> In France that seems to have been the case until recently, but not any longer

As your source notes, that does not affect "irreplaceable and expensive drugs". Cost-effectiveness concerns are typically about cases such as reimbursement for a generic vs. an equally-effective non-generic medicine.

> Putting compulsory licensing aside, which does not happen in many countries, I don't think this contradicts the poster above - a government agency makes a judgement about the cost-effectiveness of a particular drug, and they either decide to buy or not to buy depending on whether the price is above or below that limit.

Regarding compulsory licensing: contrary to what some people think, Europe isn't actually all that socialist; compulsory licensing exists as an ultima ratio in cases where the greed of a pharmaceutic corporation would endanger access to an important new drug or procedure (which was exactly the case in the BRCA case). It's a Sword of Damocles, not a weapon that's being wielded routinely. Whether the option actually exists is irrelevant: with a government-backed negotiator, it can always be legislated into practice (again, which is what happened in the BRCA case after a firestorm of criticism).

Regarding government agencies making a cost-effectiveness judgement about the purchase of a drug, that's again an inaccurate generalization.

First of all, recall that private insurance companies make such decisions all the time (see the article I referenced above), with the difference that they are interested only in their own bottom line, and that the health of their customers only enters into that calculation insofar as it affects profit or regulations force them to (the latter of which is the exact same situation as government backed-negotiation, except for the lack of the bargaining power of a monopsony).

Second, a minority of drugs are actually monopolies; price a drug for which there is competition out of the market, and you simply put money in your competitor's pocket. Likewise, drug companies don't just want to sell one single drug. Ask for too much for their one new irreplaceable drug, and they may see the income from their other ones go down (there is, after all, only so much money to be had in total).

Third, this is not necessarily how it happens. For example, it may not be a government agency (as in Germany, for example), and the agency may not have the power to make a decision whether to buy or not to buy a drug, but only whether the drug is effective (including relative effectiveness compared to other drugs), especially for life-saving drugs.

> As your source notes, that does not affect "irreplaceable and expensive drugs". Cost-effectiveness concerns are typically about cases such as reimbursement for a generic vs. an equally-effective non-generic medicine.

There is still a cost-effectiveness judgment even for irreplacable drugs (that's the SMR/ASMR category I):

"Two criteria must be met: 1) The ASMR of the drug claimed by the company is major, important, or moderate (ASMR I, II, or III, respectively); and 2) the drug is likely to have a significant impact on the health insurance budget regarding its impact on care organization, professional practices, or patient care and, when applicable, its price. This decree also specifies the medico-economic assessment procedure. The pharmaceutical company, together with its request for inclusion (or renewal of inclusion) of a medicine on the reimbursable drugs formulary, transmits all medico-economic data related to the drug to CEESP and CEPS. CEESP will provide an opinion [reported as a ‘flash opinion’ (14)] (Fig. 1) on the predictable or established efficiency of the drug and its coverage by health insurance. This opinion is based on comparative analysis, between the different therapeutic alternatives, of the ratio of the cost compared to the expected or observed benefit for patient health and quality of life."