It's not clear from the first article what sort of "research" is covered by the calculation, nor how they've come up with these figures. Lots of basic research

It's not clear from the first article what sort of "research" is covered by the calculation, nor how they've come up with these figures. Lots of basic research is done which furthers our understanding of biology; whether this research results in new drugs is unclear. ----- There are different routes by which new drugs are discovered, but very simplistically, we can break them down to two: 1) a researcher somewhere (not in Pharma) comes up with some positive research that might eventually result in the new drug for some disease. To develop this finding, often they will create (spin off) a new private company, and somehow generate funding. The extent to which the original funding institution benefits from this research is likely variable, based on the legal framework and contracts under which the researcher was operating. (And frankly, if a university or other institution doesn't benefit from such a discovery it funded in its labs... they're not doing something right.) 2) a pharmaceutical company funds its own discovery operations, which work to discover new targets and/or drugs for those targets -- #1 is unpredictable, but can result in great new discoveries which are brought to market --and thus patients-- relatively cheaply (like in the hundreds of millions, rather than billions) #2 is difficult, and very expensive, but mostly works - some pharmaceutical companies grind out drug after drug. However, ultimate success is less certain, and cumulatively, considering the cost of discovery operations as well as subsequent development, the cost is usually in the billions per successful new drug. This is also why we're seeing a switch to a mixed model: big Pharma is increasingly spending big on smaller biotech operations with promising drug candidates, as a way of minimising risk.