Ebola as a market failure

[Simon_Jester]

It's interesting; the problem is defining preventable death and explaining how it's the pharmaceutical company's fault.

With pollution (and deaths due to same) we can say definitively where the pollution came from, or at least we can say definitively what it comes from (e.g. automobile exhaust). So we can always say definitively that this pollution was caused by this person or that activity directly and hold them accountable. It's a crime of commission: you created this hazard, we know it was you that did it, so we can come to you to pay the bill.

But now, you're proposing to penalize someone for a crime of omission. You did not somehow (by unspecified means) prevent this tragedy, therefore we require that you pay for it.

We can say that a seemingly incurable disease might be curable if we tried harder, but it's impossible to defend yourself against that accusation because you're trying to prove a negative. Moreover, if a disease's cure doesn't already exist, how can we know that the pharmaceutical company would have found that cure by trying harder? Maybe the cure could not exist until an unrelated breakthrough occurs in a different field.

There's also a jurisdiction issue- how can the US fine an American company for preventable deaths that happen in Liberia? Liberia isn't part of American legal jurisdiction.

[Irbis]

With pollution (and deaths due to same) we can say definitively where the pollution came from, or at least we can say definitively what it comes from (e.g. automobile exhaust). So we can always say definitively that this pollution was caused by this person or that activity directly and hold them accountable. It's a crime of commission: you created this hazard, we know it was you that did it, so we can come to you to pay the bill.

Can we? Sure, if there's a conspicuous factory, we can measure what if outputs, but a few weeks ago, there was big news one (or several) countries might be breaking Montreal Convention and no one knows who that is. What about pollution in the sewage no tests exist to detect it? Do we give up and claim we can do nothing, or actually try and correct that?

But now, you're proposing to penalize someone for a crime of omission. You did not somehow (by unspecified means) prevent this tragedy, therefore we require that you pay for it.

This argument is absurd. We do it in fact, all the time. Say, when patient in hospital dies, do argue it's disease and no one is responsible, or try to establish if that wasn't doctor's error? If newly built house collapses, do we blame gravity or investigate if architects and builders did something wrong? When Ford rolled out Pinto, do reasonable people blamed physics or omission of crucial parts?

In a lot of fields doing something means responsibility for everything that goes wrong, even if it would go wrong if you did nothing.

Moreover, if a disease's cure doesn't already exist, how can we know that the pharmaceutical company would have found that cure by trying harder?

I said possibly preventable. Not 'preventable at any cost'. If company can show reasonable effort in a given field, it's cleared of any blame. We can even give them a lot of benefit of doubt as long as any sound program exists.

There's also a jurisdiction issue- how can the US fine an American company for preventable deaths that happen in Liberia? Liberia isn't part of American legal jurisdiction.

That would be issue that could be better solved by improving international law, but frankly, the framework for doing that, like Hague Service Convention, already exists. Still, for something like the proposal above to work, all you need is the host country where the pharmaceutical companies are located to sign it the law. That's what, 4 or 5 countries to get all the big players?

Yes, expecting say USA to sign it would be a pipe dream (even though to be fair USA did some things barring its healthcare companies from shoddy practices that were legal at the time) but that was more idealistic, not realistic proposal. But then again, carbon tax 40 years ago would be idealistic, too.

[Simon_Jester]

Can we? Sure, if there's a conspicuous factory, we can measure what if outputs, but a few weeks ago, there was big news one (or several) countries might be breaking Montreal Convention and no one knows who that is. What about pollution in the sewage no tests exist to detect it? Do we give up and claim we can do nothing, or actually try and correct that?

Understood.

My basic point is just that with pollution it is, as a matter of principle, knowable what was put into the environment, and why, and by whom. We may fail to detect past pollution but we can create tests to detect it in the future. We may fail to accurately predict the costs of pollution, but we can recalculate accurately in light of future information.

But now, you're proposing to penalize someone for a crime of omission. You did not somehow (by unspecified means) prevent this tragedy, therefore we require that you pay for it.

This argument is absurd. We do it in fact, all the time. Say, when patient in hospital dies, do argue it's disease and no one is responsible, or try to establish if that wasn't doctor's error? If newly built house collapses, do we blame gravity or investigate if architects and builders did something wrong? When Ford rolled out Pinto, do reasonable people blamed physics or omission of crucial parts?

In a lot of fields doing something means responsibility for everything that goes wrong, even if it would go wrong if you did nothing.

Then you actually have to look at the criteria for criminal liability.

Basically, we have to establish proof of negligence- that professionals could reasonably have been expected to do X, and did not do X. The problem is that if someone dies of a disease and your argument is "it's all your fault for not inventing the cure ten years ago," then it's hard to establish that it was in the pharmaceutical company's power to actually do what you said.

We cannot reasonably say that every person who dies of liver cancer was 'murdered' by pharmaceutical companies who 'failed' to invent the pill that cures liver cancer. Even if there is such a pill, we have no way of knowing when it will be invented, what will have to be developed first, or whether we can meaningfully accelerate the timetable of developing it. Nor do we know how to use our limited resources to strike the right balance between inventing new treatments for liver cancer, heart failure, and Ebola. We don't know what will pay off in advance.

So basically, if you start trying to fine companies for doing their R&D wrong, you make it extremely unpalatable for anyone to engage in R&D at all. By definition, research and development are unpredictable journeys into the unknown. And you cannot reasonably claim "you could have invented this faster and if you had people wouldn't have died" when talking about an unpredictable development process.

Moreover, if a disease's cure doesn't already exist, how can we know that the pharmaceutical company would have found that cure by trying harder?

I said possibly preventable. Not 'preventable at any cost'. If company can show reasonable effort in a given field, it's cleared of any blame. We can even give them a lot of benefit of doubt as long as any sound program exists.

How do we define 'reasonable?' Is there a fixed amount of money that must be spent to be 'reasonable?' Who sets that value? At what point would it make more sense to just nationalize the pharmaceutical companies, if we're going to dictate to them the details of what research they are and are not allowed to fund?

There's also a jurisdiction issue- how can the US fine an American company for preventable deaths that happen in Liberia? Liberia isn't part of American legal jurisdiction.

That would be issue that could be better solved by improving international law, but frankly, the framework for doing that, like Hague Service Convention, already exists. Still, for something like the proposal above to work, all you need is the host country where the pharmaceutical companies are located to sign it the law. That's what, 4 or 5 countries to get all the big players?

In that case you need, essentially, a code of international law to prosecute negligent homicide, which is a pretty drastic expansion and change of existing customary policy.

Yes, expecting say USA to sign it would be a pipe dream (even though to be fair USA did some things barring its healthcare companies from shoddy practices that were legal at the time) but that was more idealistic, not realistic proposal. But then again, carbon tax 40 years ago would be idealistic, too.

Well, my point is that it is questionable whether it is even legal to do this in a system designed around nation-states. My nation can tax carbon burning all it wants; that is a reasonable extension of its power to regulate its own environment and commerce. But to allow my government to sue me for allegedly failing to do enough research to prevent deaths from a plague on another continent is very intrusive. And it sets a precedent for a lot of harassment and unjustified lawsuits for negligence in cases that are not negligence under today's law.

[PainRack]

It may not have been common, but we've all been sitting around for years saying how scary it is. An outbreak was inevitable.

Painrack, how is a systematic failure not a market failure?

From google.
An economic term that encompasses a situation where, in any given market, the quantity of a product demanded by consumers does not equate to the quantity supplied by suppliers. This is a direct result of a lack of certain economically ideal factors, which prevents equilibrium.

Remember that before Ebola broke out, the entire world, including said countries affected by Ebola were more busy demanding resources for other medical conditions. Malaria. AIDS. TB. These were all important diseases that were badly hurting countries such as Nigeria.

Insisting that these countries divert scarce medical resources away from treating other diseases on the scope, to treat a POTENTIAL problem when NOBODY is demanding it....... how is that a market failure?

Are markets now prescient? The Free Hand of the market knows all?


Now, we elect experts and leaders to help prepare for potential problems. However, given the scope of other diseases, again, Malaria, TB, AIDs, just how is it justifiable, to put scarce public dollars into creating a vaccine for Ebola?
...
Now, Professor Hill idea does have merit. We should consider funding a subsidiary branch like say the WHO to increase resources into treating potential outbreak as opposed to the current mechanicism of monitoring and detecting outbreaks, a fire prevention service so as to speak as opposed to our current firefighting approach but that's not the fault of companies, because there simply were no buyers back THEN. If we had used public money to create what Professor Hill suggest, THEN we can start blaming the free market for not catering to said customer.

I think we're pretty close to agreement. The free market is just emergent behaviour. Agents participating in the free market can be prescient, it's why we tend to build power stations before the lights go out. As you said, that's the role of experts and leaders. In my previous posts I was pretty clear I blamed a lack of foresight on both sides. A drugs company, were they not beholden to short term shareholder interests could easily predict a large outbreak would create a market for a drug, and since the development speed is not quick, starting before the outbreak and having it on the books (or even better, being paid for vaccination work) would give them a huge lead on the competition. I also agree with you that funding a long horizon WHO department is probably the best way to create a stable market demand for such research.

At this very moment however, we have a huge unmet, yet reasonably predictable, market demand that the market cannot fulfil, and will be unable to fulfil for some time. By your own definition, that's a market failure.

What market demand?
There wasn't any demand for an Ebola vaccine. You asking companies to forsee that a potential demand exists for Ebola in 2014,so,let's invest in creating this in 2008...

[PainRack]

Tell me, do you think that if USA had rare, endemic disease striking only occasionally, but killing hundreds/thousands at each strike very messily, people would just shrug and focus on "more pressing issues"? Or would something be done with it, without waiting for benevolent market miracle?

The US can afford to spend 100 million dollars to investigate whether power lines cause Leukemia.
Nigeria doesn't have that money lying in its pockets to treat an EXISTING problem, just WHY the fuck would Nigeria have committed funds to purchasing or creating the demand for an Ebola vaccine?

Again. A market failure exists only when there's DEMAND for a product. There was no demand for an Ebola vaccine in previous years and I'm trying to stretch my mind to imagine how this could have been seen as a major failure, given the existing health crisises the world faced then.
Let's not forget that the WHO was facing H1N1, Avian Flu, MERs-COV, was still finishing up research into SARS and a whole bunch of emerging infectious diseases then, as well as completing the battle to crush polio and of course, the now forgotten battle against measles.

For Africia, TB, AIDs, Malaria.............


Just tell me how this is a failure of experts to have predicted that Ebola will emerge in 2014.

They are more pressing issues only because funds that could solve the problem go instead on researching kitten butt balding - as kitten owners in first world can pay more than Africans do anyway. You might shift goalpost "but that's will of people" but that's bullshit. Someone choose to move resources from critical fields to trivial ones trying to chase limited purchasing money. They put an option frittering away funds on trivial matters on table, without which people could have spent it more responsibly, by choice their own or administrative one.

And said demand exists.

You're trying to argue that a demand for Ebola vaccine existed before this epidemic. Show it to me. Then you can argue that a market failure exists.

Call you what you want, progress, economic growth, choice, but if more common good could be made on not doing something, then system encouraging it is flawed by any sensible definition and should be changed. People found out that denying a few extra non-essential luxuries to improve lot of many was almost always better choice thousands of years ago, sadly through the eons people on receiving end of forced redistribution for the good of society had also the means to buy most skilled liars to try and convince the others the 1% hoarding all the goods is in fact proper way to go and there is nothing bad in it.

Except that you SEEM to have forgotten that the world was also investing money in fighting malaria, TB and HIV. Problems that affect Africia and was the criticism of Professor Hills Ebola vaccine attack.


Let me just remind you of this again. There ARE existing research in medicine to fight diseases that affect Africia and the poor. Its just that said research was focusing on problems that were extent and major, NOT a potential problem in the future.

You're arguing that research should have been diverted away from other areas to help fund a potential research. Then PONY up the cash FIRST. Professor Hill request isn't flawed. We should create an agency to invest into treatment for emerging epidemics.

Of course, we SHOULD also note that the existing efforts of the world in the past decade was into avian flu, H1N1 and MERs-COR. Hell, MERS-COR isn't over yet!

Ah. So, we're finally getting somewhere. And just why their medical resources are scarce? Are they lazy, unwilling to just work harder, flips some burgers, get extra degrees? Or maybe, just maybe it has something to do with faults in the system?

Because they were focusing on OTHER diseases you mother-fucker.

Poland did not. Here, have a link:

http://www.flu-treatments.com/h1n1_vaccine.html

You know why? Because our experts didn't listen to siren song of big pharma and actually did think on what would be most sensible resource allocation, axing useless, superfluous spending.

And guess what? It worked! Not our dear hand of market, but an expert telling salesmen to shove it and do something useful instead!

HOW the FUCK does this prove your point?

Hey GENIUS. The H1N1 vaccine falls into the SAME FUCKING NICHE AS EBOLA VACCINE. Hell, the H1N1 vaccine was more fucking justifiable because there was an existing pandemic emerging when scientists ENGINEERED it.

There was NO Ebola epidemic emerging before this year. Just how many Polands would have rejected an Ebola vaccine?

[madd0ct0r]

There wasn't any demand for an Ebola vaccine. You asking companies to forsee that a potential demand exists for Ebola in 2014,so,let's invest in creating this in 2008...

Yes I am. It's pretty fucking simple to predict an outbreak at some point this decade, unlike the emergence of a brand new disease like SARS. It's (almost) exactly the same maths we use for predicting earthquake magnitudes, rainfall intensity, power grid peak demand sizes ek-fucking-cetra.

I've spent a few years of my life working on market niche exploration and the last 6 months on long horizon modelling, so perhaps this stuff isn't common knowledge, but a pharma corp should be able to say X amount to develop, Y amount profit from an outbreak in Z years time, followed by a likely immunisation program giving Y+ each year for a further Z years. Maybe these huge corps that can predict the upcoming exact demand for hair treatment based off demographics and the google adwords value of the word 'shampoo' missed a trick here?

maybe a developed vaccine means a promptly treated outbreak, which is worth less then an emergency injection of aid cash followed by sales to a much broader number of countries affected. but hey, that's business right?

[Esquire]

Do we actually have the data necessary for that kind of prediction, though? My understanding was that East Africa generally hasn't got the kind of functional, data-focused medical system that gives you comprehensive and continuous data to extrapolate from. Factor in all the other contributors to an Ebola outbreak, things like climate change making potentially-infected bush meat more plentiful or expanding the habitat for a particular species of Ebola-carrying bat - none of which we necessarily have useful data for - and the prediction models get a lot less trustworthy.

Plus, of course, pharmaceutical companies exist to make a profit and there's less of one in preventing diseases than in providing a vaccine after the 24-hour news cycle creates a panic.

[Simon_Jester]

Not actually knowing the data analysis 'industry' I can easily believe that maddoctor has expert knowledge that overrides Esquire's question... but it seems relevant. The success of 'big data' in analyzing markets and predicting market trends has a lot to do with collection of massive, comprehensive data and with virtually everything being recorded and written down and annotated.

With this, there are numerous hidden variables in play, some of them predictable (population density continuing to rise in Africa) and some not. It seems very unreliable to assume that we will just automatically know everything and be able to predict an Ebola epidemic well ahead of time, any more than we can predict earthquakes.

Sure, we know that powerful earthquakes can and will occur, but we have absolutely no idea when. With earthquakes, we design buildings to withstand the largest foreseeable quake precisely because ideally we want them to last forever. If the quake this building was designed to survive doesn't happen for 200 years, so much the better- if it does happen, we'll be ready.

With Ebola, though, not knowing how much time you have, and knowing that a vaccine cannot pay for itself unless there's an epidemic, does create an incentive to NOT panic and throw all your money at it. It's not like there aren't other pressing medical problems that need R&D funding to solve. It's not even as if many of these problems (like AIDS) don't affect poor countries or are somehow diseases of affluence.

[Esquire]

Just to clarify, I don't have any particularly relevant knowledge and will cheerfully concede to maddoctor's. I'm genuinely curious about the ways data feeds into vaccine funding - I spent the weekend writing grad school applications for epidemiology, so this topic is directly up my alley.

[madd0ct0r]

Fair points. I chose earthquakes as an analogy for that reason. Imagine each outbreak as a small perturbation to the system. We have good models for infective diseases spreading, so we can say reasonably confidently for a given infectivity and population density an outbreak of x size is the largest likely in a decade return period. Same as earthquakes. In essence the pharmaceutical Corp has to take a punt on the investment paying back within that time period. It's the sort of long range investment decision they have to make daily.

Esquire - I'm not necessarily talking about a model as detailed as that, I'm not sure we know enough about the carrier species(s). A simpler model of population density, slum population, infectivity range and possibly transport patterns should capture most of it. One of my colleagues was building a transport model of south Africa and ended up modelling aids in some detail: apparently there was a key tipping point where the population lost to the economy due to illness or caregizing destroyed the country's ability to develop, undermining the financial case for the motorway project. Bssically, this sort of modelling is done frequently and the capability is there.

I come into this from a civil Engineering background, I just got lost in marketing for a while.

[Esquire]

Ah, that makes perfect sense. I was thinking more of epidemiological models - my bad, and thanks for the explanation.

[PainRack]

Tell you what maddoctor, source me a statement from the WHO, or any other health entity that requested for an Ebola vaccine and I concede .

[madd0ct0r]

Have you not read a ducking thing I've written? Companies are allowed to anticipate market demand you know.

[Simon_Jester]

Fair points. I chose earthquakes as an analogy for that reason. Imagine each outbreak as a small perturbation to the system. We have good models for infective diseases spreading, so we can say reasonably confidently for a given infectivity and population density an outbreak of x size is the largest likely in a decade return period. Same as earthquakes. In essence the pharmaceutical Corp has to take a punt on the investment paying back within that time period. It's the sort of long range investment decision they have to make daily.

Yes, but...

Put this way. Often, the data that goes into models like this is inaccurate. For example, let's take a subject dear to your heart: hydrology, flood control, and distributing water for public use.

Most of the US west of the Appalachians is part of the drainage basins of the Mississippi, or of the West (i.e. west of the continental divide along the Rockies). In both those regions, the government did a lot of work surveying, estimating rainfall and aquifer levels, and distributing water rights in the early 20th century. Likewise, they built up flood control infrastructure based on historic trends.

The problem, it turns out, is that most of their planning was based on a period of time that was anomalous- high rainfall in the West, low burst rainfall in the Mississippi, and so on. The period in which scientific measurements were first taken and used was very different either from the centuries before it (we now know) or the decades after it.

So now, "once in a century" floods are happening every decade or so along the Mississippi, and towns protected by levees that were designed to stop anything BUT the 'flood of the century' are instead getting inundated twice in a generation. And water rights distributions that worked fine in 1950 are proving unsustainable today, with the result that (for instance) the Colorado River no longer reaches the ocean.

Likewise, while we can sort of predict what the biggest Ebola epidemic of the century will be, if our model turns out to be wrong then that epidemic might actually be the biggest of the millennium or the biggest of the decade. If it's the biggest of the millennium than a pharmaceutical company would be foolhardy to make a plan that relied on it happening, because the odds are 90% or more that it won't happen in the lifetime of the company. If it's the biggest of the decade, then they've just overestimated the amount of time they have to develop a cure.

Esquire - I'm not necessarily talking about a model as detailed as that, I'm not sure we know enough about the carrier species(s). A simpler model of population density, slum population, infectivity range and possibly transport patterns should capture most of it. One of my colleagues was building a transport model of south Africa and ended up modelling aids in some detail: apparently there was a key tipping point where the population lost to the economy due to illness or caregizing destroyed the country's ability to develop, undermining the financial case for the motorway project. Bssically, this sort of modelling is done frequently and the capability is there.

I come into this from a civil Engineering background, I just got lost in marketing for a while.

The problem is that while AIDS' dynamics were by that time well understood- we know how it spreads, how fast it spreads, what tends to retard its spread, and what encourages it. With Ebola, we knew in theory how this would work, but we've never observed it on anything like this scale. It would be very easy to construct a model like this and have it be inaccurate... as those early 20th century flood control and water rights planners learned in the US.

[Coop D'etat]

Regardless of accurate information about Sub-Saharran Africa, pharma development is notoriously unpredictable, being the land of trying out 100 good ideas to 10-20 years later end up with the one robust enough to work out in the field. Its not something where you can just budget 100 million for development and have much certainity of there being a useable product in a few years.

Pair an uncertain and expensive long-term development process with no certainity of there being much of a market at the end of the road and you can see why this isn't something private money wouldn't have that much interest in. Even if you're talking about social reponsibility in the thrid world, investing R and D into the likes of Malaria, AIDS and TB makes much more sense because you can be pretty sure there are millions of people who would be helped by it a generation from now.

Also keep in mind that in the biggest Ebola outbreak in history the death toll is still in the 4 digits. That's a rounding error for many third world infectious diseases. Ebola is more newsworthy for being a new and horrific way to die, rather than its massive mortality rate.

[madd0ct0r]

True, but there were a number of good trials happening around 2000 before the US defence industry pulled the plug. Googling gets several articles and papers, and I believe seaskimmer mentioned it earlier in the thread.

Ugh. Maybe I should put my money where my mouth is and build a model. 20 year horizon seems reasonable. Potential tripling of GDP in the region and less that for population change so it's not as unstable as some I've done

[Simon_Jester]

What I mean is, it's hard to be confident enough in the reliability of your projection to say "look, we practically know there's going to be a huge Ebola outbreak some time in the next fifteen to twenty years, so it'll pay off for us to spend huge sums and lock down key research personnel for that time working on a vaccine purely on spec."

Because if you turn out to be wrong and Ebola continues to be a disease associated with only tiny outbreaks, then you've just wasted a lot of money.

Whereas with other diseases that are known, ongoing killers (like AIDS), anything you spend has definite potential to pay off...

[madd0ct0r]

True, but are we discussing whether to fund one or the other OR whether market demand for Ebola could have been forecast? I acknowledge that pharma companies don't have unlimited research budgets so at some point a decision has to be made.

I'm surprised no-one's brought up this article yet, since it undermines a lot of my arguments http://www.bbc.co.uk/news/science-environment-19112510

I'm also looking at historic outbreaks - and it looks like a small village size outbreak was occuring every three to 4 years. The curve looks interesting, not many that are less then a dozen cases, although it's possible those never enter the statistics. http://en.wikipedia.org/wiki/List_of_Ebola_outbreaks