I really can't get my head around your logic. What does it matter to the drug companies what the Canadian legislation says or the purchasing policies within the healthcare regions. That point is entirely irrelevant in that there is no obligation whatsoever that demands they sell into Canada at all, let alone at a loss. Again, these company's are not charities and rarely will they sell at a loss unless that policy suits their individual needs. In the event that they are willing to sell to Canadians at a lower cost, well, all that indicates is that they are overcharging in the US market, not that they are subsidizing Canadians.
I pass no moral judgment on Canada's policies. I do not even know if the OP is correct. The author cites no sources nor presents any numbers. Rather, I am merely presenting an argument made by someone else. My response to the argument when I first heard it from a Canadian context was , "Yeah. So?"
From a purely rational standpoint, Canada is doing what it should be doing. If America is going to leave money on the table and Canada can pick it up, Canada would be stupid not to do so. That is what a good negotiator does.
However, Canadians should have an understanding how this all works. The piece was written for an American audience, given the debate that is raging down here on healthcare. The author - if he is factually true - is correct in that the US could not implement Canada's drug pricing policies. If America were to force the prices of drugs substantially lower, at some point, either drugs would be sold at a higher price in Canada or not be sold at all.
In terms of the economics being bandied about, your assumptions and description is incomplete. No where has there been mention of the reality that their R&D costs are amortized over the livable life of the patent or the nature of the fixed/variable expenses that they are including in the formula nor is there any focus attached to having the exclusive opportunity to manufacture and sell their product for a 25-50 window.
I think there is a couple of things. First, the ammortization of R&D costs are a sunk cost and thus a fixed cost. That cost must be recovered for the drug to be profitable. It is no different than a plant being built and then depreciated over a period of time.
Second, there is a significant fixed cost structure. If you take a quick look at an income statement of, say, Pfizer, you will see that Pfizer had $48 billion in revenues in 2008 and made $8 billion for a profit margin of 17%. Of the $40 billion cost structure, $8 billion is R&D and $5 billion is depreciation. $15 billion is selling, general and administrative expenses, of which at least half will be fixed. So at least half the cost structure is fixed. (It is here where I expect Nilfmir to start quibbling about whether or not each line item in the P&L is truly fixed or if everything is really variable.)
Finally, I do not think it is unfair to say that the patent protection affords the drug companies excess profits. The profit margins of drug companies are roughly double the average company in America. But even if you halved the profits by reducing prices such that the drug companies earned the average margin, you would lower Pfizer's revenues from $48 billion to $44 billion, or an 8% decline. And to answer what I'm sure will be Niflmir's argument, even if you gutted out $2-$3 billion from the cost structure - a significant amount for anyone who has analyzed the synergies of pharmaceutical mergers - you would be looking at a 15% decline, or the prices you paid roughly four years ago. So, it is not like there would be a dramatic fall in prices.