharma company developing a bunch of drugs, shutting down big pharma to save money (in theory) strikes me as a risky scheme. Some drugs provide incremental/marginal benefits over the standard-of-care, but but some are miraculous. If you delay the next Herceptin by just three to five years, that’s a real human and economic cost. I’d much rather screw up smart phone innovation than drug innovation, if I had to choose.
4. As an aside, I’ve never exactly understood why US payers have not historically taken a harder line against the most blatant/bogus evergreening strategies (Nexium!). Sure, pharma X can put out an extended release version of the old drug a year before patent expiry and try to switch everyone over, but I’ve never understood why US payers don’t just act like NICE in the UK and refuse reimbursement. Formulary tiering, whereby different drugs have different co-pays based on value is a step in this direction, but can be evaded/gamed. At some point, someone needs to tell patients that just because they want something doesn’t mean they can have it for free. One can solve this by refusing to license safe, effective drugs. But that seems like a dumb way to deal with the problem…
Gareth Wilson 05.25.17 at 8:47 pm
“Can you spell this out? Obviously it’s different now, given the way it’s organized, but you seem to be claiming that it’s inherently unamenable to the methods of academic research.”
Drug development is the process of creating a product that’s useful and safe for millions of people, and finding a way to mass-produce it as cheaply as possible. I suppose academics could do that, but they wouldn’t get many papers out of it. I usually ask people to name a drug created entirely by academia at this stage. To save time I’ll admit there are a few, but the universities usually hired drug companies to do some of the work.
As for penicillin, it was indeed developed, mass-produced and distributed by the government. During World War 2, when there was serious discussion of the government producing socks. Before the war, it was produced by Merck.
Sebastian H 05.25.17 at 8:52 pm
Sorry I didn’t mean to be rude. It was more an exasperation at hearing a kind of zombie argument about pharma. Have you read Derek Lowe’s blog. It has a bunch of great articles on the topic. Ill link when I’m not on my phone.
Patrick S. O'Donnell 05.25.17 at 10:03 pm
Those interested in other works directly and indirectly related to the business practices of Big Pharma, medical research in a neoliberal world, and the question of drugs and intellectual property rights, might find my bibliography on “Biological Psychiatry, Sullied Psychology, & Pharmaceutical Reason†useful: https://www.academia.edu/13799578/Biological_Psychiatry_Sullied_Psychology_and_Pharmaceutical_Reason_A_Basic_Bibliography
Jerry Vinokurov 05.25.17 at 11:25 pm
This right here is the key that unlocks the mystery and I wish more people would talk about these aspects of scientific research.
Ebenezer Scrooge 05.25.17 at 11:45 pm
I think that there is a third path. The US used to make its own weapons, in factories called “arsenals.” Or it would buy things developed by private manufacturers. It now contracts the process out, after deciding what it wants. The process has its problems, but more-or-less works.
Why can’t the same kind of thing work for drugs? The government, using universities and its own resources, does the low-cost research. (There are a lot of national labs in the physical sciences: a model.) If something looks promising, it then tenders out the development: feeding the pills to rats and then (with luck) people. This is very “contractable” stuff, I think. You can have defined protocols and end-points, developed with the bidders. If everything works, you have a generic pill which can be manufactured by anybody that meets FDA standards.
There are some people on the thread who know the industry. What’s wrong with this idea?
Sebastian H 05.26.17 at 2:40 am
BTW, I’m not arguing against the idea that a bunch of patent garbage in extending patents through silly fiddling with dosage levels or recombining different drugs together in one pill is patent garbage.
Lots of it is. The patent office should be a lot harder on companies than it is. But a huge portion of this is driven by craziness at the FDA. If you have two drugs that are commonly used together combined into one pill, that shouldn’t take massive re-trials leading to re-patenting. The companies arguably need the repatenting to pay for the massive re-trials, but we typically shouldn’t need those at all.
Which is part of why I’m skeptical about your “just let the government do it” approach. The government is a contributing to the problem now. (Though knowing the way the federal government works it would just exempt itself from the need for re-trials in such cases, which raises the question of why we can’t just do that now.)
Faustusnotes 05.26.17 at 3:19 am
I agree with Sebastian, those numbers seem way off. Also I wonder how much research from outside the NIH gets used by pharmas. Further, I think private companies should be forced to do at least some of their own product development – if we’re going to have the govt fund their research then the govt should get the profits too. And I think there ar elikely some efficiencies and/or necessary reasons why pharma should do its own research for drug approvals – again at least because so long as they’re private entities they should be subject to broadly similar regulatory principles as other companies. If you think that these companies are so critical to society or so inefficient that they need to be nationalized then say so!
I would be interested to see how the us development process compares to other countries. My understanding is that the hpv vaccine was developed by an oz govt organization (CSL). Was that a more efficient process? Did GSK develop their own or buy the Aussie one? How did the process compare? Does the book only consider the us process?
Faustusnotes 05.26.17 at 3:20 am
Also tims suggestion is hilarious. Tim, if ever you have pause to wonder why libertarian views aren’t taken seriously, consider your proposal that drug companies be able to market useless drugs. Brilliant!
Peter T 05.26.17 at 5:18 am
The suggestion that the US government do this attracts a lot of push-back. Quite rightly – the US government is not a paradigm of efficiency in many areas. But the US – deeply patronal, decentralised, institutionally divided, highly politicised, is not the place to experiment with ideas like this (in the same way that letting the Russians take the running with socialism was a bad idea).
As Britain, Australia and NZ have largely dismantled the capacities they had, I suggest the idea be put to France, or perhaps Iran.
Pavel A 05.26.17 at 6:32 am
@33
“I usually ask people to name a drug created entirely by academia at this stage. To save time I’ll admit there are a few, but the universities usually hired drug companies to do some of the work.”
The primary research is often academic, scaling production and marketing the product are the corporate responsibility. But just to list a few “small” academic discoveries:
– Penicillin (you mentioned this one, but I may as well note again that this one was quite big)
– Insulin (Banting and Best at UofT)
– AZT (NIH)
– Hepatitis vaccine (NIH)
– HPV vaccine (NIH)
– Lithium (NIH)
– Aspirin (Discovered in the mid-19th century by a chemist)
– Nitrous Oxide (Discovered in the 18th century)
– Codeine (Mid 19th century)
– Epinepherine and autoinjector delivery systems in general (19th century Polish chemist and US military, respectively)
– Ephedrine (traditional Chinese medicine formalized in the 1920’s)
Then there is this tidbit: “In 2000, the Joint Economic Committee of Congress reported NIH research, which was funded at $16 billion a year in 2000, that some econometric studies had given a rate of return of 25 to 40 percent per year by reducing the economic cost of illness in the US. It found that of the 21 drugs with the highest therapeutic impact on society introduced between 1965 and 1992, public funding was “instrumental” for 15.”
“Drug development is the process of creating a product that’s useful and safe for millions of people, and finding a way to mass-produce it as cheaply as possible.”
I would amend this to say that corporate drug development is the process by which you also attempt to profit from your product regardless of its actual effects, cancelling out the points about useful and safe (see Paxil, Benzos, Thalidomide). This can usually be achieved by performing low-quality trials or specifically refusing to publish raw data or even finished studies showing that your drug has serious negative side-effects or perform no better than placebo. This is fairly well detailed in some rather large cases against GSK (specifically with regard to Paxil).
Just one small example: a review of published and unpublished trials of 12 antidepressants placed on the market by a variety of corporations (Pfizer, etc) showed that corporations chose to publish roughly 38 trials that showed their drugs as having a positive effect or working better than placebo and “failed” to publish 36 studies showing that their drugs performed no better than or even worse than placebos or had serious side effects (http://www.nejm.org/doi/full/10.1056/NEJMsa065779).
Faustusnotes 05.26.17 at 6:42 am
It’s worth noting that many drugs have their development process limited not by the FDA but by the disease they’re used on. For example, if you are testing AZT (the first ant-AIDS drug) you ideally want to see that it stops progression to AIDS, but this takes on average 9-10 years from acute infection. So you need to either recruit enough people in your study that you are guaranteed that a sufficient proportion of them are near enough to this clinical endpoint that you can observe a statistical difference in one year, or you need an interim clinical marker that is closely related to the endpoint. I think at the time AZT was developed they understoood the role of cd4 cells in HIV progression so they could use those as an imperfect proxy (not sure), but if they hadn’t that knowledge they would need to recruit a very large sample. Basic application of information on rates tells us they would need approximately 10 times the minimum sample size for statistical power in order to apply a survival analysis to the clinical endpoint over a single year. That’s a lot of people. Fortunately for people with HIV, AZT had already been tested for safety in patients with cancer so I assume it had a slightly expedited approval process.
If you’re dealin with cancer that is slow growing and lacks clear clinical proxies then you’re going to face a long, slow development process because the trials will be slow. This means that drug companies need to have a lot of drugs under development at once to ensure a regular flow of new products – no doubt their research budgets cover development of multiple years’ products. Drug companies also aren’t alone in having a large stable of products of which very few will be successful – the publishing and music industries also face this problem, with the subtle difference that their products don’t financially cripple you or kill you (although I would argue the broadway musical can kill rational people). It’s worth noting that despite the vagaries and complexity of this system we have eliminated smallpox using it, made HIV into a chronic non fatal disease, and are on the cusp of eliminating a bunch of other killers. Many of the diseases we could eliminate – rubella, polio, river blindness, for example – are still around due to political and social problems not medical or drug approval problems. While there are no doubt many ways the process could be further improved, even the high costs that Americans pay are a function of their political decision not to have decent health financing – you can bet he drug companies would suck it up and continue to do business if they were forced to negotiate lower prices by a strong federal government. So long as we work within a capitalist system I don’t think the drug approval and patent laws are killing medical innovation or access to good drugs for Americans – that’s a political decision to run things stupidly.
Tim Worstall 05.26.17 at 7:57 am
“This kind of appeal to the magic of the market doesn’t work any more, Tim. If you think competition between drug companies is pushing prices down, spell out how it does so.”
You specifically mention that drug companies will compete with each other to find a new drug. Or a treatment for a condition. “Competitive duplication.” That’s actually something you want to eliminate.
The drug companies of course have patents for the first 10 years of the useful life of the drug (20 year terms, often 10 years to be allowed to sell). There is thus little limit, other than what the market will bear, to the price they can charge for said new drug or treatment.
Unless, of course, some other drug company has another treatment which is also approved. Each still has their own patent but they face competition. That competition limits prices. That Bayer has something to treat “freemarketism” limits the price Genentech can charge for their rival treatment for “freemarketism.”
I don’t regard any of that as particularly controversial either.
reason 05.26.17 at 8:09 am
Doesn’t anybody think that the basic problems are:
1. Science and business really don’t mix very well (as I mentioned before science progresses as much by public failures as by successes – after all that is what it is trying to do – disprove things).
2. Intellectual property fundamentally is creating monopolies and so is anything but “the magic of the market”.
Tim Worstall 05.26.17 at 8:11 am
“Tim, if ever you have pause to wonder why libertarian views aren’t taken seriously, consider your proposal that drug companies be able to market useless drugs. Brilliant!”
I believe that’s the way the entire vitamin industry works, isn’t it?
Gareth Wilson 05.26.17 at 9:39 am
“What’s wrong with this idea?”
If you want to control excess profits and lower costs, US defence contractors wouldn’t be the model I’d choose. And I think you’re bringing the pharma companies in too late – they’re quite likely to look at the compound you want tested and declare it’ll never be a viable drug. Structure’s too wacky, never survive the stomach, whatever.
Layman 05.26.17 at 12:11 pm
Sebastian H: “The patent office should be a lot harder on companies than it is. But a huge portion of this is driven by craziness at the FDA.”
This argument shows up in pretty much any discussion of the wrongdoings of corporations. A big pharma corporation tinkers with a once-high-flying drug, making minor adjustments in a transparent bid to extend the patent life in order to prop up the high retail price for a few more years, and that’s the government’s fault. Adjust and repeat as necessary. The FDA isn’t the bad actor here.
Jerry Vinokurov 05.26.17 at 2:55 pm
But it isn’t, actually. There’s no need to have government run all R&D, just some of it, and actually government labs have a fine record of doing so. Got us to the moon and everything! And in addition to R&D, we could easily have government factories producing cheap generics, like, say, insulin, something that millions of people need to live, and for which they are being mercilessly extorted by pharma companies.
There’s no reason to believe that the average academic lab funded by the NIH is the model of greatest efficiency. Indeed, there are many reasons to believe that it is not.
Now this is definitely a non sequitur. Because iPhones are not the kinds of things that people need to live, while drugs most certainly are. There’s no reason for the government to be involved in the production of basic consumer goods, but every reason for it to be involved in the production of lifesaving medicine.
The first part of your proposal is inconsistent with the second.
But no one is shutting down pharma. A government-owned lab would exist alongside already-existing labs, except that all intellectual property stemming therefrom would be in the public domain and the drugs produced sold at cost. If pharma can’t compete with that model, tough shit.
Too bad no one knows where the next Herceptin will come from. Pharma loves to play up this line of reasoning because it perpetuates the false idea that because method X led to advance Y it will also lead to advance Z. But there’s absolutely no reason to believe that the current model of pharma R&D is the best one. If it truly is, then it should easily win out against a public R&D model and pharma has nothing to fear! If it isn’t, then it deserves its fate.
Because the agencies responsible for this are regulatorily captured or prevented by law from negotiating medicine prices, and insurers often have perverse incentives to actually pay out more in order to, for example, increase their medical loss ratio. See Elizabeth Rosenthal’s An American Sickness for more details.
Jerry Vinokurov 05.26.17 at 3:00 pm
Regulatory capture is a hell of a drug, pardon the pun. Let’s start by purging all the industry remoras attached to it and putting actual public servants in charge. Actually, let’s do that for more than just the FDA.
Well, I guess we’ll just keep going with the current system until health care costs consume 100% of GDP. Or, perhaps we could re-examine the a priori belief that the US government can’t do anything right and start working on figuring out how it actually can. I know which path I want to take.
Faustusnotes 05.27.17 at 12:19 am
Tim, just a little heads up for you: the vitamin industry doesn’t work. It’s a sham. Doctors don’t prescribe or recommend vitamins, because we know they don’t work. If this is your model for cancer cures, you really can’t even be described as phoning it in. This is why libertarian ideas are a joke.
Collin Street 05.27.17 at 4:38 am
There’s no reason for the government to be involved in the production of basic consumer goods
Well… property rights? I mean, I can “buy” an iPhone because the government has groups of heavilly-armed people who’ll kill — literally, shoot-until-dead — anyone who takes the one the government has assigned to me away, and iPhones can be “sold” and thus become worth producing because both the iPhones themselves and the exchange material are protected the same way. The government is intimately involved with the production of Every Fucking Thing Ever.
[it’s like… the air! Air buoyancy reduces the measured mass of everything around us; it’s a fairly minor factor that usually doesn’t matter, but when you’re looking at changing huge elements of stuff some of the otherwise-invisible elements are things you have to start thinking about]
Gareth Wilson 05.27.17 at 11:08 am
“Drug development is the process of creating a product that’s useful and safe for millions of people, and finding a way to mass-produce it as cheaply as possible.â€
“I would amend this to say that corporate drug development is the process by which you also attempt to profit from your product regardless of its actual effects, cancelling out the points about useful and safe.”
You do understand that you can describe what barbers do without having to include Sweeney Todd?
John Garrett 05.27.17 at 1:46 pm
I have more proposals that won’t happen: revise the law to include broad intellectual property rights to the government for specified categories of drug development (life saving, new anti-biotics, orphan diseases) for all research funded by the government and let the government place them automatically in the public domain. That wouldn’t cover everything (e.g. Hepatitis C treatment, the only disease we have cured since….measles) but it would cover lots. Then government fund an early stage biostartup fund on the same principles – $300 million would change the world.
Jerry Vinokurov 05.27.17 at 4:20 pm
Sure, yes, I agree with all of that. I was obviously being a bit imprecise in my statement, with the hope that it would be understood that what I meant was that I don’t think there’s any reason for the government to directly produce iPhones.
Nia Psaka 05.28.17 at 1:33 am
Your ideas are in line with things the Pirate Party was saying over a decade ago.
I agree with the Pirate Party critique: The underlying problem is intellectual property itself. or rather making the ownership of “IP” the priority around which other activities flow. If the objective is access to technology, and IP gets in the way, IP must be treated as counter-productive.
Drug research should be publicly funded, and drug manufacturers should base the viability of their business model on their physical capital, not on restraint of competition.
Collin Street 05.28.17 at 7:14 am
Sure, yes, I agree with all of that. I was obviously being a bit imprecise in my statement, with the hope that it would be understood that what I meant was that I don’t think there’s any reason for the government to directly produce iPhones.
I get this. I, however, think that there are reasons for the government to directly produce iPhones: all profit is derived from rent, and rent is a consequences of the actions of the community — the government — not the nominal “manufacturer”. I don’t think that this is a compelling reason — a broad-based rent tax, 50% of all returns-to-capital over 3%, say, might be preferable — but that there are decent reasons, that a person might reasonably think it a good idea, seems pretty solidly demonstrated to me.
Sebastian H 05.28.17 at 2:55 pm
I’m sorry to use this method to communicate, but there doesn’t appear to be another one.
I have no idea how my comment with various links to Derek Lowe falls afoul of the comment policy, but if you don’t want to get into that I’m fine with that.
I donated quite a bit of research time selecting the links through dozens of his posts which I would like to be able to recapture in the future and had expected to be able to informally reference here, which is frustrated by you deciding not post the comment. If I had realized that was a likely outcome I would have saved the comment locally. Would you mind sending it back to me?
Thank you for your time.
Bill Murray 05.28.17 at 6:54 pm
If you delay the next Herceptin by just three to five years, that’s a real human and economic cost.
and what if you out the next Herceptin kills or seriously injures many of the people for whom it is prescribed in that 3-5 year delay? How many people are you willing to sacrifice since without the trials you don’t actually know the human and economic cost
Comments on this entry are closed.