[darawk]

Yes, that is true, but irrelevant. A purchasing entity can price the drug fairly, and an entity doing original research can price the drug unfairly. Whether or not the IP was bought and sold is a non-sequitur.

What actually matters is that we have two problems:

1. How do we incentivize drug discovery?

2. How do we treat all the people that need treatment?

Right now our answer to #1 is that we allow people to charge the people that need treatment, for a limited time. This system works ok, but there may be better ones. If you have any, i'd love to hear them.

[nradov]

One approach would be to place an upper limit on the price of each drug or medical procedure based on the expected gain in Quality Adjusted Life Years (QALY) for an average patient. Although it's uncomfortable to discuss, we already place a dollar value on human life in civil wrongful death lawsuits. So we know roughly what an extra year of life is worth. And that value can be adjusted based on whether the patient is able to live a normal life or will still be bedridden (or other limitations) even with the treatment.

[darawk]

Ya, but the effect of that will be to diminish investment in treatments. If you can charge a million dollars per pill for a cancer cure, then you'd be willing to invest an enormous sum to find that cure. If your price is capped at $1000/pill, you will be willing to invest considerably less.

This isn't out of any malice or malfeasance, i'd like to add. This is just a rational calculation. A real person at the top somewhere is investing their own money in this, in the hopes that they'll be able to recover that in the marketplace, and presumably make a profit.

[waterhouse]

This article might be interesting reading: http://levine.sscnet.ucla.edu/papers/ip.ch.9.m1004.pdf

Skimming it, it doesn't really explain the means by which pharmaceutical companies succeed under weak or absent patent laws, it just says that they do. It seems to leave open the explanation, "Chemical and pharmaceutical companies in countries with weak or no patent laws make most of their money by selling into countries that have such laws." I haven't read it too carefully, though. It does say that they exaggerate their costs (presumably so as to argue for patent laws); that lots of those costs are perverse (e.g. bribing doctors); and that the patent laws inhibit innovation in lots of ways.

Meanwhile, I vaguely remember reading someone's proposal: Health insurance companies, at least in theory, basically make a bet that their customers will have good health. Let's suppose a health insurance company has 100k customers, and it's expected that 0.1% of them (let's say there's no way to tell who) will develop disease X sometime during the term of their insurance plan, and the treatment for disease X costs $50k. Then the expected total cost per customer is .001 * $50k = $50. The insurance company would therefore be fine agreeing to cover all costs of treating X for any insurance premium exceeding $50 over the term of the plan. Let's say they charge $70 to cover that particular item, and risk-averse customers agree to it. This is the classical notion of health insurance (I think in practice it's diverged significantly from that).

Once these plans are agreed to, the insurance company stands to bear an expected $50 * 100k = $5m of costs (for $7m in revenue). Now, if the insurance company has some way of, say, spending $2m to fund medical research that cuts the cost of treating X from $50k to $25k, then that would save them $2.5m in costs. (Making no reference to patenting or even selling the drug.)

Now, research is almost never guaranteed to produce the results you expect (arguably it wouldn't be called "research", merely "implementation"), and furthermore it takes time, perhaps longer than the term of the insurance plans. The uncertainty might be handled with a research consultant who makes "educated guesses" as to the expected value of different research paths. As for the time taken, (a) you could project how many insurance plans of this type you expect to have at the time the research completes, and (b) you could work something out with other insurance companies that provide similar plans. Which brings me to the biggest point:

I chose the figure of 100k customers. What if we imagine either a very large health insurance provider, or a joint effort by multiple insurance companies? Company A expects to have 5m insurance plans covering disease X when the research completes; company B expects 10m; company A will contribute $z and B will contribute $2z, or perhaps they'll just commit to funding the research and then find out how many plans they have when the research completes and divide the costs at that point. With a total of 15m customers, the research is now profitable if it costs anything less than 15m * $25 = $375m. The wider the consortium can be, the wider the range of treatments it will now become profitable to research.

[darawk]

> It does say that they exaggerate their costs (presumably so as to argue for patent laws); that lots of those costs are perverse (e.g. bribing doctors); and that the patent laws inhibit innovation in lots of ways.

That's certainly true. Pharma companies spend enormous sums of money on bribing doctors, so to speak (usually it's slightly indirect, because direct bribery is illegal). However, there's an interesting economics point here, which is that: A company should only spend a dollar on advertising it if expects to make 1 + epsilon dollars in return. Which is to say that each dollar of advertising spend shouldn't really be factored into the cost profile of the drug, since each marginal advertising dollar exists only to generate > 1 dollar of profit in return. This is a little bit difficult to wrap your head around, but it's true.

One way to summarize this is that while advertising dollars may cause prices to be higher than they otherwise would be, they don't impact what companies ought to be willing to invest in. Although, that's of course only if you assume that all drugs are equally marketable. Which is almost certainly false. Hmm...will have to think about this a bit.

> Now, if the insurance company has some way of, say, spending $2m to fund medical research that cuts the cost of treating X from $50k to $25k, then that would save them $2.5m in costs. (Making no reference to patenting or even selling the drug.)

This is an interesting idea, but one problem I see here is that it would incentivize insurance companies to silo and keep secret their medical research, because better treatment would be a competitive advantage for them, and cheaper treatment would allow them to charge lower premiums, which is also a big competitive advantage.

> The uncertainty might be handled with a research consultant who makes "educated guesses" as to the expected value of different research paths.

That's exactly what pharmaceutical companies do :).